Backfield in Motion?

The Edmonton Oilers added Scottie Upshall on a PTO yesterday and things are far more interesting because of it. As I mentioned in an article for The Athletic Sunday, general manager Peter Chiarelli has a few more options than he did previously. The combination of Andrej Sekera’s injury and Upshall’s addition makes things…interesting.

THE ATHLETIC!

Give The Athletic as a gift or get it yourself and join the fun! Offer is here, less than $4 a month! I find myself reading both the hockey (Willis, Dellow, Pronman, et cetera) and the baseball coverage a lot, it’s a pure pleasure to visit. We’ll sell you the whole seat, but you’ll only need the edge.

WHAT’S GOING ON?

The two recent news items have changed things quite a bit, bringing all kinds of names into play. Andrej Sekera has moved from the “Certain” list to the “Distant Bells” list, and only because I don’t have a “Injured before camp” list to put him on.

CERTAIN OILERS (DOWN ONE NAME)

  • These 18 names are secure unless one of them gets traded for defensive help. The Sekera injury opens a window.
  • I mentioned yesterday about Peter Chiarelli’s quote surrounding his knowledge of the Sekera injury, and if it is long term that opens up an interesting set of possibilities. One of the names here might be headed out of town for a defenseman.
  • The kicker, the thing we could not see until now, is the player acquired could be an expensive blue and the Oilers might be able to send back an inexpensive young player (plus a pick) for that rearguard. This is a fascinating development and worth contemplating.
  • I think the goalie, all five defensemen, the centers and Nuge, Lucic, Rattie, Rieder and Puljujarvi are safe.
  • That leaves Zack Kassian, Jujhar Khaira and Drake Caggiula. You can see why a team would value each, for me losing Khaira and his inexpensive contract would be the least attractive option. Therefore, assume it is Khaira heading out.
  • One of the keys to the season will be Leon Draisaitl’s line. Jesse Puljujarvi would make life easy for his coach if he can slide up to No. 2 RW and allow Tobias Rieder to help the No. 3 line.

UNCERTAIN OILERS

  • Scottie Upshall has been added to the list and as of now there are five open jobs: Backup goalie, No. 6 and No. 7 defense, No. 13 and No. 14 forward. Edmonton may aim higher (trading for a second pairing defenseman like Torey Krug or Justin Faulk) but those are the jobs currently in play.
  • I believe Upshall has a job. I don’t think any of the other forwards is a lock.
  • I believe Mikko Koskinen has a job. I don’t think Al Montoya will be in the organization opening night.
  • I have no idea how the defense is going to shake out.

DISTANT BELLS

  • William Lagesson and Caleb Jones are the only two defensemen who have even a snowflake’s chance in hell among the blue.
  • Goal is completely up in the air after the NHL group.
  • I would nominate Tyler Benson, Joe Gambardella and Tyler Vesel as long shot NHL recalls for 2018-19.

POSSIBLE ROSTER 2018-19

  • Ryan Nugent-Hopkins—Connor McDavid—Ty Rattie
  • Milan Lucic—Leon Draisaitl—Tobias Rieder
  • Drake Caggiula—Ryan Strome—Jesse Puljujarvi
  • Scottie Upshall—Kyle Brodziak—Zack Kassian
  • Extras: Pontus Aberg, Brad Malone
  • Oscar Klefbom—Adam Larsson
  • Darnell Nurse—Justin Faulk
  • Kris Russell—Matt Benning
  • Kevin Gravel
  • Cam Talbot (Mikko Koskinen)

I think the Oilers would be interested in a trade that sends away a young inexpensive winger and a draft pick for Justin Faulk, even if there’s a risk later in the year re: Sekera coming back. It’s vital for this organization to make the playoffs, sending away Zack Kassian in February is a smallish price to pay (and I like Kassian but you get the point). Will this happen? I think it may.

10 SECONDS TO GLORY

Tyler Dellow wrote another tremendous article, it’s out today. If you’re an Oilers fan, it’s a must-read and offers incredible insight into one of last season’s most baffling statistics. Lordy.

LOWDOWN WITH LOWETIDE

We’re back, baby! At 10 this morning we hit the air with 10 days of pent-up verbal! Who knows how this will go???? Scheduled to appear, TSN1260:

  • Rod Pedersen, voice of the Saskatchewan Roughriders. A fabulous football game last night, SK winning over Calgary to end the Stampeders hopes of a 18-0 regular season. It also means the Labour Day classic in Calgary has some intrigue.
  • John Keim, ESPN reporter on the NFL’s Washington franchise. We’ll chat about the team, injuries, and the division.
  • Andy Strickland, Fox Sports Midwest. We’ll chat about the newest Oiler, Scottie Upshall, and what he’ll bring to the club this winter.

10-1260 text, @Lowetide on twitter. I’ll also be on the roundtable this morning with Dusty, Iwanyk and Yaremchuk. I’ll be trying to get a word in! See you on the radio!

written by

The author didn‘t add any Information to his profile yet.
Related Posts

134 Responses to "Backfield in Motion?"

  1. Jaxon says:

    I really don’t feel Rattie and Khaira belong on the certain list.

  2. Jaxon says:

    I’d be okay with Edmonton waiting for the preseason waivers and offering something to Toronto for Justin Holl. He may have been one of the best right hand D in the AHL last season he’s 26, he’s big (6’3″, 205lbs), he’s fast, he’s a defensibly responsible veteran pro, and last year he added an offensive element to his game. He had one of the best 5v5 points/60 (1.58), one of the best 5v5 primary points/60 (.73), one of the best 5v5 GF (59.77%), And he was logging top pair 5v5 minutes against the toughest competition (16.51 minutes). He’s also played some forward in the past so he’s quite versatile. I think he’d be a great addition as 3RD who could keep Russell on the left side. He’s the right hand D version of Josh Leivo, if you will.

  3. OriginalPouzar says:

    “One of the keys to the season will be Leon Draisaitl’s line. Jesse Puljujarvi would make life easy for his coach if he can slide up to No. 2 RW and allow Tobias Rieder to help the No. 3 line”

    ———————

    YES!

    This has been my position for a while. In my opinion Rieder is best suited to a middle 6/3rd line role and if he is being slotted in as a top 6 wingers then he’s being asked to perform above his established levels.

    Of course, Puljujarvi hasn’t performed at those established levels yet but he is a developing player and, to me, he is more of a top 6 talent and Rieder more of a middle 6/3rd line talent.

    To the extent we are looking for less line juggling and I’m sure we all are (even McDavid alluded to the same a few weeks back), here is hoping the coach gives this make-up a shot in pre-season and he hasn’t pigeon-holed Jesse in to start the season on the 3rd line and earning a move up in-season.

    My money is on Jesse earning at top 6 position if given the opportunity.

  4. OriginalPouzar says:

    Is there any way that Koskinen is not a “certain Oiler”?

    Even if outplayed by Montoya in preseason and exhibition, is there any realistic chance that, if healthy, management actually looks to re-assign Koskinen and keep Montoya on the roster?

    I mean, Koskinen was in the letter……

  5. Woodguy v2.0 says:

    Not that The Athletic needs any more pimping here, but this piece by Katie Strang outlines the issues for the upcoming CBA negotiations: https://theathletic.com/480706/2018/08/20/the-big-issues-facing-nhl-nhlpa-in-upcoming-cba-talks/?redirected=1

    Also,

    If you subscribe to The Athletic you should read everything Strang writes.

    She’s one of the best and covers topics that others do not.

    Fabulous writer.

  6. OriginalPouzar says:

    As far as Caleb Jones and Willie Lagesson go, I see quite a few posts (on various platforms) about Jones getting some NHL minutes this year. That very well might be the case but, personally, I have Lagesson over Jones as far as readiness for the NHL (not necessarily potential but readiness).

    I just see Lagesson as the more polished player at this point but I do acknowledge that Jones played in the tougher league last year (AHL vs. SHL).

    Jones really struggled to start last year but then got his game together and had a nice stretch before his injury and, after coming back from injury, he wasn’t quite as effective – Willis had a nice write up on this at the Athletic.

    To my understanding, Lagesson was simply trusted more and more by his coach throughout the year earning more ice as the year went on. I think he averaged 19 minutes and change but was often playing over 20 minutes down the stretch.

    It will be interesting to see how these two develop in Bakersfield this year.

    Hopefully the big club isn’t so devastated by injuries that we can allow these two fine prospects the year, or the better part of the year, to develop.

  7. OriginalPouzar says:

    Woodguy v2.0:
    Not that The Athletic needs any more pimping here, but this piece by Katie Strang outlines the issues for the upcoming CBA negotiations: https://theathletic.com/480706/2018/08/20/the-big-issues-facing-nhl-nhlpa-in-upcoming-cba-talks/?redirected=1

    Also,

    If you subscribe to The Athletic you should read everything Strang writes.

    She’s one of the best and covers topics that others do not.

    Fabulous writer.

    I saw this tweeted out earlier and am excited to read it later – Strang is indeed fantastic.

  8. OriginalPouzar says:

    To me the Upshall PTO doesn’t really change much as we all knew that Chiarelli was going to acquire another forward for the bottom of the lineup.

    Lots of talk out there about it being all but a certainty that he gets a contract and the plan is to play him (i.e. not the 13th/14th forward).

    That very well might be the case but here is hoping Chiarelli doesn’t rush in to it and gives the waiver wire a chance to provide some nice options – there always are.

  9. Woogie63 says:

    Jaxon:
    I really don’t feel Rattie and Khaira belong on the certain list.

    Do you think Connor is “bragging” to his dad about the great right winger the Oiler’s have found to make him better?

  10. nvan97 says:

    Woodguy v2.0,

    That article by Dellow is bonkers for the Oilers. I won’t spoil it for people who haven’t read but it’s absolutely crazy that they did what they did.

  11. Westchester Oil says:

    Because of Faulk’s second year of remaining term, I think I’d prefer to sign Enstrom for ~$2 million rather than trade for Faulk.

    Next best, IMO, would be a Faulk for Kassian plus either a 2nd or a 3rd or a Caleb Jones. Unloading Kassian’s contract would mitigate the risk of not being able to move Faulk’s contract in the summer of 2019.

    Another option if the Oilers feel they are comfortable with Bouchard or a waiver pick-up for the whole year, might be to use the LTIR money to beef up the forward ranks and sign Rick Nash to a 1 year, $3 mln contract – if he was willing to come play on 97’s team for that money.

  12. Woogie63 says:

    OriginalPouzar:
    As far as Caleb Jones and Willie Lagesson go, I see quite a few posts (on various platforms) about Jones getting some NHL minutes this year. That very well might be the case but, personally, I have Lagesson over Jones as far as readiness for the NHL (not necessarily potential but readiness).

    I just see Lagesson as the more polished player at this point but I do acknowledge that Jones played in the tougher league last year (AHL vs. SHL).

    Jones really struggled to start last year but then got his game together and had a nice stretch before his injury and, after coming back from injury, he wasn’t quite as effective – Willis had a nice write up on this at the Athletic.

    To my understanding, Lagesson was simply trusted more and more by his coach throughout the year earning more ice as the year went on.I think he averaged 19 minutes and change but was often playing over 20 minutes down the stretch.

    It will be interesting to see how these two develop in Bakersfield this year.

    Hopefully the big club isn’t so devastated by injuries that we can allow these two fine prospects the year, or the better part of the year, to develop.

    Last year the Oiler’s used 11 dman… The Washington Capitals and the Golden Knights each used 9 dman last year

  13. Melvis says:

    “I am quite confident it will give us reliable data on 26 dimensions of skating from real time play.”

    “Information isn’t knowledge isn’t wisdom.”

    A cynic might employ this data in a myriad of ways that have nothing to do with hockey, and everything to do with a contemporary surveillance economy. There’s the gold mine.

  14. dustrock says:

    Westchester Oil,

    Yes, Enstrom for nothing but money makes the most sense to me.

    Regarding Upshall, do you really think he has a definite job LT?

    Seems more like insurance to me, yes possibly with a mind to moving someone like Caggiula, but right now, is he a more effective hockey player than Caggiula, Aberg or Khaira?

