GENTLE BEN

I like Ben Scrivens, cheer like hell for him every game. He’s clearly a good human and the causes he champions are top drawer. Now, that doesn’t get him off the hook for this season but it’s still great to see him play so well for the Edmonton Oilers  (as he did last night). His SP went from .895 to .897, nothing really in a long season of disappointment. For one night though, it was splendid. Good for him. It’s important to use the lesson of Devan Dubnyk when applying blame to Ben Scrivens. Is he an .895 goalie or a .915 man? No doubt Craig MacTavish has to find a No. 1 goalie this season but I think Scrivens is a suitable choice for the other goalie in 2015-16. Thoughts?

marincin common

I thought Marincin—Fayne had a good night and Todd Nelson clearly agreed, as that pairing was sent out late to shut down the Wild. If the Oilers can be trusted, then a strong run here at the end of the season should elevate Marincin into the conversation for next season:

  • Klefbom—Schultz (most minutes, soft parade pairing)
  • Marincin—Fayne (tough opposition pairing)
  • Ference-Nikitin (lottery pairing)

One thing Nelson does that NO others Oilers coach I’ve seen this century do? Trust his former AHLers. He wasted no time putting Anton Lander into different spots and last night he showed tremendous confidence in Martin Marincin. That’s a very encouraging item.

YOU CAN’T STOP TODD NELSON

11-12-3 now with a ragtag bunch that boasts a quality top line, some emerging defensemen (well, one) and last night Dominik Scrivens. The club is winning more under Nelson (Eakins had them 7-19-5 when MacT pulled the chute) but the possession numbers favor the Eakins group.

COACH NUMBERS FEB 25

One thing I’m not going to do is argue who has the better roster. MacT clearly believes the Nelson group is superior and they did add Derek Roy, Rob Klinkhammer, Anton Lander and Martin Marincin—plus Oscar became Oscar (or more Oscar) as the season wore along. That said, David Perron is gone, Taylor Hall has been out for some time and there’s little doubt men like Ryan Hamilton and Matt Fraser and Iiro Pakarinen aren’t quite established NHLers. I’m honestly not certain who has the better 23 and it really doesn’t matter in the grand scheme of things: They’re not a good team. I will say that Dallas Eakins Oilers got killed by goaltending and I’ll go to my grave convinced of it. So, yeah. (Numbers via war-on-ice).

petry common

According to the Fourth Period, teams on Petry include Los Angeles, Tampa Bay, Chicago (all three teams were scouting last night) and Detroit, Boston, Pittsburgh and Anaheim have made inquiries. The report says MacT is asking for a second-round pick and a prospect, which seems dear based on prices this and last deadline.  We’ll see, and it’s worth mentioning no one in the credible media is suggesting there’s a snowball’s chance in hell the young man signs here. Source

WHAT DOES THE OILERS DRAFT LIST LOOK LIKE?

A lot of chatter surrounding the names and the order after McDavid and Eichel. The NHL draft industry appears to be fairly universal in their support of Noah Hanifin as No. 3 overall with Mitch Marner and Dylan Strome as candidates for top five overall. Lawson Crouse is a divisive figure in this year’s draft, some have him inside the top five and others don’t have him near that high. Red Line Report still has him outside their top 10 overall and they’re the best list (imo) so that’s certainly a tell from here.

Each Saturday on this blog I publish a Top 30 for the 2015 draft and on the Monday repeat the Top 10 overall. That’s my list, doesn’t reflect anything but a fan’s opinion (it’s a moo point!) but it’s fun and we get to learn about the prospects together. YEARS of watching the Oilers tells me their list probably has two guys in the top six overall that fans will wonder about:

  1. Connor McDavid
  2. Jack Eichel
  3. Noah Hanifin
  4. Dylan Strome
  5. Lawson Crouse
  6. Pavel Zacha

I’ll bet a 2-4 that list is pretty similar to the one Edmonton has in its inner laptop this morning. If the Oilers continue to win a little, and pass Arizona and Toronto (as an example), and end up picking No. 4 or No. 5? It could happen and I’m not even kidding.