    Don’t think so.

  15. Ribs says:

    Off the wall prediction I’ve had in my head for the last month or so. Pontus Aberg doesn’t play a regular season game with the Oilers this season. Another bet placed against him with Upshall coming to camp. It feels like the coaching staff are not fans of his work (ethic, maybe?). Complete speculation on my part. I hope he blows the training camp doors off and forces himself on to the roster. They could definitely use his speed.

    They can’t let Upshall play. I will never get it out of my head. Every time they say his name on television I hear (and usually have to shout out myself)

    “Scotty – Otty – Otty – Upshallllll!!!!!”

    Every time.

    I hate myself.

    lol.

  16. slopitch says:

    If the Oilers get Faulk for sweet fa, they make the playoffs. He’d help the PP.

  17. geowal says:

    Jaxon:
    I really don’t feel Rattie and Khaira belong on the certain list.

    So you think Khaira could be waived before the season begins? Not a chance in hell imo. I think they have given him a longer leash than that.

  18. VOR says:

    Melvis:
    “I am quite confident it will give us reliable data on 26 dimensions of skating from real time play.”

    “Information isn’t knowledge isn’t wisdom.”

    A cynic might employ this data in a myriad of ways that have nothing to do with hockey, and everything to do with a contemporary surveillance economy. There’s the gold mine.

    My associates keep calling it “spot the hidden weapon”. It comes from the fact that in game, though we are focused on the lower body we can tell if the player has changed sticks, added tape, has on new gloves, is sweating heavily during the game, etc. By the way, the choice of stick and taping actually has a real impact on many dimensions of skating. That is of course not news to anybody who plays hockey but the effect is more powerful than I would have guessed.

    Oh and hand grip and forearm strength are critical to some dimensions of skating. But I think even casual fans of hockey know that. Our work may be very helpful to people who coach young players. If nothing else it could be used with young players to provide feedback. Laval has already reached that stage. Though they can’t do it in game.

    However, we are years of hard work from anything predictive. No skater has yet measured the same from game to game. Skaters up to the junior and college level are not consistent. There are differences in shift, in period, in situation, in game, in season, and presumably from season to season. That is allowing for injury, fitness, equipment, quality of team, and quality of competition.

    There is this great line in Eiger Sanction (I think) about the hero “not qualified but passed” in Martial Arts training. There will always be hockey players who aren’t qualified in skating but passed.

    As an example, a junior player who was part of earlier test runs was drafted this year. He has some of the highest scores we’ve seen to date. But he is also among the most inconsistent skaters we have yet measured. Every scouting report talked about his lacklustre skating. The kid ranges from below 50th percentile to 99+ percentile within a single period. We don’t know why (no obvious explanation) but his skating biomechanics change literally from shift to shift in a pattern we couldn’t predict.

    In other words the more we learn the less we know. Surveillance might be a much easier nut to crack than hockey. My point stands, we are producing mountains of data but beyond the obvious teaching tool application we can’t offer coaches, scouts, or GMs any meaningful insight.

    To date when we’ve given players access to the film, data, and analysis their skating has gotten measurably worse. That is typical of the use of biomechanical feedback, it takes time to payoff, at first it just makes you more aware of what you are doing and you start thinking about it, and you can’t consciously think about a high speed physical skill. Thus, the drop off in performance at the beginning of the process. It can take months or even years to pay off.

  19. Westchester Oil says:

    Fun fact for a slow day:

    The screenwriter for ‘Crazy Rich Asians’ is named Peter Chiarelli.

    The GM for ‘Crazy, Overpaid, Up Against the Salary Cap, Oilers’ is also named Peter Chiarelli.

  20. Oilman99 says:

    geowal: So you think Khaira could be waived before the season begins? Not a chance in hell imo. I think they have given him a longer leash than that.

    Waiving Khaira would be making the same mistake they made with Brodziak, though I think JJ has a higher upside.

  21. Lowetide says:

    Oilman99: Waiving Khaira would be making the same mistake they made with Brodziak, though I think JJ has a higher upside.

    Waiving Khaira would be insane.

  22. Wilde says:

    VOR,

    What’s the visualisation process for this data?

    Is there one?

    There’s a drinking game to be had here if you could use it on an NHL game.

  23. VOR says:

    Several years ago I predicted that conventional hockey analytics would be replaced by tracking data and video analysis. Now we have reached that point. That isn’t to say there won’t be tons of work for data scientists and analysts sorting the wheat from the chaff and figuring out what it all means.

    I will now go out on a limb, however, and predict that the impact on hockey of the wide scale use of tracking data and digital video analysis will be essentially zero. Any impact will be front end loaded as early adopters and skilled adapters gain advantage and then lose it as every team starts using that same data in that same way. The race for advantage will move on, probably into neurological and psychological approaches.

    Biomechanical analysis holds out some hope for significant advancement given it is in its infancy in hockey. However, applying biomechanical knowledge in effective ways in hockey will be a real challenge. But because it could be integrated into both scouting and coaching, particularly at lower levels I am sure resources will continue to flow into hockey biomechanics.

    You can see all this played out in the amazing body of work produced on the NHL draft in the last decade. I believe you can see a real improvement overall in first round drafting. But there have been some off setting loses in later round efficiencies. What there has not been is huge breakthroughs.

    We can’t even agree on who the top ten drafting teams are. The winners should leap out from the data. The advantage should be obvious and consistent. Michael has shown some teams can outperform Central Scouting consistenly but Namita has shown that no single team is closer to perfect drafting than any other.

    Many people in the analytics community believe if we had more stats from junior levels we could do a better job of drafting. This is almost the exact opposite of what will happen. More data merely means more data points and more confusion over ranking. Though it will take years for the analytics community to acknowledge this. Data is to data scientists what crack cocaine is to addicts. Thus all data is good data, we’ll just develop cleverer ways of figuring out what it all means. Yeah, sure!

    I am not sanguine about the prospects of my own approach to drafting. First, I am an amateur. Second, I know AIDA is producing a bizarre mix of real insight and utter non-sense, very much like real life scouts. I have no way of knowing in advance which is which.

    I think anybody who doesn’t acknowledge that the advanced stats community doesn’t bat for average, rarely hits a home run, and spends most of its time striking out and hitting into double plays has their head buried deep in the sand.

    That does not mean I hate numbers. Or that I personally plan to stop using possession metrics and other advanced stats. It does however mean that I think the advanced stats community should welcome criticism and embrace doubters. Yet defensiveness remains the default position.

    I also think that some methodological improvements are passed due. But that is an entire other post.

  24. VOR says:

    Wilde:
    VOR,

    What’s the visualisation process for this data?

    Is there one?

    There’s a drinking game to be had here if you could use it on an NHL game.

    We use a “modified” version of Mokka. The quotations are because I am pretty sure Mokka’s creators wouldn’t recognize their baby now. We are also moving into some virtual reality work but it is very clunky at this point.

    As a company we have wanted for years to be able to give home viewers the ability to control their gaze (turn your head and move the cameras as it were). Also to see the game from each players view point. Not to mention to pull up real time data on each players in game, season, and career data. I think we first discussed it with scientists at the UofA in 2000. It may have been even earlier.

    Imagine watching tracking and biomechanical data piling up in real time.

  25. OriginalPouzar says:

    Jaxon:
    I really don’t feel Rattie and Khaira belong on the certain list.

    I agree with Rattie but disagree with respect to Khaira – in my mind he established himself as an every day NHL player last year.

    Not to mention his league minimum cap hit is a huge value.

  26. Wilde says:

    VOR: the impact on hockey of the wide scale use of tracking data and digital video analysis will be essentially zero. Any impact will be front end loaded as early adopters and skilled adapters gain advantage and then lose it as every team starts using that same data in that same way. The race for advantage will move on, probably into neurological and psychological approaches.

    This is a statement that’s missing some initial framing to make it work:

    By ‘impact’ here you’re meaning to refer to the competitive advantages teams will have that others do not.

    But, outside of that narrowing, I think what we’re going to see is evolutions in how the game is played and taught, in any and every competitive environment.

    (notice how I didn’t mention which way)

    The length of this process will also reach until the end times.

    VOR: We can’t even agree on who the top ten drafting teams are. The winners should leap out from the data. The advantage should be obvious and consistent. Michael has shown some teams can outperform Central Scouting consistency but Namita has shown that no single team is closer to perfect drafting than any other.

    Agreed on this, but how are we going to get through the swamp that is development?

    VOR:
    Many people in the analytics community believe if we had more stats from junior levels we could do a better job of drafting. This is almost the exact opposite of what will happen. More data merely means more data points and more confusion over ranking. Though it will take years for the analytics community to acknowledge this. Data is to data scientists what crack cocaine is to addicts. Thus all data is good data, we’ll just develop cleverer ways of figuring out what it all means. Yeah, sure!

    Hard, hard disagree.

    Time on ice, VOR.

    That’s just one.

    Also, non ‘quality’ stuff like usage could be sussed out.

    VOR: I think anybody who doesn’t acknowledge that the advanced stats community doesn’t bat for average, rarely hits a home run, and spends most of its time striking out and hitting into double plays has their head buried deep in the sand.

    This is entirely too extreme.

    People are making very good money predicting hockey games with their models.

    VOR: That does not mean I hate numbers. Or that I personally plan to stop using possession metrics and other advanced stats. It does however mean that I think the advanced stats community should welcome criticism and embrace doubters. Yet defensiveness remains the default position.

    The vast majority of the main contributors to the advanced stats community’s defensiveness arises solely and only in response to bad faith criticism from people who put more time into typing up said critique(still not very much) than they did reading the material they’re attacking.

  27. LadiesloveSmid says:

    Jerabek seems like he’s got the skill level of Gravel, no?

  28. BornInAGretzkyJersey says:

    Edmonton Oilers Retweeted

    Bob Stauffer

    Verified account

    @Bob_Stauffer
    1m1 minute ago
    More Bob Stauffer Retweeted Edmonton Oilers
    Jerabek gets a 1-way deal.
    Can move the puck a little.
    6/7 D man with a some upside

  29. BornInAGretzkyJersey says:

    Jakub Jerabek
    Defense — shoots L
    Born May 12 1991 — Plzen, Czech Rep.
    [27 yrs. ago]
    Height 5.11 — Weight 200 [180 cm/91 kg]

    2017-18 Montreal Canadiens NHL 25 1 3 4 6 -1
    2017-18 Washington Capitals NHL 11 1 3 4 0 -1

    Source: http://www.hockeydb.com/ihdb/stats/pdisplay.php?pid=120313

  30. Lowetide says:

    The #Oilers have signed Jakub Jerabek to a one-year contract. The 27-year-old Czech defenceman split time between Montreal & Washington in 2017-18, including two playoff appearances with the Capitals.

  31. BornInAGretzkyJersey says:

    Assets:
    Is mobile and an excellent puck mover from behind the blueline. Can also post excellent offensive numbers at lower levels. Can play a physical brand of hockey, as well.
    Flaws:
    Must prove capable of making an impact in North America, specifically at the highest level. Lacks ideal size for an NHL defender, so he may struggle to play a physical game.
    Career Potential:
    Solid, puck-moving defenseman with a little upside.

    Source: http://forecaster.thehockeynews.com/player/7659

  32. OilClog says:

    This D isn’t setting up well for fast puck movement, maybe Yawney is bringing the Flying V to town.

  33. LadiesloveSmid says:

    LadiesloveSmid:
    Jerabek seems like he’s got the skill level of Gravel, no?

    I wouldn’t venture he’s as good as Auvitu..?

  34. russ99 says:

    OilClog:
    This D isn’t setting up well for fast puck movement, maybe Yawney is bringing the Flying V to town.

    The wingers are is as much to blame. See Dellow’s post today on Athletic. Good video of a total lack of positional awareness or where they need to be.

  35. godot10 says:

    russ99: The wingers are is as much to blame. See Dellow’s post today on Athletic. Good video of a total lack of positional awareness or where they need to be.

    Don’t worry, the guy coaching the forwards last year is the same guy who will be coaching (for the first time ever) our entire prospect pool in Bakersfield. And the head coach in Edmonton is still the head coach. Everything will be fine.

  36. hunter1909 says:

    I saw a pretty girl with a puppy today. Normally I’m shy around women but the puppy was cute so I told her so. Then we both go into the local store and I keep bumping into her, as well as another hottie who was wandering through the store’s narrow aisles.

    I love summer.

  37. Side says:

    godot10: Don’t worry, the guy coaching the forwards last year is the same guy who will be coaching (for the first time ever)

    Troll.

  38. OilClog says:

    russ99: The wingers are is as much to blame. See Dellow’s post today on Athletic. Good video of a total lack of positional awareness or where they need to be.

    Sorry but can’t blame the wingers for the cement feet D behind them.

    They signed a guy that was being healthy scratched while the Montreal D was looking uglier then Edmonton’s.

    Maybe Russ can do some circle backs and 47 resets before he decides to pass the puck off the boards.

  39. OriginalPouzar says:

    Westchester Oil:
    Because of Faulk’s second year of remaining term, I think I’d prefer to sign Enstrom for ~$2 million rather than trade for Faulk.

    Heck ya – that $4.8M cap hit for the 2019/20 season is not something this team can take on without subsequent cap disposition and there is too much risk involved, for me.

  40. VOR says:

    Wilde: This is a statement that’s missing some initial framing to make it work:

    By ‘impact’ here you’re meaning to refer to the competitive advantages teams will have that others do not.

    But, outside of that narrowing, I think what we’re going to see is evolutions in how the game is played and taught, in any and every competitive environment.

    (notice how I didn’t mention which way)

    The length of this process will also reach until the end times.

    Agreed on this, but how are we going to get through the swamp that is development?

    Hard, hard disagree.

    Time on ice, VOR.

    That’s just one.

    Also, non ‘quality’ stuff like usage could be sussed out.

    This is entirely too extreme.

    People are making very good money predicting hockey games with their models.

    The vast majority of the main contributors to the advanced stats community’s defensiveness arises solely and only in response to bad faith criticism from people who put more time into typing up said critique(still not very much) than they did reading the material they’re attacking.

    I did say it will take the analytics community a long time to acknowledge the data smog problem. But let’s rift on TOI data for a moment. Is 100 points acquired in 30 minutes a night in 60 games more or less predictive of a future NHL star than 100 points in 25 minutes a night in 60 games? Why? What compounding factors might there be? How do we allow for those? For that matter are points per minute played predictive of points per minute in the NHL. Sounds like you’ve made the assumption that it is. But where is the proof?

    I can go on and on if you’d like. Or you could just acknowledge time on ice data will just give us another thing to argue about but in a game like drafting provide no advantage because it will be less than definitive and everybody will have the same data. It will just add complexity.

    But I am quite content to leave the argument and return in a decade and say, see I told you so. It took five years for people to admit I was right that Corsi is a poor way of evaluating and rating Dmen. And there are still hold outs. I am looking at you, LT.

    As for people making money predicting the outcomes of hockey games while I don’t gamble I can consistently outperform chance at picking the winners of NHL games. All I need to know is venue, referees, and streaks. Try it for yourself.

    I am not sure why you’d think you need fancy stats or that gambling success was prove of value in hockey analytics. Every successful system gambler and horse racing and baseball produce quite a number is an inveterate lire and you couldn’t pry their secrets from them by tearing their fingernails off. They also all know their success is a result of being smarter than the mugs who make up the vast majority of the gambling population. They aren’t about to let the mugs into the game.

    But that does bring up one of my points about methodology in hockey analytics. The lack of publication of the majority of the work in hockey analytics in peer reviewed academic journals. The reason I use referees, venue, and streak information to predict hockey games is because there are peer reviewed papers, multiple papers actually, that explain the relationship and give reliable insight into the methodology. I can be reasonably confident in the workmanship.

    Why won’t the hockey analytics community submit their work to be scrutinized by their peers in a proper review format? Or do they submit and get rejected over and over again. Every other major professional sport’s analytical community has a voluminous published research literature. Why not hockey?

    Where are the Stanley Cup winning teams that are throwing parties for their analytics departments. That isn’t to say they don’t have analytics departments just that they are part of an integrated whole. Many bad teams have large analytics departments. But the analytics community seems determined to blame all their failures on poor management and champion their modest successes.

    The game of course will to be to tear me down. Which merely proves my fundamental contention which is that the hockey analytics community is antagonistic to even moderate criticism.

    Just wait until I get around to criticizing their basic methodology.

  41. VOR says:

    Wilde: This is a statement that’s missing some initial framing to make it work:

    By ‘impact’ here you’re meaning to refer to the competitive advantages teams will have that others do not.

    But, outside of that narrowing, I think what we’re going to see is evolutions in how the game is played and taught, in any and every competitive environment.

    (notice how I didn’t mention which way)

    The length of this process will also reach until the end times.

    Agreed on this, but how are we going to get through the swamp that is development?

    Hard, hard disagree.

    Time on ice, VOR.

    That’s just one.

    Also, non ‘quality’ stuff like usage could be sussed out.

    This is entirely too extreme.

    People are making very good money predicting hockey games with their models.

    The vast majority of the main contributors to the advanced stats community’s defensiveness arises solely and only in response to bad faith criticism from people who put more time into typing up said critique(still not very much) than they did reading the material they’re attacking.

    The final statement really deserves a post of its own. I can personally attest to it being false. When I dared to suggest Darnell Nurse might be a valuable NHL defensemen several years ago I got ripped into by Woodguy and GMoney and LT told me the smartest guys he knew told him Nurse would already have arrived if he was ever going to (these smart guys based this on where Darnell Nurse was drafted – true story). It was clear nobody was bothering to read and think about my arguments. So go preach the gospel of the saintly, picked upon, hockey analytics community to the choir.

    For now I need to focus on solar farms so I will leave it there.

  42. OriginalPouzar says:

    Woogie63: Last year the Oiler’s used 11 dman… The Washington Capitals and the Golden Knights each used 9 dman last year

    Yup, even in our playoff year when we were relatively healthy I think we used 12.

    This is another reason why I’m stupified by the Gryba buyout.

    We are set up for each of Keegan Lowe, Ryan Stanton, Kevin Gravel and Jones and/or Lagesson to get NHL minutes (in addition to Bear who could likely use another year prior to being forced in to action).

  43. Revolved says:

    VOR,

    I would like to hear about your criticisms of the basic methodology of hockey analytics!

  44. OriginalPouzar says:

    Ribs:
    Off the wall prediction I’ve had in my head for the last month or so. Pontus Aberg doesn’t play a regular season game with the Oilers this season. Another bet placed against him with Upshall coming to camp. It feels like the coaching staff are not fans of his work (ethic, maybe?). Complete speculation on my part. I hope he blows the training camp doors off and forces himself on to the roster. They could definitely use his speed.

    I have a similar feeling (re: McLellan’s thoughts on the player). I have nothing to go on but I just “feel” that they won’t give him a legit shot at a top 6 position and I think its b/c of the missed practice incident last year. Again, that is pure baseless speculation but I just have this “feeling”.

    It would be disappointing as he scored at 1.99 P/60 as an Oiler (granted it was almost all in a 3-game heater) and, similar to Khaira, at league minimum, that contract has the potential to be huge value.

    I hope he gets a real shot.

  45. leadfarmer says:

    OriginalPouzar: Yup, even in our playoff year when we were relatively healthy I think we used 12.

    This is another reason why I’m stupified by the Gryba buyout.

    We are set up for each of Keegan Lowe, Ryan Stanton, Kevin Gravel and Jones and/or Lagesson to get NHL minutes (in addition to Bear who could likely use another year prior to being forced in to action).

    Willis’s article on Jones has the GF% numbers from the AHL team. Guess who was dead last?
    Gryba with 34.4 %
    I think that explains the buyout fairly well

  46. OriginalPouzar says:

    slopitch:
    If the Oilers get Faulk for sweet fa, they make the playoffs. He’d help the PP.

    I’m not even sure he’s a 3rd pairing NHL d-man at evens though….

  47. flea says:

    leadfarmer: Willis’s article on Jones has the GF% numbers from the AHL team.Guess who was dead last?
    Gryba with 34.4 %
    I think that explains the buyout fairly well

    I also think they didn’t want Gryba in the organization anymore. Everyone talked about them doing Gryba a “solid”, but a buyout is hockey’s version of getting fired. It would have been more of a solid to him to let him compete for a job and collect his whole paycheck. The relationship had obviously broken down.

    Is Jerabek the Sekera response? Gawd I hope not! This already is starting to feel eerily like last year, hope i’m wrong though!

  48. stephen sheps says:

    VOR,

    Any chance we can chat offline about some of this stuff? (stephen dot sheps at ryerson dot ca)

    LT – apologies if making such requests is bad form on the blog. Let me know if that’s not ok. Thanks to you both.

  49. geowal says:

    Does this Jerabek fellow mean the trade scenario is pretty much done?

  50. Lowetide says:

    geowal:
    Does this Jerabek fellow mean the trade scenario is pretty much done?

    I don’t think so, seems a reasonable signing for the 6-7D, competition for Gravel, Lowe, et cetera. I think he might be a favourite for No. 7D but that’s a guess.

  51. bendelson says:

    hunter1909:
    I saw a pretty girl with a puppy today. Normally I’m shy around women but the puppy was cute so I told her so. Then we both go into the local store and I keep bumping into her, as well as another hottie who was wandering through the store’s narrow aisles.

    I love summer.

    One of my dogs is a blue merle Australian Sheppard with those ice blue eyes… a fine looking pup, to be sure. Once in a while, when ladies walk by, one will say to the other something along the lines of “Wow! He’s sooooo cute” to which I love to brazenly retort, “Thanks! What do you think of my dog?” The momentary panic in their eyes is priceless I tell you… priceless!

  52. Woodguy v2.0 says:

    VOR: The final statement really deserves a post of its own. I can personally attest to it being false. When I dared to suggest Darnell Nurse might be a valuable NHL defensemen several years ago I got ripped into by Woodguy and GMoney and LT told me the smartest guys he knew told him Nurse would already have arrived if he was ever going to (these smart guys based this on where Darnell Nurse was drafted – true story). It was clear nobody was bothering to read and think about my arguments. So go preach the gospel of the saintly, picked upon, hockey analytics community to the choir.

    For now I need to focus on solar farms so I will leave it there.

    To be clear:

    People were projecting Nurse as a monster top pairing Dman.

    ~95% of those Dmen announce their presence in the NHL with authority.

    I was on record then and am now saying that he is not tracking to be that.

    His breakout season in 17/18 was excellent and now I’m comfortable saying he probably has a long career as an Actual Top 4 NHL Dman.

    He still isn’t tracking to anchor a top pair.

  53. Zelepukin says:

    geowal:
    Does this Jerabek fellow mean the trade scenario is pretty much done?

    It does make it feel like Chia is again looking for a Hail Mary D-man to step up and either make the team as a rookie or establishing themselves as a top 6 D for the first time in their career.

  54. Wilde says:

    VOR: Sounds like you’ve made the assumption that it is. But where is the proof?

    I intentionally left out any mention of a second factor in the equation to let you paint the target if you so chose, and you applied such snarl in a refutation of a statement you made up(I did not mention points) that it’s entirely demonstrative of the poor faith you meant to engage me with.

    VOR: but in a game like drafting provide no advantage because it will be less than definitive and everybody will have the same data.

    This could indeed be so and a valid contention if I was making a statement that a team would acquire a competitive advantage out of superior interpretation of the same data, but I did not.

    When you were refuting “we could do a better job of drafting”, I took it to mean ‘we’ was the NHL in general, and was applying the same concept of the tracking data; widening it to not an increase in one team’s abilities over the other but to a general increase in efficiency in drafting across the board.

    I apologise for the confusion caused by not very literally making that connection and distinction.