ronstadtlinda1

LOWDOWN WITH LOWETIDE

10 this morning, we get it rolling on TSN 1260. Scheduled to appear:

  • Bruce McCurdy from the Cult of Hockey at the Edmonton Journal. Last night’s game, the Petry thing, Marincin.
  • Matt Kassian, former NHL player and now at Vimy Hockey. We’ll talk about his career and his new career.
  • Andrew Berkshire, Habs Eyes on the Prize. Smith-Pelly trade and do the Habs need Petry?
  • Jeff Krushell, Krush Performance. Spring training! And should seasons be shortened?

10-1260 text, @Lowetide_ on twitter. Talk soon!

 

written by

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

238 Responses to "GENTLE BEN"

« Older Comments
  1. G Money says:

    LMHF#1: I don’t object to the idea of variation, I object to the idea that it is “random” or “luck”, but then again there may be a difference between what you and I think when those terms are used.

    I actually regularly rail against the use of the word ‘luck’, because it creates an implication (“no skill involved”) that is misleading.

    Random variation on the other hand, is the correct term, and is inseparably tied to skill. “Skill” is what you’re capable of. Random variation is how your performances actually occur over time around that base skill level.

    Random variation is why a .880 goalie can pitch a shutout while facing a .920 hall of famer who gets pulled after giving up three goals on four shots.

    If you dig into each individual event, the specific goals or shots are not ‘luck’ or ‘random variation’. But the end result when put in context very much is, and can be meaningfully analyzed that way. A .880 goalie may pitch a shutout, but you can bet that a .920 goalie will pitch a lot more, and get pulled a lot less often.

    You just can’t predict when or why.

    That’s as close to a real world definition of ‘random variation’ as you’re going to get.

  2. rickithebear says:

    Franson
    2nd comp
    1.27 EVP/60
    3.27 EVGA/60
    High end EVP production.
    top 30 EVGA/60 facing 3rd comp firsttime in NSH

    Sekura
    1st comp
    .72 EVP/60
    2.51 EVGA/60
    1st comp EVP D.
    will have superior EVGA/60 in a 2nd comp role.

    Petry
    2nd/3rd comp
    .52 EVP/60
    3.04 EVGA/60
    none of the above!

  3. G Money says:

    SwedishPoster: But it’s hardly ever luck that decide who wins the game, it comes down to execution.

    Sorry, my friend, but this is demonstrably false. As last night’s game showed, where the better team kicked the teeth in of the shittier team, and the shittier team won.

    Scrivens won the game for the Oilers who were by far the shitter team on the night – just as he has lost the game for the Oilers when they were the better team on the night.

    Skill and execution win out, but only over the long run.

    And that in the end is why we look at shot metrics and the like, because they are actually the best indicators we have of long run skill rather than short run variation.

  4. LMHF#1 says:

    G Money: I actually regularly rail against the use of the word ‘luck’, because it creates an implication (“no skill involved”) that is misleading.

    Random variation on the other hand, is the correct term, and is inseparably tied to skill.“Skill” is what you’re capable of.Random variation is how your performances actually occur over time around that base skill level.

    Random variation is why a .880 goalie can pitch a shutout while facing a .920 hall of famer who gets pulled after giving up three goals on four shots.

    If you dig into each individual event, the specific goals or shots are not ‘luck’ or ‘random variation’.But the end result when put in context very much is, and can be meaningfully analyzed that way.A .880 goalie may pitch a shutout, but you can bet that a .920 goalie will pitch a lot more, and get pulled a lot less often.

    You just can’t predict when or why.

    That’s as close to a real world definition of ‘random variation’ as you’re going to get.

    This is why I asked and I appreciate the clarification.

  5. Snowman says:

    rickithebear,

    Who plays against the toughs for us if its not Petry? And where do you find that information?

  6. Eustace Matthews '16 says:

    G Money,

    Slightly related to the discussion:

    True randomness is incredibly difficult to create digitally.

    http://www.popularmechanics.com/technology/security/how-to/a11278/the-future-of-cryptography-is-outdated-nokia-phones-17199208/

  7. Halfwise says:

    Jon K:
    I sincerely thought and still think that Petry is worth a 2nd and a decent prospect. I think the value for the players moved so far supports that valuation. They perhaps even suggest Petry could have been worth more. A few of the major buyers are done now, however.

    If Petry gets dealt Monday at the last minute, after Sekera, Franson, and Michalek, and for less than those players, people will draw the inference that MacT waited too long and lost value as the buyer’s market fulfilled their demands elsewhere.

    This series of events has the potential be another negative on MacT’s resume…

    Let me get this right.
    If he trades Petry before other trades, “he jumped the gun.”
    If he trades Petry during the other trades, “he panicked, could have held out for more.”
    If he trades Petry after the other trades, “he waited too long.”
    If no deal gets made, “MacT is stubborn and overvalues his players”
    If MacT spends the rest of the season negotiating an extension, “he’s just trying to cover for his negligence”

    What is the scenario where it’s “MacT handled that pretty well”?

    Answering my own question, if it is Petry 5x$5M or Petry for Antti Raantta or Sobotka’s rights plus either a 2015 2nd or any 21 year old prospect with size, speed and a brain, I will say “MacT handled that pretty well”.

  8. G Money says:

    godot10: So my hypothesis, because the improved Corsi was NOT confirmed by any other indicator, that Eakins was “cheating” for Corsi. There is still some undetermined factor or factors that broke the correlation between Corsi and results.
    There is now a feedback loop that didn’t exist before. Once one is aware of advanced stats, and begins changing tactics based on awareness of advanced stats, all previous results may no longer be quite the same.

    As far as I’m concerned, if Eakins was somehow able to find a way to coach to an improved Corsi without actually improving results and without the markers of a ‘coach to Corsi’ showing up in any other shot measures, he must be the most brilliant coach ever to walk the face of the earth.

    There is only one real Corsi management metric I know of – a defensive system that pushes shots to the perimeter. This will result in a higher CA measure i.e. a LOWER not a higher Corsi, but the distance adjusted metrics will be in your favour because of the distance and location of the shots.

    In theory, you could also manage to Corsi by having your team take perimeter shots instead of trying to get higher quality looks.

    Yet, none of these things are true.

    Eakins teams did not take a disproportionate number of perimeter shots.
    Eakins teams did not give up a disproportionate number of close-in shots.

    In fact, looking at the shot location data, the one thing that clearly shows up is that the Oiler goalies gave up a (grossly) disproportionate number of goals from the blueline.

    Again, contradicting the idea that the Oiler goalies have improved because they’re facing an easier type of shot. They are not.

    So either the biggest factor for Eakins was that he was horribly impacted by bad goalering as the data strongly suggest, or he is in fact an amazing hockey mind, having figured out how do to something no-one else has ever done: improve Corsi AND decrease performance, all the while having the shot distance data contradict the idea that he’s coaching to Corsi.

    Logical conclusion?

    Godot thinks Eakins is the most brilliant coach to ever walk the face of the earth!!!!!!!!!

    Do you hear that everybody!?!?!

    Godot thinks Eakins is the most brilliant coach to ever walk the face of the earth!!!!!!!!!

  9. godot10 says:

    G Money: 1 – I’m looking at it objectively.Data and analysis are tools to help in that regard.

    2 – If Nelson’s system is better, then it should show up somewhere other than the scoresheet.It doesn’t.

    Likewise, if Eakins wasn’t cheating for Corsi, it should show up somewhere elsewhere also.

    Nelson, unlike Eakins, has more than a single up arrow. Nelson has better results, better power play, slightly better PK, players like Yak not regressing and playing to potential (Eberle, Lander, Schultz, Gazdic, etc), better starts, more time playing with the lead, etc. Nelson seems to be able to identify which are the good defensemen and which are the bad ones.

    Eakins has a marginally better Corsi, and a team and city living in hell.