    (Jaxon’s work created an interesting dialogue regarding his TOI/GP cutoff when he was trying to eliminate common threads for point producers that busted)

    VOR: But I am quite content to leave the argument and return in a decade and say, see I told you so

    I could certainly see us returning to the scorched battlefield littered with your target dummies’ corpses, and I could certainly see you projecting what you believe I said onto them again just as you’ve done now.

    VOR: As for people making money predicting the outcomes of hockey games while I don’t gamble I can consistently outperform chance at picking the winners of NHL games. All I need to know is venue, referees, and streaks. Try it for yourself.

    And in doing this, if you were outperformed in turn by someone using a model, would you consider that to be striking out most of the time?

    VOR: being smarter than the mugs

    And by what ability are they smarter than the mugs? Again, it’s not just venues, referees and streaks.

    VOR: Why won’t the hockey analytics community submit their work to be scrutinized by their peers in a proper review format? Or do they submit and get rejected over and over again. Every other major professional sport’s analytical community has a voluminous published research literature. Why not hockey?

    This is a good question I don’t have the information to answer, in fact I don’t know if the premise is true. Thank you for the jumping off point.

    VOR: But the analytics community seems determined to blame all their failures on poor management and champion their modest successes.

    This is an attempt to hold the leaders (and the people worth engaging with) accountable for those who are members to the community at large by your own distinction.

    Similar to earlier when you asked me to defend things I didn’t say, you here ask people to defend things said by others they don’t agree or identify with.

    You can do this type of critique only when a community has hard boundaries and hard association, there’s no way for say, a model maker to police some random from using his model in an arg improperly or demonstrating poor faith

    VOR: The game of course will to be to tear me down. Which merely proves my fundamental contention which is that the hockey analytics community is antagonistic to even moderate criticism.

    This is pre-emptive self-victimisation to shield yourself from someone responding eye-for-eye to your imbalanced takedown attempts.

    What you’ve said this morning only loosely translates to criticism of ideas held by others, and in no way adds up to moderate. It is your posts that are antagonistic.

    VOR: The final statement really deserves a post of its own. I can personally attest to it being false. When I dared to suggest Darnell Nurse might be a valuable NHL defensemen several years ago I got ripped into by Woodguy and GMoney and LT told me the smartest guys he knew told him Nurse would already have arrived if he was ever going to (these smart guys based this on where Darnell Nurse was drafted – true story). It was clear nobody was bothering to read and think about my arguments. So go preach the gospel of the saintly, picked upon, hockey analytics community to the choir.

    Forgive me if your recollection of this argument isn’t entirely believed by myself. Given the twisting you’ve attempted on my own words in just the short time it’s been since I’ve written them, I could easily see much more muddying proportionate to the much longer time it’s been since.

    This was ugly, VOR, and I didn’t enjoy any part of this interaction, which is especially lame given the usual enjoyment I derive from speaking with you.

  55. hunter1909 says:

    Zelepukin: It does make it feel like Chia is again looking for a Hail Mary D-man to step up and either make the team as a rookie

    Oh, you’re talking about the number 2 drafted Canadian(yes, some of us still count Canadians as special players).

  56. Woodguy v2.0 says:

    Lowetide: I don’t think so, seems a reasonable signing for the 6-7D, competition for Gravel, Lowe, et cetera. I think he might be a favourite for No. 7D but that’s a guess.

    His results with Benn suggest he might be able to handle 2nd pair.

    Benn was solidly 2nd pair QoC wise in MTL and he played mostly with Benn and did ok there.

    Played mostly with Orpik in WSH so those results are meh and you can’t come to conclusion via those results.

    You’re right that it doesn’t close the door to a trade, but I think Pete is done.

    Check Jon Willis’ twitter feed for some good links with Athletic MTL writers pieces on him.

  57. Woodguy v2.0 says:

    This might be a really good signing.

  58. Wilde says:

    We need to clone Larsson and just make every 1 on 1 drill for forwards to be played vs. a Larsson.

  59. godot10 says:

    Woodguy v2.0: To be clear:

    People were projecting Nurse as a monster top pairing Dman.

    ~95% of those Dmen announce their presence in the NHL with authority.

    I was on record then and am now saying that he is not tracking to be that.

    His breakout season in 17/18 was excellent and now I’m comfortable saying he probably has a long career as an Actual Top 4 NHL Dman.

    He still isn’t tracking to anchor a top pair.

    A physical Jay Bouwmeester.

  60. Wilde says:

    Woodguy v2.0:
    This might be a really good signing.

    I think it’s good already, just based on what we know, but the coaching staff will have to like him for it to turn into a true success

  61. leadfarmer says:

    Woodguy v2.0,

    Although you weren’t very high on Seth Jones either

  62. leadfarmer says:

    Woodguy v2.0,

    Anything at this point that doesn’t mortgage futures or have future cap implications is a good signing.
    Although how much does this leave for Nurse contract?

  63. Lowetide says:

    Woodguy v2.0: His results with Benn suggest he might be able to handle 2nd pair.

    Benn was solidly 2nd pair QoC wise inMTL and he played mostly with Benn and did ok there.

    Played mostly with Orpik in WSH so those results are meh and you can’t come to conclusion via those results.

    You’re right that it doesn’t close the door to a trade, but I think Pete is done.

    Check Jon Willis’ twitter feed for some good links with Athletic MTL writers pieces on him.

    Yeah, I read both but man we’re talking small sample sizes. Good bet to make, I’ll agree on that though.

  64. blainer says:

    Man people have long memories here..

    I can barely remember what I had for lunch…

    Damn… the 1970’s were good ..

  65. Ivan says:

    Do we know what Jerebek’s cap hit is yet?

  66. Pescador says:

    Woogie63: Last year the Oiler’s used 11 dman… The Washington Capitals and the Golden Knights each used 9 dman last year

    Quality>Quantity
    We have 3 top 4 D
    And 7 sevens, again
    Sans balance

  67. leadfarmer says:

    Does Markov want to come back to the NHL?

  68. jp says:

    hunter1909: Oh, you’re talking about the number 2 drafted Canadian(yes, some of us still count Canadians as special players).

    You know, Mitch Moroz was the number 15 drafted Canadian in 2012.

  69. JimmyV1965 says:

    OriginalPouzar: I have a similar feeling (re: McLellan’s thoughts on the player).I have nothing to go on but I just “feel” that they won’t give him a legit shot at a top 6 position and I think its b/c of the missed practice incident last year. Again, that is pure baseless speculation but I just have this “feeling”.

    It would be disappointing as he scored at 1.99 P/60 as an Oiler (granted it was almost all in a 3-game heater) and, similar to Khaira, at league minimum, that contract has the potential to be huge value.

    I hope he gets a real shot.

    I’m not sure where this is coming from. Aberg played 18 min his first game back from suspension, on the second line with Drai. He had a great game and continued to play well and kept playing 18 min. It didn’t last long though and his play faded and his time was reduced to about 13 min. His play continued to fade and he was finally benched near the end of the season.

    Aberg’s role on the team will be determined by Aberg. He has to show up game in and game out. He’s just another prospect who has to do something to make himself stand out from the rest of the prospects. If not, he’s out the door.

  70. OriginalPouzar says:

    hunter1909: Yet, you seriously question my stating that Bouchard has a shot at the Oilers next season.

    Bouchard, who already looks 31 years old, will arrive at training camp as the flat out puck moving defenceman and for this desperate team I’m sure he makes the cut.

    I’ve never questioned that Bouchard has a shot at making the Oilers – not once.

    In fact, I’m quite confident he will be there in Gothenburg.

  71. OriginalPouzar says:

    leadfarmer:
    Woodguy v2.0,

    Anything at this point that doesn’t mortgage futures or have future cap implications is a good signing.
    Although how much does this leave for Nurse contract?

    This won’t effect the cap at all. If he’s on the roster in will be in replacement of another d-man that would have otherwise been on the roster (and who’s cap hit will disappear – if under $1M).

  72. Woodguy v2.0 says:

    Lowetide: Yeah, I read both but man we’re talking small sample sizes. Good bet to make, I’ll agree on that though.

    Good bet.

    RiversQ says he has slow feet (Belovesque I’d assume)

    I don’t know.

    Never seen em.

  73. Woodguy v2.0 says:

    Here’s one link on Jerabek: https://theathletic.com/249153/2018/02/21/jakub-jerabek-went-from-shutdown-to-shipped-out-in-a-month/

    Here’s another: https://theathletic.com/248788/2018/02/21/jakub-jerabek-played-good-hockey-relative-to-his-teammates-and-stuff-like-that/

    Neither writer does much to couch the results in QoC, but he played mostly with Benn which was 2nd pair.

    His results with Benn have me hopeful that he can cover in the Top 4 when injury hits.\

    I expect:

    77-6
    25-4
    Jerabek-83
    Gravel

    Also,

    Jerabek was a 1st team all star in the KHL.

    He’s motivated to play in the NHL if he re-signed in the NHL.

    That means something.

  74. leadfarmer says:

    OriginalPouzar: This won’t effect the cap at all. If he’s on the roster in will be in replacement of another d-man that would have otherwise been on the roster (and who’s cap hit will disappear – if under $1M).

    But isn’t minimum salary like 650k or something like that. Keegan Lowe has 675k cap hit and we don’t know what Jerabek has but If like you said is a million if Keegan Lowe replaces him you still lose 325k.

  75. Bank Shot says:

    leadfarmer: But isn’t minimum salary like 650k or something like that.Keegan Lowe has 675k cap hit and we don’t know what Jerabek has but If like you said is a million if Keegan Lowe replaces him you still lose 325k.

    If Keegan Lowe replaces anyone the Oilers will lose out on a lot of W’s.

  76. hunter1909 says:

    jp: You know, Mitch Moroz was the number 15 drafted Canadian in 2012.

    Such bitterness.

  77. VOR says:

    Stephen,

    I just sent you an email. Might need to check your junk mail.

  78. leadfarmer says:

    Bank Shot,

    Well people are questioning the Gryba buyout and Lowe was more than 20 percentage points higher in the goals percentage department. Not bringing in a top 4 RHD is going to cost us a lot of wins no matter who we bring in

  79. Scungilli Slushy says:

    OriginalPouzar: Yup, even in our playoff year when we were relatively healthy I think we used 12.

    This is another reason why I’m stupified by the Gryba buyout.

    We are set up for each of Keegan Lowe, Ryan Stanton, Kevin Gravel and Jones and/or Lagesson to get NHL minutes (in addition to Bear who could likely use another year prior to being forced in to action).

    OP maybe he wanted out. I think he’s done at the NHL level. PC might have taken one on the chin to give him a last shot, which would go miles in player circles and the Oilers need that after the gong show of the past. He also should have learned how to skate.

    If Gryba is the answer the question is wrong as far as this season goes.

  80. Scungilli Slushy says:

    VOR:
    Stephen,

    I just sent you an email. Might need to check your junk mail.

    I’m sure it’s very good!

  81. Scungilli Slushy says:

    JimmyV1965: I’m not sure where this is coming from. Aberg played 18 min his first game back from suspension, on the second line with Drai. He had a great game and continued to play well and kept playing 18 min. It didn’t last long though and his play faded and his time was reduced to about 13 min. His play continued to fade and he was finally benched near the end of the season.

    Aberg’s role on the team will be determined by Aberg. He has to show up game in and game out.He’s just another prospect who has to do something to make himself stand out from the rest of the prospects.If not, he’s out the door.

    YUUUP!!!

  82. OriginalPouzar says:

    leadfarmer: But isn’t minimum salary like 650k or something like that.Keegan Lowe has 675k cap hit and we don’t know what Jerabek has but If like you said is a million if Keegan Lowe replaces him you still lose 325k.

    Well, yes, but Keegan Lowe isn’t going to be on the opening night roster, at least not unless there are at least a couple injuries.

    Jarabek is likely to replace Gravel or Bouchard.

  83. OriginalPouzar says:

    hunter1909,

    No worries.

    To be clear, I’ve questioned if he is ready (never said he is or isn’t, I simply don’t know, he’s yet to skate a rookie camp let along main camp let alone an X game let alone a late camp X game) but I have no doubt he’s in the conversation.

  84. Kinger_Oil.redux says:

    Pescador: Quality>Quantity
    We have 3 top 4 D
    And 7 sevens, again
    Sans balance

    – I agree with this

    – It’s a good pick-up, regardless of the status of Sek: he is competition with Russell, Benning, Gravel, Bouchard on the lower end.

    – It will be interesting to see what happens though: I’m sure they owuld like to get another bona-fide top-4, but the cost is dear no doubt

    – Maybe Benning is read for top-4,we know what Russel is (maybe top-4 with a better D).

    Is it:

    Klef-Larsson
    Nurse-Russel
    Gravel/Jerabek-Benning

    – It’s only Russell whose out of place, but a wily vet who does it: and maybe Benning moves up?

    – Pretty dangerous to go into season with this though IMO

  85. Woodguy v2.0 says:

    godot10: A physical Jay Bouwmeester.

    Bingo.

    I think Nurse is working hard to be more.

  86. Woodguy v2.0 says:

    leadfarmer:
    Woodguy v2.0,

    Although you weren’t very high on Seth Jones either

    No, that’s another misunderstanding.

    At the time there was RNH for Jones rumours I said;

    Jones has killed third pair, but that doesn’t mean he’s ready for 1st pair.

    RNH is the only center on the Oilers who could play vs toughs and losing him was not a good idea.

    That’s a far cry from “you weren’t high on Jones” but people remember their opinions much easier than written words.

  87. fireantz says:

    Ribs,

    On a team that wants to “move the puck faster” we are already stuck with Lucic on the left and now we add Upshall. How do we get any of Khaira Caggiula or Aberg into the lineup. If we give up on Khaira I believe it would compare to not signing Brodziak in 2009. Kassian and a pick for Faulk would be a smarter option

  88. leadfarmer says:

    Woodguy v2.0,

    Pretty sure the discussion was you didn’t know his upside and since he wasn’t elite early he wasn’t going to be because elite players are elite early and we should be pursuing Ryan Ellis. Or something close. Maybe I don’t remember it correctly. Happens once in a while. But not very often

  89. fireantz says:

    JimmyV1965: I’m not sure where this is coming from. Aberg played 18 min his first game back from suspension, on the second line with Drai. He had a great game and continued to play well and kept playing 18 min. It didn’t last long though and his play faded and his time was reduced to about 13 min. His play continued to fade and he was finally benched near the end of the season.

    Aberg’s role on the team will be determined by Aberg. He has to show up game in and game out.He’s just another prospect who has to do something to make himself stand out from the rest of the prospects.If not, he’s out the door.

    You can say the same for Caggiula but the coaching staff never took him out of the lineup

  90. OriginalPouzar says:

    Upshall is signed to a tryout, not an NHL contract.

    Lots of talk like he’s on the team. I understand some of the insiders like Staufer has posited that its all but a done deal but, if that was really the case, he’d be signed to an actual contract.

    We shall wait and see.

  91. Dominoiler says:

    I find it confusing, perplexing and borderline hilarious how Rattie can be a lock for the roster, yet if he had compete for 2rw, 3rw or 4rw he would be on the outside looking in.. but since the oilers have a monster #1rw hole that no one can bare to look at, everyone seems to summon up a gigantic blind spot that ignores that absolutely massive hole on the roster..

    Confusing, perplexing and borderline hilarious..

    Having Rattie #1 on the RW depth chart since June has been, and remains to be, an odd choice, imo.. sure, they have no better options and don’t have the assets to address this, but at least face it with clarity..

  92. Dominoiler says:

    (I’ll be happy to eat crow if he scores 20)

    Over under on 20 games this season?!..

  93. Georgexs says:

    Yesterday, I posted a look at the offensive results for forwards picked in the top, middle, and bottom thirds of the first and second rounds of the draft. I’m putting the numbers up again and adding one more field, the proportion of players selected in each group who end up playing in the NHL.

    Segment, Ave. PPG, Median PPG, NHL Player Count, % of Players Selected Who Made it to the NHL

    1-10, .67, .60, 118, 1.00
    11-20, .49, .39, 91, .87
    21-30, .49, .41, 98, .90
    31-40, .41, .30, 61, .67
    41-50, .43, .30, 62, .68
    51-60, .44, .34, 58, .64

    You can see that the stats for top 10 picks look different from the rest. Picks 11-20 look like picks 21-30. And second round picks look very similar regardless of actual draft position.

    This suggests to me that NHL teams are mostly able to figure out who belongs at the top of a draft. After that, what matters is that you’re some team’s first round pick or you’re some team’s second round pick. That matters more than where you were picked in the first round (if you were picked outside of the top 10) or where you were picked in the second round. Where a forward is picked doesn’t predict how he’ll do once we know whether he was picked in the top 10, 11-30, or 2nd round. What matters is some team picked him as their (likely) first or second choice.

    So there’s less information in the order in which teams select forwards than I previously thought. If teams follow pre-draft rankings and pre-draft rankings are sorting by talent, then these results suggest that outside of the top 10, and for forwards in similar tiers, teams are able to develop forwards to similar levels despite differences in perceived talent.

  94. Pescador says:

    fireantz:
    Ribs,

    On a team that wants to “move the puck faster” we are already stuck with Lucic on the left and now we add Upshall. How do we get any of KhairaCaggiula orAberg into the lineup. If we give up on Khaira I believe it would compare to not signing Brodziak in 2009. Kassian and a pick for Faulk would be a smarter option

    If your concerned about Khaira getting reps, I’d suggest your worried about the wrong guy.
    Upshall will be competing for 4th line minutes with Kassian, Caggiula and Aberg.
    Khaira is your 3Lw if Lucic is 2Lw

  95. godot10 says:

    OriginalPouzar:
    Upshall is signed to a tryout, not an NHL contract.

    Lots of talk like he’s on the team. I understand some of the insiders like Staufer has posited that its all but a done deal but, if that was really the case, he’d be signed to an actual contract.

    We shall wait and see.

    His competition is Caggiula and Rattie and Aberg. The bar for Upshall to jump over is pretty low.

  96. selective sam poll says:

    VOR: I did say it will take the analytics community a long time to acknowledge the data smog problem. But let’s rift on TOI data for a moment. Is 100 points acquired in 30 minutes a night in 60 games more or less predictive of a future NHL star than 100 points in 25 minutes a night in 60 games? Why? What compounding factors might there be? How do we allow for those? For that matter are points per minute played predictive of points per minute in the NHL. Sounds like you’ve made the assumption that it is. But where is the proof?

    I can go on and on if you’d like. Or you could just acknowledge time on ice data will just give us another thing to argue about but in a game like drafting provide no advantage because it will be less than definitive and everybody will have the same data. It will just add complexity.

    But I am quite content to leave the argument and return in a decade and say, see I told you so. It took five years for people to admit I was right that Corsi is a poor way of evaluating and rating Dmen. And there are still hold outs. I am looking at you, LT.

    As for people making money predicting the outcomes of hockey games while I don’t gamble I can consistently outperform chance at picking the winners of NHL games. All I need to know is venue, referees, and streaks. Try it for yourself.

    I am not sure why you’d think you need fancy stats or that gambling success was prove of value in hockey analytics. Every successful system gambler and horse racing and baseball produce quite a number is an inveterate lire and you couldn’t pry their secrets from them by tearing their fingernails off. They also all know their success is a result of being smarter than the mugs who make up the vast majority of the gambling population. They aren’t about to let the mugs into the game.

    But that does bring up one of my points about methodology in hockey analytics. The lack of publication of the majority of the work in hockey analytics in peer reviewed academic journals. The reason I use referees, venue, and streak information to predict hockey games is because there are peer reviewed papers, multiple papers actually, that explain the relationship and give reliable insight into the methodology. I can be reasonably confident in the workmanship.

    Why won’t the hockey analytics community submit their work to be scrutinized by their peers in a proper review format? Or do they submit and get rejected over and over again. Every other major professional sport’s analytical community has a voluminous published research literature. Why not hockey?

    Where are the Stanley Cup winning teams that are throwing parties for their analytics departments. That isn’t to say they don’t have analytics departments just that they are part of an integrated whole. Many bad teams have large analytics departments. But the analytics community seems determined to blame all their failures on poor management and champion their modest successes.

    The game of course will to be to tear me down. Which merely proves my fundamental contention which is that the hockey analytics community is antagonistic to even moderate criticism.

    Just wait until I get around to criticizing their basic methodology.

    This is the post of the century !
    For some reason, the people who ran a few macros and pivot tables off their Excel all of a sudden believe they rediscovered the wheel again. What could be an argument about data sample and methodology turns into ” I have Excel and you don’t” So my argument is stronger and now Eberle is worth $6M.
    QED

    That’s fine as an opinion but somehow an opinion was deemed fact on many occasions here and I find it very difficult to read many of these posts.

    The thin skinned responses were evident and plentiful in the past. Unfortunate the criticism was coming from many of the so called unwashed ON crowd. Otherwise the criticism may stand a better chance of being argued appropriately.

    To avoid confusion, my post is not directed at Wilde.

    Video analysis is one form a analysis. But I get the feeling a few prominent posters believe Excel is greater than Youtube as an example. It’s not which is greater but so much more about how much more we don’t know.

    While I am on fire here making friends with my above post, perhaps we could also discuss whether a person who has played competitive hockey has additional insight into the game vs a non player. For example, intimidation factor is a huge factor only in hockey but you wouldn’t truly know unless you’ve been there.

  97. geowal says:

    OriginalPouzar:
    Upshall is signed to a tryout, not an NHL contract.

    Lots of talk like he’s on the team. I understand some of the insiders like Staufer has posited that its all but a done deal but, if that was really the case, he’d be signed to an actual contract.

    We shall wait and see.

    PTO also allows them to defer the cap until Sekera is LTIR. And not take any risk should Upshall get hurt himself at some point, or just be surprisingly bad. But I think it’s an@0% done deal.

  98. Woodguy v2.0 says:

    leadfarmer:
    Woodguy v2.0,

    Pretty sure the discussion was you didn’t know his upside and since he wasn’t elite early he wasn’t going to be because elite players are elite early and we should be pursuing Ryan Ellis.Or something close.Maybe I don’t remember it correctly.Happens once in a while.But not very often

    Pretty sure you’re conflating things.

    I said he killed 3rds but that doesn’t mean he can play firsts.

    Don’t put words in my mouth.

    Get a link to where I said what you think I said.

  99. jtblack says:

    Pete knew Reinhart would kill 2nd pair ….. in the AHL ..

  100. jtblack says:

    If Pete sits toght and Russell is playing 2nd pair, I think we have too much wobble.

    Many have mentioned here, that other injuries will happen to our D core.

    So if Larsson, Klef or Nurse is out for say 2 months, how does our D hold up? Answer: it doesn’t.

  101. v4ance says:

    VOR, I can only speak for myself so I’ll talk about why I believe in analytics.

    I’ve always been a numbers guy, went to engineering, done some number crunching things. When it came to discussing one of my favorite subjects of hockey, I preferred to try and find the best impartial ways of ranking players so I could be better than my competitors at winning hockey pools. I was always surrounded by fellow fans who talk about this coach’s assessment of a player or that scout’s rating of a player or this TV analyst’s report but that always came down to the “eye” test which was highly subjective to that coach or that scout or that analyst. They could all agree or they could disagree but you wouldn’t know who was actually “right”. even then, they could be correct in one instance and totally wrong on another player

    My belief in analytics came about because I saw how analysing bulk data revolutionised and upended various fields and industries with the increasing availability of computing power. Analytics can get things wrong because of the garbage in garbage out principle but as time goes on, we get better at building better models, using the right types of data and making better software tools to evaluate things impartially and correctly.

    Analytics at this point isn’t a great predictive tool but it has shown some value in spotting “who’s doing it wrong”. One great example is Vancouver Canucks scouting/management VS a potato that CanucksNation did a few years back. By any objective measure, a potato that simply picked the highest scoring draft eligible player remaining, outperformed the Vancouver staff.