  10. G Money says:

    LMHF#1: This is why I asked and I appreciate the clarification.

    You also asked why we look at these things, and this is such a deeply philosophical issue that I completely avoided (and am still avoiding) the question!

  11. blainer says:

    By the looks of the trades today the Oilers are looking at some decent returns on a trades. Petry is getting us a First and prospect. Gordon maybe a first also. Hope we sign both though. If petry is traded we are gonna have some nice assets to trade with at the draft. Hopefully we win the lottery get Mcdavid and use the other two first rounders and prospects on D goaltending and a vet center…maybe Sobotka… Would love to see Sobotka centering Hall and Yak next year.

  12. Eh Team says:

    Why are we saying Scrivens and Fasth have played better with Nelson as coach? Right now Scrivens is still at .897 save % and Fasth at .888%. Isn’t that were they were with Eakin as coach also.

    They were bad and are still bad.

  13. Halfwise says:

    G Money,

    Or there is more going on than Corsi can explain at its current resolution, and despite its attractiveness at an overall league level there is something else going on.

    I’m not diminishing the validity of Corsi, but implying that it fully models hockey is like saying GCMs fully model the atmosphere or this really expensive investment algorithm that we’re trying to sell you fully models the stock market.

    Chaos-rich systems work like their models, until they don’t. That’s nature’s way of telling us to keep trying to improve our models.

  14. G Money says:

    godot10: Likewise, if Eakins wasn’t cheating for Corsi, it should show up somewhere elsewhere also.

    Nelson, unlike Eakins, has more than a single up arrow.Nelson has better results, better power play,slightly better PK, players like Yak not regressing and playing to potential (Eberle, Lander, Schultz, Gazdic, etc), better starts, more time playing with the lead,etc. Nelson seems to be able to identify which are the good defensemen and which are the bad ones.

    Eakins has a marginally better Corsi, and a team and city living in hell.

    – Better power play? Yes. Vastly better. Sustainably better. Shows up in the shot metrics. If this is your argument for why Nelson is a better coach, I agree with you. This is the ONLY “Nelson is better” argument that is actually supported by data.

    – Better results? Easiest explanation: better PP and better goalering. And goalies are voodoo.

    – Better PK? No. It’s slightly worse. Easiest explanation is same as for the overall results: better goalering.

    – Better starts? No, the temporal charts I posted yesterday show the opposite. At 5×5, Nelson’s teams get snowed under more in the early part of the game. Easiest explanation for why they have the lead more often: better PP, better goalering. And goalies are voodoo.

    – Better player development? Yes. I think so too. It is why I think Nelson is a better coach than Eakins. Then again, could it be Roy? Yes. It could.

    Let’s be clear here. If you tell me you think Nelson is a better coach, I do not disagree. I think so too. Particularly in the PP and player development, because the first is demonstrably better and the second is subjectively better in a way that at least isn’t contradicted by the objective data.

    But if you tell me Nelson is better because “Eakins coached for Corsi”, or “look at the results”, you are spouting populist nonsense.

  15. jake70 says:

    Related to Oiler trades……David Perron update: Last 8 games

    1 goal
    1 assist
    +1
    avg TOI ~17min

  16. G Money says:

    Halfwise,

    Agreed. Corsi isn’t perfect, and no-one has claimed it is. What it is, though, is the best we’ve got right now.

    Or rather, situational shot metrics are the best we’ve got right now for looking into the game deeper than just “hey, we won, that’s all that matters! MOAR HITS!”

    And while not perfect, they do an impressively good job at identifying sustainably good/bad teams and unsustainably good/bad teams. (And to a lesser extent, players)

    But like all models, they will improve over time, both as we improve them and as we understand them better.

    And as I’ve said a few times in the last few weeks, I’m doing my own part to actively try and enhance those models by rolling in one underutilized dataset, which is the time component.

    Recognizing the value, while also recognizing the shortcomings and working actively to improve on them, are not incompatible!