    If people in the analytics field appear thin skinned, a lot of it’s due to main stream media types like Marc Spector or Eric Francis who get on national TV or Twitter and openly deride analytics as having absolutely no value at all. But then there’s all the other people who are not TV analysts but who all have their own biased opinions which they tend to share loudly about how they disagree with the numbers guys. Quite venomously, I might add, if you look at some of the twitter responses.

    In the end, I believe the eye test has some value. I believe analytics have some value. When both agree, it’s easy. When one disagrees with the other, that’s when we get interesting questions and more nuanced evaluations. To me it’s all data. The human perspective and the numbers. Both can “see” some things clearly but clearly miss other aspects. I try to keep an open mind but the results from analytics tend to make more sense to me and tends to get things right more often.

  102. jtblack says:

    v4ance,

    Analytics is “New” to the NHL. But it’s not as new to Basebell. More so it’s not new to business and the stock market. The stock market is full of “eye test”, “gut feel” people; who in General get slaughtered at the market.

    The long term brokers who deliver above average results year over year run a system (Analytics) that takes out the emotion and the eye test. There are always exceptions, buts that what they are.

    It doesn’t take Scotty Bowman to run McD out there 22 mins a nigth. Eye test AND analytics ppl are correct in Connor. It does take more than an eye test to figure out what D man is getting caved, what PK options are NOT working, etc etc.

    The eye test will always matter; we all watch the game – Coaches, fans, announcers et al. And as with all walks of life, some people will be better than others and some will be excellent (most coaches are probably excellent to get where they got).

    Bit it’s 2018. Most NHL coaches have learned their craft in the pre analytics days. And that makes the “leap” to a more Analytical approach more difficult.

    Fast forward 10 years and I beleive almost all coaches / GM’s will rely and value more heavily on Analytics.

  103. jtblack says:

    Random question for those that are awake.

    When do you think the NHL will adopt tracking chips in the players jersey’s?

    I think that will also add a whole level to Analytics

  104. Pescador says:

    jtblack:
    Pete knew Reinhart would kill 2nd pair ….. in the AHL ..

    Except he never watched him play,
    What a fuck job, still to this day

  105. jtblack says:

    Pescador: Except he never watched him play,
    What a fuck job, still to this day

    Bahahaha … there are bad trades and then there isnthe Reinhart trade.

    Lets just hope Pete keeps his 2019 1st rounder. deep draft like 2015

  106. JimmyV1965 says:

    fireantz: You can say the same for Caggiula but the coaching staff never took him out of the lineup

    Meh. Coaches play favourites. What’s new. Life isn’t fair. If Aberg wants to be an NHL player he has to be that much better than the Drake, or any other prospect. The bar isn’t exactly high. Find out what impresses the coach and go out and do it. Aberg has some very good tools. If he fails to make an NHL career, that’s on him. With his speed and shot he should be able to elevate himself above the other prospects.

  107. Glovjuice says:

    JimmyV1965: Meh. Coaches play favourites. What’s new.Life isn’t fair. If Aberg wants to be an NHL player he has to be that much better than the Drake, or any other prospect. The bar isn’t exactly high.Find out what impresses the coach and go out and do it. Aberg has somevery good tools. If he fails to make an NHL career, that’s on him. With his speed and shot he should be able to elevate himself above the other prospects.

    BAM

  108. Georgexs says:

    WG,

    I had a look at some of the draft posts you linked to yesterday.

    One common feature seems to be a logarithmic curve that’s been fitted to some measure of player value or contribution along the y-axis and draft position along the x-axis. The fitted curve is going to be negatively sloped and it has to run from a high player value for the 1st draft pick to a low value for the 210th draft pick. Its job is to fit the entire range of data with minimal error.

    What I’m showing is that when you look at specific ranges of data (11-30, 31-60), you don’t find the negative slope. The picks in the range seem to deliver the same value, not a gradually decreasing value as suggested by the fitted curves. The ranges make intuitive sense when you take a step back.

    Teams typically have one first round pick and one second round pick. If a team is lucky/bad enough to have a high first round pick and chooses a forward, it will select one of the clear offensive talents of that cohort. If it’s picking a forward outside of the top 10, the team enters the gradually decreasing portion of the logarithmic curve. But the only reason it’s gradually decreasing is because value decreases BETWEEN rounds (after the top 10) rather than WITHIN rounds.

    Here’s how to think about it. Once we’ve selected the clear stars, each first round forward draft pick is SOME TEAM’S first round forward draft pick. So they get ALL of the attention, care, and opportunity that teams are able to give first round picks. At that point, the results are no longer predictable by draft position; they become pretty uniform. Outside of the top 10, teams develop their first round picks to the same level.

    This same logic applies to second round picks in that each player is some team’s second choice and gets the attention and opportunity that’s due. The 57th pick is not competing with the 32nd pick in his cohort; he’s not in the 32nd pick’s shadow. Instead, he’s his team’s second round pick for that year and he gets as much support and opportunity in his team as the 32nd pick gets on his. This is why, again, draft position doesn’t predict results in the second round. The results are uniform. Teams develop second round picks to the same level, regardless of draft position.

    The pattern isn’t as strong for later rounds. I think this is because there are, on average, between 2-3 forwards that make their debut per team every season. Teams seem less committed to developing picks in the later rounds because the higher picks get most of the attention and opportunity and there are only so many jobs. (The 3rd round looks different from both the 2nd round and from later rounds. It seems like an intermediate range. I’ve chosen to throw it in with the later rounds to limit the categories and keep things simple.)

    From my reading of your links, I think my approach here is a novel way of looking at the draft. Fitting logarithmic curves, as is commonly done, leads to the (mistaken) belief that value is constantly dropping WITHIN rounds when the drops are really happening BETWEEN rounds. The logarithmic approach misses the intuitively satisfying explanation that these forwards are SOME TEAM’S first and second choices; and, once they’re picked, they get equivalent development within their organization. Which leads, in aggregate, to equivalent results.

    That’s my theory.

  109. rickithebear says:

    It is scary how much people value dmen like Nurse and Klefbom avoiding quick transition to skate up the ice to face NZ traps.

    Let’s pick the hardest path to Zone entry.

    Crazy!

  110. GMB3 says:

    I really like the Jerabek signing, similar to how I felt about Auvitu. I hope he gets a good opportunity to earn his keep.

    I truly believe that if you keep making enough of this low risk bets, one of them will pan out in a big way. Mind you I also thought Anton Belov was a huge pick up at one point in time.

  111. rickithebear says:

    leadfarmer:
    Bank Shot,

    Well people are questioning the Gryba buyout and Lowe was more than 20 percentage points higher in the goals percentage department.Not bringing in a top 4 RHD is going to cost us a lot of wins no matter who we bring in

    In structure thier is only one that has 2 Dmen.
    The other has a rover.

    When I presented the concept of a 3 axis player situation matrix system based on Desjardins comp, Team, zone start Right after Behind the net appeared
    I stated there was a group of situation avg to measure players against.
    Everyone hammered the small sample size.

    For me it set a standard for what a skater was.
    Goal diff performance versus the mean partnered with comp based WOWY told me right away weather a player could handle the situation.

    Then when I married it with shot density, there was no question about performance.

    You cannot handle 1st comp you are not a PvP player.
    Therefore not a 1st line player.

    You are defined comp level you are a positive performance relative to teammates and zone start situation avg.

    I continued to beat this drum.

    All 3 axis are required.
    Or else you do not properly define the avg to measure against.

  112. rickithebear says:

    GMB3:
    I really like the Jerabek signing, similar to how I felt about Auvitu. I hope he gets a good opportunity to earn his keep.

    I truly believe that if you keep making enough of this low risk bets, one of them will pan out in a big way. Mind you I also thought Anton Belov was a huge pick up at one point in time.

    Auvitu was a HD dman who could established a strong GA base.
    Providing more room for lesser performance at the top end of HD comp end.

    As Fistric, Gryba,Davidson did.

  113. VOR says:

    v4ance:
    VOR, I can only speak for myself so I’ll talk about why I believe in analytics.

    I’ve always been a numbers guy, went to engineering, done some number crunching things.When it came to discussing one of my favorite subjects of hockey, I preferred to try and find the best impartial ways of ranking players so I could be better than my competitors at winning hockey pools.I was always surrounded by fellow fans who talk about this coach’s assessment of a player or that scout’s rating of a player or this TV analyst’s report but that always came down to the “eye” test which was highly subjective to that coach or that scout or that analyst.They could all agree or they could disagree but you wouldn’t know who was actually “right”.even then, they could be correct in one instance and totally wrong on another player

    My belief in analytics came about because I saw how analysing bulk data revolutionised and upended various fields and industries with the increasing availability of computing power.Analytics can get things wrong because of the garbage in garbage out principle but as time goes on, we get better at building better models, using the right types of data and making better software tools to evaluate things impartially and correctly.

    Analytics at this point isn’t a great predictive tool but it has shown some value in spotting “who’s doing it wrong”.One great example is Vancouver Canucks scouting/management VS a potato that CanucksNation did a few years back.By any objective measure, a potato that simply picked the highest scoring draft eligible player remaining, outperformed the Vancouver staff.

    If people in the analytics field appear thin skinned, a lot of it’s due to main stream media types like Marc Spector or Eric Francis who get on national TV or Twitter and openly deride analytics as having absolutely no value at all.But then there’s all the other people who are not TV analysts but who all have their own biased opinions which they tend to share loudly about how they disagree with the numbers guys.Quite venomously, I might add, if you look at some of the twitter responses.

    In the end, I believe the eye test has some value.I believe analytics have some value.When both agree, it’s easy.When one disagrees with the other, that’s when we get interesting questions and more nuanced evaluations.To me it’s all data.The human perspective and the numbers.Both can “see” some things clearly but clearly miss other aspects.I try to keep an open mind but the results from analytics tend to make more sense to me and tends to get things right more often.

    I appreciate your well thought out response.

    I want to clarify.

    I am not anti analytics.

    I am opposed to doing it badly and wasting time, resources, and opportunity.

    I believe hockey analytics up to this point in time has consisted of bright people drilling a series of dry wells. But if you stop drilling you will never find oil.

    I am being such a huge jerk because I have some very important things to say and I want people’s attention when I say them. Tomorrow I am going to launch a major challenge to the methodology of the that currently dominates hockey analytics. My goal however is not to destruct but to reconstruct.

    Like you numbers speak to me. I love hockey. This is why I played for more than fifty years. And numbers should be able to help inform it and maybe ever change the game for the better.

    Where I differ with the hockey analytics community is that I don’t believe for a moment we can get to the future I see and we all claim to desire if we stay on our current path.

    I believe the first step is for the hockey analytics community to confront some very hard truths.

    I can’t imagine a community less prepared to hear criticism.

    I mean we’ve sold ourselves on the idea that anyone who criticizes us is our intellectual inferior and thus has nothing of value to contribute. So we attack our critics. Watch what happens to me tomorrow when I point out the emperor has no clothes.

    If everybody was approaching the problem with your spirit we’d be much further ahead though I am sure I will test your patience mightily tomorrow. All I ask of you or any other reader is that you think about what I have to say.

  114. rickithebear says:

    VOR:
    The season GAA goes up as the HD:LD ratio reduces.
    But
    1.we cannot eliminate the Open shot aspect of game.
    2. The existence of defence of HD area.
    3. The major influence of Fwd NZ trap

    The reduction of goalie equipment to cause a major increase in open space area in net elevation is the critical change needed.

    But we will not see 80,s Numbers.
    Unless stick technology shortens the reaction time available to goalies.

    But this was obvious from my theories.

    High ratio goal diff influences (decisions)

    I have presented them for years!

    Lots on here Do not like counter thought to traditional system.
    They have always wanted the solution to fit the existing system which allows poor Goal diff decisions.

    VOR, I look forward to the high goal diff influences ( game winning decisions) that you are going to present.

    My major
    1. HD ratio
    2. Open shot %
    3. Corsi agdjustment for
    3a. FO ZS
    3b. Bench change ZS
    4. Fwd NZ trap run
    5. 2-1 Def structure vs 1-1-1 structure.
    6. Poor rover Even production efficiency
    7. Targeting open holes based on reduction of pocession turnover counterattack potential.
    8. Critical pocession change skills.

  115. Woodguy v2.0 says:

    Georgexs,

    That’s my theory.

    Interesting.

    Ties to development.

    I look forward to more from you on this

  116. Chelios is a Dinosaur says:

    VOR,

    My work is on cities and regions, especially with respect to mineral extraction.

    There are fewer more complex human systems than our built environment and how we interact with it.

    However, in the past 40+ years, increasingly we have moved away from technocratic or purely quantitative interpretations in order to explain and plan, and instead developed either mixed methods or those approaches found in the likes of ethnography or political ecology. Because of this we have a much better idea of how we interact with our environment and make plans to deal with anticipated problems. This seems obvious enough.

    It has struck me as odd or unfortunate that, while we come to terms with the degree to which hockey is perhaps more complex than other team sports, like baseball, and therefore more difficult to measure, there has instead been a doubling-down on analytics as a means of explaining what we are seeing instead of a furthering of qualitative lenses. Instead of making better use of the eye, we throw out the eye test as some vestige of old, grey-haired general managers who have been chewing the same cigar since 1978. People in hockey don’t always make the right decisions, but we hardly ever know why the decision is made, in full awareness of the all the necessary information.

    This is, quite certainly, because we aren’t in the room, therefore our population is out of reach and, it hasn’t helped that the gatekeepers of this research: the GMs, scouts, coaches and players, have 0 reason or incentive to let us into their world.

    So I am not against statistical analysis in hockey. But it needs to be mixed or its useless. That’s why teams hire analytics people, they don’t just follow analytics blogs. They need those people in their mix.
    Is Lucic depressed? Is Talbot not sleeping enough? Is Puljujarvi lonely? These are immeasurably important factors in evaluation and planning. And we have zero clue.

  117. Melvis says:

    VOR: I am being such a huge jerk because I have some very important things to say and I want people’s attention when I say them. Tomorrow I am going to launch a major challenge to the methodology of the that currently dominates hockey analytics. My goal however is not to destruct but to reconstruct.

    Or you could have just kept your powder dry.

    I’ve always read your posts with a considerable amount of interest, but this?

    Historically, the minute some prof intimated an intention to “reconstruct”, usually meaning “instruct”, I walked. Personally, a teaching moment had passed. How to think had been usurped by what to think.

    Unfortunately, I’m anticipating some backlash based on tone, rather than content.

  118. who says:

    Woodguy v2.0: No, that’s another misunderstanding.

    At the time there was RNH for Jones rumours I said;

    Jones has killed third pair, but that doesn’t mean he’s ready for 1st pair.

    RNH is the only center on the Oilers who could play vs toughs and losing him was not a good idea.

    That’s a far cry from “you weren’t high on Jones” but people remember their opinions much easier than written words.

    I don’t remember what your opinion of Seth Jones was 2 years ago, but the question is were you for or against the RNH for Jones trade that was rumored at that time.
    I am a big fan of the Nuge but I stated at the time that I would make that trade.
    Were you against that trade at the time because Jones hadn’t proven himself as more than a 3rd pairing dman? Or because Drai wasn’t a proven center yet and Nuge was? Or both of the above?
    I think this is a common failing when using analytics to make roster decisions. Most models don’t take potential upside into account. If you are basing your decisions only on what the player has done you end up acquiring a lot of players that are past their prime. Wasn’t Lucic an analytics star before the Oilers signed him? Extreme example I know.
    My point is, that because Jones and Drai hadn’t reached their ceiling yet, and Nuge had, some people considered this a bad potential deal for the Oilers.

  119. russ99 says:

    OilClog: Sorry but can’t blame the wingers for the cement feet D behind them.

    They signed a guy that was being healthy scratched while the Montreal D was looking uglier then Edmonton’s.

    Maybe Russ can do some circle backs and 47 resets before he decides to pass the puck off the boards.

    Watch the video on Athletic, it’s so damning. Our D are in the right place, our wingers either turn up ice or float in no-mans land.

    All our defensemen did the reset, it’s systemic. Besides, The breakout is a 5-man responsibility, not rush up the ice as fast as you can and maybe there will be a window to receive a good pass.

    McLellan used the term “cherry-pickers” a few times in postgame pressers.

  120. VOR says:

    Chelios is a Dinosaur:
    VOR,

    My work is on cities and regions, especially with respect to mineral extraction.

    There are fewer more complex human systems than our built environment and how we interact with it.

    However, in the past 40+ years, increasingly we have moved away from technocratic or purely quantitative interpretations in order to explain and plan, and instead developed either mixed methods or those approaches found in the likes of ethnography or political ecology. Because of this we have a much better idea of how we interact with our environment and make plans to deal with anticipated problems. This seems obvious enough.

    It has struck me as odd or unfortunate that, while we come to terms with the degree to which hockey is perhaps more complex than other team sports, like baseball, and therefore more difficult to measure, there has instead been a doubling-down on analytics as a means of explaining what we are seeing instead of a furthering of qualitative lenses. Instead of making better use of the eye, we throw out the eye test as some vestige of old, grey-haired general managers who have been chewing the same cigar since 1978. People in hockey don’t always make the right decisions, but we hardly ever know why the decision is made, in full awareness of the all the necessary information.

    This is, quite certainly, because we aren’t in the room, therefore our population is out of reach and, it hasn’t helped that the gatekeepers of this research: the GMs, scouts, coaches and players, have 0 reason or incentive to let us into their world.

    So I am not against statistical analysis in hockey. But it needs to be mixed or its useless. That’s why teams hire analytics people, they don’t just follow analytics blogs. They need those people in their mix.
    Is Lucic depressed? Is Talbot not sleeping enough? Is Puljujarvi lonely? These are immeasurably important factors in evaluation and planning. And we have zero clue.

    While I have wandered academically you can’t shake off the academic subject you first imprint on. I trained first as an ecologist. The observation that clinched my interest in the field. I was in the high desert of northern Nevada. It is about as hostile a place as I would care to be in. Nothing lives there.

    Well not until you stop, look, explore. Then you find a complex ecological system shaped by ants. As far as you can see there are pimples Those are ant hills. They have a unique mathematical relationship to each other.

    That relationship can be explained by math. Nope it is down to the social behaviour and chemistry of this specific species of ant. The hills cooperate in a myriad of complex tasks including picking the next hill site.

    It just seems to me the tools used to study these sorts of problems could usefully be deployed in hockey analytics. But you said it so much better than I could.

  121. VOR says:

    Melvis: Or you could have just kept your powder dry.

    I’ve always read your posts with a considerable amount of interest, but this?

    Historically, the minute some prof intimated an intention to “reconstruct”, usually meaning “instruct”, I walked. Personally, a teaching moment had passed. How to think had been usurped by what to think.

    Unfortunately, I’m anticipating some backlash based on tone, rather than content.

    I am hoping the tone will come off as warm and supportive, maybe even funny. I am trying for the way I talk when I am coaching. That will stand in stark contrast to what I have done up to date today and be more like the stories I tell here from time to time.

    This version won’t be mathy or have references as I plan to do a long series of posts with that information on my own blog.

    I am going to post it here in this thread so as not to derail whatever LT is writing about today.

  122. Melvis says:

    VOR,

    Great. Thank you. I’m looking forward to it.

  123. Woodguy v2.0 says:

    who: I don’t remember what your opinion of Seth Jones was 2 years ago, but the question is were you for or against the RNH for Jones trade that was rumored at that time.
    I am a big fan of the Nuge but I stated at the time that I would make that trade.
    Were you against that trade at the time because Jones hadn’t proven himself as more than a 3rd pairing dman? Or because Drai wasn’t a proven center yet and Nuge was? Or both of the above?
    I think this is a common failing when using analytics to make roster decisions. Most models don’t take potential upside into account. If you are basing your decisions only on what the player has done you end up acquiring a lot of players that are past their prime. Wasn’t Lucic an analytics star before the Oilers signed him? Extreme example I know.
    My point is, that because Jones and Drai hadn’t reached their ceiling yet, and Nuge had, some people considered this a bad potential deal for the Oilers.

    None of my take had a thing to do with “analytics” unless you include “thinking through a decision” as “analytics”

    EDM didn’t have a C to take on the tough assignments except RNH. Drat’s ability was a black box in 2015. Even today 3 years later he hasn’t been above 50% GF when centering his own line

    Jones killing 3rd pair didn’t mean you could thrust him into 1st pair on EDM.

    Trading the only tough minute C that you have for a Dman who projects well, but will take a couple years doesn’t make sense in all situations.

    When he went to CBJ he played mostly 3rd with Murray for the rest of the 15/16 season.

    In 16/17 he played 2nd pair with rookie Werenski while Savard/Johnson had tougher minutes. They killed their minutes.

    CBJ plays more of Top 4 like EDM does than 1-2-3 in terms of Dpairs.

    Jones/Werenski were the more sheltered of the top 4 in 16/17.

    In 17/18 he and Werenski were 1st pair in the top 4 last year and did well.

    My take was:

    -If you trade RNH you’ll just need him again as who knows when/if Drai can take his job
    -Jones is killing 3rd pair, but that doesn’t mean he’ll come to EDM and slot into the top pair. That’s usually a bad way to do business
    -If Jones does progress to top pair, how many years will it take?

    Keeping RNH was more conservative and in retrospect Hall would still be an Oiler.

    Hall and Drai could have been a killer 2nd line as well.

    If that deal was available and I knew then what I know now, I would do that deal.

  124. VOR says:

    I am going to do this in chunks.

    I said I was going to ask hard questions and here is the first. How much would you like access to SportsLogiq’s data?

    Not where you thought I would start a critique of hockey analytics?

    Well let’s try a second one. What would it take for the hockey analytics community to get that access?

    I am sure many cynical souls would say it is impossible. Put on your rose coloured glasses for a moment and come along for the ride.

    I would hypothesize that there would have to be something pretty sweet in it for SL. Can we imagine something sweet enough? Why not offer that all IP the hockey analytics community generates using SL’s data remains the exclusive property of SL. Well that seems a bit restrictive and not transparent at all.

    How about SL gets right of first refusal but if they say no then the idea, though not the data, becomes public domain?

    So hard question three, why not, why wouldn’t this work?

    As an entrepreneur I love the idea of many bright little minions crunching my data for free. A lot. It lets me focus on what makes money, selling useful tools to hockey teams. It also keeps my in house analysts on their toes and highly motivated to keep delivering value. Sadly I don’t have the cool data SL does.

    On the other hand who the hell am I negotiating with?

    There is no association of hockey analysts.

    And where would I find them? I guess I could go to MIT/Sloan. But frankly I stopped going because it was becoming repetitive and really turning into a meet/meat market. That is not a criticism of one of the best conferences on Earth and one that is entirely student run. Just describing the reality.

    I am looking for the free spirits, the one of a kind minds, the people who don’t just think outside of the box but have put the box down somewhere and forgotten where. I am in the business of creating new IP, old math minds need not apply. So where do I find weirdos who love numbers and hockey?

    I know the Oilers tried this once. Well they tried several approaches. I don’t know what went wrong and maybe I’m walking into a minefield but if I let that stop me I wouldn’t be much of an entrepreneur.

    So once I find a place that hockey, numbers, and incredibly creative people congregate who do I talk to, what happens when I say, “take me to your leader?”

    And my point is the ad hoc nature of the hockey analytics community needs to be rethought. That is my first criticism.

    Come back later for more, many wonderful surprises await.

  125. frjohnk says:

    VOR:
    I am going to do this in chunks.

    I said I was going to ask hard questions and here is the first. How much would you like access to SportsLogiq’s data?

    Not where you thought I would start a critique of hockey analytics?

    Well let’s try a second one. What would it take for the hockey analytics community to get that access?

    I am sure many cynical souls would say it is impossible. Put on your rose coloured glasses for a moment and come along for the ride.

    I would hypothesize that there would have to be something pretty sweet in it for SL. Can we imagine something sweet enough? Why not offer that all IP the hockey analytics community generates using SL’s data remains the exclusive property of SL. Well that seems a bit restrictive and not transparent at all.

    How about SL gets right of first refusal but if they say no then the idea, though not the data, becomes public domain?

    So hard question three, why not, why wouldn’t this work?

    As an entrepreneur I love the idea of many bright little minions crunching my data for free. A lot. It lets me focus on what makes money, selling useful tools to hockey teams. It also keeps my in house analysts on their toes and highly motivated to keep delivering value. Sadly I don’t have the cool data SL does.

    On the other hand who the hell am I negotiating with?

    There is no association of hockey analysts.

    And where would I find them? I guess I could go to MIT/Sloan. But frankly I stopped going because it was becoming repetitive and really turning into a meet/meat market. That is not a criticism of one of the best conferences on Earth and one that is entirely student run. Just describing the reality.

    I am looking for the free spirits, the one of a kind minds, the people who don’t just think outside of the box but have put the box down somewhere and forgotten where. I am in the business of creating new IP, old math minds need not apply. So where do I find weirdos who love numbers and hockey?

    I know the Oilers tried this once. Well they tried several approaches. I don’t know what went wrong and maybe I’m walking into a minefield but if I let that stop me I wouldn’t be much of an entrepreneur.

    So once I find a place that hockey, numbers, and incredibly creative people congregate who do I talk to, what happens when I say, “take me to your leader?”

    And my point is the ad hoc nature of the hockey analytics community needs to be rethought. That is my first criticism.

    Come back later for more, many wonderful surprises await.

    I have seen some of Sportloqiqs data.

    Have also seen some data that the skills analysts use for NHL players. Some of the skills analyst charge into 5 figures. The players data I saw the skills analyst charged I think about 10 grand. The most I heard is 50 grand. Big money in analytics.

    The data that I have seen is Miles above anything I have seen in the public domain. And I just had a snippet. But would also say that there is so much that is undiscovered.

    I thoroughly enjoy your posts. You really need a blog gonna keep bugging you till it happens 😀

  126. BornInAGretzkyJersey says:

    Agree with Padre, VOR, if you could link us to your blog, that’d be great.

  127. VOR says:

    frjohnk: I have seen some of Sportloqiqs data.

    Have also seen some data that the skills analysts use for NHL players. Some of the skills analyst charge into 5 figures. The players data I saw the skills analyst charged I think about 10 grand. The most I heard is 50 grand. Big money in analytics.

    The data that I have seen is Miles above anything I have seen in the public domain. And I just had a snippet. But would also say that there is so much that is undiscovered.

    I thoroughly enjoy your posts. You really need a bloggonna keep bugging you till it happens

    I am literally days away on the blog. I have come up with a unique model for the blog. And I do mean unique. Let’s say I am totally rethinking the concept of hockey blog.

  128. VOR says:

    The Strange Case of the Corgis in the Night

    “Sherlock, how many times do I have to tell you it is Corsi, not Corgi,” Watson said through clenched teeth. He was grinding them at the same time. That is very difficult to do. As his employer this concerned me greatly. I am the one paying for his dental benefits. He really has to learn to relax. I wonder if I paid for yoga classes if he’d go?

    “Corsi, Corgi, they all have a snout, for paws, and a tail.” I said.

    “Sherlock, don’t make sport. Our client is in real trouble. The masses are massing. They have pitchforks and there are rumours of tar and feathers.”

    “You do realize it is possible our clients are a shade paranoid, that they see enemies everywhere including under their beds. After all what sort of name is AHA? Sounds like they are into the sort of recreational pursuit you object to me enjoying?” I responded.

    “It is an antonym,” Watson offered up gamely.

    “I think you mean an acronym. I know it stands for the Association of Hockey Analysts. And I still think they are a prize bunch of paranoids. They’ve imagined some super villain with a huge organization and given them a ridiculous name. I mean the “seen him good” gang. Really?”

    “Moriarity,” Watson said, under his breath. I pretended I hadn’t heard him.

    “In any case I just came back from the Convention of People who Follow Paranoids so maybe I am being a little harsh.” Watson didn’t laugh. He never gets my jokes. “But the answer to why Corsi was predictive of winning hockey games in the past and isn’t anymore is simple, you might even say elementary, my Dear Watson.”

    “Oh!” Watson said, “you’ve cracked the case.” Why does he always doubt me?

    “All it took was a few teams to try to deliberately increase their Corsi. They confused the measured with what it was a proxy for, possession. So they increased their shot volume without increasing their actual possession of the puck. Having a higher Corsi no longer correlated with possession. And it was possession that was correlated with winning.’

    “So what are we going to tell the clients?” Watson asked.

    “To stop using proxy measures and to get professional help with their paranoia.” You could see Watson relax. His shoulders slumped down to even with his ears.

    The hockey analytics community needs to stop using proxy measures. Apparently hockey teams are silly enough to confuse the the proxy with the real thing.

    And it needs to stop waging war with the imaginary foes. It has reached beyond quixotic and is climbing toward deranged. I get that the sense of us against the world is part of the culture of hockey analytics but that needs to change. That sense of isolation is a useless anachronism.

  129. Ribs says:

    This has been the most fascinating thread in forever. Thanks for starting the conversation, Vor!

    Analytics in hockey is still in it’s infancy, but is leaping ahead at a very good clip. I am still a believer!

    Watching the progress of deep learning has been exciting through the past three years or so. It is becoming very evident that the more information you have, the better predictions you can make. It is actually slightly alarming how things are developing!

    Inevitably, the truth tends to present itself through the data. I agree that parsing it is the challenge, but we’re getting better and better at it as time passes. We must remember that the human personalities behind the analytics we are seeing are only temporary. Math and logic win over time. Speed bumps.

    The hockey analytics community needs to stop using proxy measures. Apparently hockey teams are silly enough to confuse the the proxy with the real thing.

    There’s a lot of truth to this, but unfortunately widespread awareness is not something that develops overnight. In reality, some things need to go through the trial process before they can be advanced upon, for a multitude of reasons. I think it’s encouraging that the early data is becoming accepted and that teams are making attempts to use it.

    Speed of advancement seems to be your biggest issue, and I think that you are correct that we won’t see the type of immediate gains that could be possible, but I think that over time the game of hockey will be enhanced greatly by the analytics that we have and will see in the (near!) future.

    I am excited to read more of your ideas!

  130. VOR says:

    Years ago I wrote a book called How to Clone Yourself. And actually it is about exactly what it says. How somebody could clone themselves with some cheap lab equipment, a step by step recipe, and a willing surrogate.

    It has a bizarre publication history. I intended it as a funny book that taught basic science while having enough social commentary to hold the attention of professional scientists and policy makers. It was also something of a warning to big science that a new era was dawning. I foresaw the rise of something I called the home based scientist. Publishers saw it as a handbook for science in a post apocalyptic world. I re-wrote it 15 times for 7 different publishers all of whom passed in the end mainly because I don’t do dystopian very well. Though nobody asked for there advance back so it is actually my most commercially successful book.

    Each publisher sent it out to several scientists for review. Somewhere in there it slipped into the public domain. Now each year I get dozens of enquiries from everybody from academics and grad students to home hobbyists asking if it is really possible to clone a human. I used to be a big deal in the cloning world. I ran the cloning version of Lowetide.

    My answer is always the same, sure but why would you? You see cloning doesn’t lead to duplication it leads to replication. Think of yourself as a science experiment. Sperm meets egg and fertilizes it. DNA recombines. Embryo develops. Unique individual is born and grows up.

    Turns out cloning replicates that experiment. If embryonic development and uterine environment and where and how you were raised didn’t matter you’d get an exact replica. But the experiment proves all that nurture matters. And it matters a lot. We know this from cloning experiments in animals.

    If we extrapolate those experiments to humans we could reasonably say cloning Albert Einstein might give you Groucho Marx. The effect is so profound it isn’t even necessarily a good way to get donor organs. This is why all the focus now is on stem cells. Which are still plastic enough to be useable in many applications.

    It used to be scientists all had a firm conceptual grasp on replication. First, it was beaten into our heads in biology class and later we spent large portions of our professional lives replicating other scientists’ work. Science had a three step process for ranking research and in a sense replication was a big part of all three.

    Every paper presented for publication was sent to outside reviewers, experts in the field, for comment. And believe me the comments could be savage. The system was also flawed in that it was very hostile to new voices and new ideas which are the lifeblood of science. But still it worked as a filter because most reviewers tried to do a good job.

    Then once a paper was published you sat around and waited. If it was interesting other scientists would mention it in their work. There was and is an entire system for keeping track of how many times your paper gets tagged called the Science Citation Index. The more citations the more scientists found your work important.

    But the gold standard was replication. That meant you’d pissed some other scientist so much they wanted to see if your experiment really worked the way you claimed. Every paper has a methodology section, a recipe book for performing the experiment. The Cold Fusion guys messed up in the methods and missed a tiny detail. So there reputations got destroyed when nobody could replicate their work. Fortunately for them somebody took the time to work with them, identify the problem, and successfully replicate the experiment.

    Successful replication would lead to your work being cited far more often. Papers on that subject then became more likely to get positive recommendations in peer review and so on. It all worked most of the time and the rules were clear, if unfair.

    Then changes started creeping in. Science and math education found itself under siege in many places. In others it was dumbed down in an attempt to increase overall interest in STEM. Number of publications became the criteria for rating scientists in academia and the era of the Least Publishable Unit was born. Most problematic funding for, and interest in doing replication of other people’s experiments waned and has nearly died out all together. Then there were online journals founded that didn’t use peer review and print journals that will print any paper if you pay them.

    There is now a new generation of experiment. One that looks at how these changes have eroded the value of research by letting erroneous and even fraudulent data out into public. The results are not pretty and changes are coming.

    My point for now is that the hockey analytics community is dominated by a dystopia world view. The one I couldn’t buy into when I was writing How to Clone Yourself. The community came of age in the era of least publishable units, no referees, and no replication.

    I am also a child of my time. I can’t read hockey analytics and not think there is a very good chance this is utter hogwash. I wait patiently for replication and citation that never comes. For me it all remains suspect until those hurdles are jumped over. Science worked my way for millennia. It has been the new way for just shy of two decades. Want to take a bet on which way works best?

  131. N64 says:

    VOR: The Cold Fusion guys messed up in the methods and missed a tiny detail. So there reputations got destroyed when nobody could replicate their work. Fortunately for them somebody took the time to work with them, identify the problem, and successfully replicate the experiment.

    DOE 2004:

    The claims of cold fusion, however, are unusual in that even the strongest proponents of cold fusion assert that the experiments, for unknown reasons, are not consistent and reproducible at the present time. (…) Internal inconsistencies and lack of predictability and reproducibility remain serious concerns.

    ~ Correct assessment or not….getting hockey analytics up to Cold Fusion levels of replication. Yikes ~

  132. VOR says:

    I am stuck, unsure what the right thing to do is and aware every option is bad.

    I think the hockey analytics community has done a lot of good for the game of hockey. It has encouraged scouts, coaches, and managers to confront their biases. More and more hockey people are using data in their professional lives.

    I think that is a great good.

    I am also aware that the use of analytics in hockey still has its opponents and detractors amongst hockey people, the MSM, and the fan base.

    While I doubt they could follow the mathematical and scientific argument I think needs to presented in order to advance hockey analytics as an enterprise I am sure they could follow the gist of it. And that eventually they would start using my argument to reinforce their resistance to change. Probably sooner than later.

    There doesn’t seem to be a way to communicate that while I see serious methodological flaws in how the community approaches analyzing hockey I think the impulse to analyze hockey is a very good one.

    So how the hell do I explain my concerns about the how of hockey analytics without sounding like I am criticizing the why?

    I decided on the following approach:

    1. If any of you want to discuss my concerns I will gladly meet in person or chat in email or by phone. My email is dennischute(at)hotmail(dot)com.

    2. I am going dark here for now. My partners and spouse have convinced me to focus on our own proprietary work.

Leave a Reply

Want to join the discussion?
Feel free to contribute!
© Copyright - Lowetide.ca