  17. Halfwise says:

    jake70:
    Related to Oiler trades……David Perron update:Last 8 games

    1 goal
    1 assist
    +1
    avg TOI~17min

    So, about the same point production per 60 as Klinkhammer?

  18. Woodguy says:

    I’m so glad Gmoney posts as much as he does because it means I’m not compelled to and I get a lot more work done.

    On point every post in the thread, correct, well reasoned and very read able.

    I really appreciate your posts sir.

    Also,

    Eakins wrecked Godot!!

  19. Halfwise says:

    G Money,

    I think what you are doing is awesome.

    The only line of argument that I object to is “if it doesn’t show up in Corsi it isn’t real.” To me, that assumes that Corsi is more comprehensive than it is proven to be.

  20. G Money says:

    Pouzar:
    G Money,

    Shot intensity as you call it is exactly what Fangda Li refers to as Shot Quality.

    Looking forward to your findings G. This is the holy grail for me as there is a huge correlation between these high intensity events and goals. Cheers!

    Yeah, man, we are in such early days that the terminology changes for this stuff are going to be head spinning.

    I didn’t want to use the term Shot Quality because a. a bit generic, and b. is already being studied by Chris Boyle in his Shot Quality Project, which looks at type and context of shots rather than timing. (Top notch stuff, by the way. Recommended reading, though he hasn’t posted anything in quite a while).

    Shot Intensity isn’t perfect but seems a bit closer to me.

    When I did some digging on the background to the whole aspect of the temporal aspect of Corsi, I could only find four people who’ve really dug into it:
    – “Our own” Michael Parkatti (and one other unrecalled blogger, who did something similar), who looked at sh% on rebound shots (delta < 5s)
    – "Our own" Tyler Dellow, who looked at Corsi effects after faceoffs
    – Fangda Li, who's been playing with temporal data but more noodling with it, same stage as I am at, but we seem to have different directions for the most part

    I actually emailed Fangda to get some help finding the shot location, that is to say (x,y), not distance, data.

  21. stevezie says:

    G Money,

    I’m glad you were here today.

    Woodguy: Eakins wrecked Godot!!

    Ha!

  22. G Money says:

    Halfwise: I think what you are doing is awesome.
    The only line of argument that I object to is “if it doesn’t show up in Corsi it isn’t real.” To me, that assumes that Corsi is more comprehensive than it is proven to be.

    Thanks mang.

    Yeah, you bet – sometimes it is tempting to get into the whole “it isn’t real”, and that’s a false certainty.

    Maybe more clarity from explicitly applying what is probably a more relevant standard: if it doesn’t show up in Corsi, it might still be real, but you better have some solid objective data and/or analysis to indicate why. Because most previous claims to the contrary i.e. is real but doesn’t show up in shot metrics, after further analysis, proved that the claimed effect was false, not the Corsi effect was insensitive.

    It’s kind of like alternative medicine that way – the 98% of fraudulent claims drown out the 2% of the legitimate ones, but you’re still best off being skeptical when a new claim comes up.

  23. G Money says:

    Woodguy: I’m so glad Gmoney posts as much as he does because it means I’m not compelled to and I get a lot more work done.
    On point every post in the thread, correct, well reasoned and very read able.
    I really appreciate your posts sir.

    stevezie: I’m glad you were here today.

    Aw shucks. Thanks lads!

    * Looks at ground, kicks awkwardly*

  24. Dark Asia says:

    An amusing little discussion but the line that “wins do not determine the better team” is entirely laughable on multiple counts. First off – the entire purpose of playing the game is to “win” as determined by the “score” which identifies the “better team”. If you reject this then you reject the fundamental purpose of the entire exercise. Winning the game does indeed determine the better team on any given night – unless you want to keep score using corsi or possession or whatever – but that’s not how we agree to keep score. No attempts to redefine winning will change that. A “poor” team (definition debatable) can beat a “good team” (definition debatable) on any given night due to the complexity and unpredictability of the system – but the better team will “win” on the scoreboard more often. If this were not true then there would be no point in playing game would there?

    Almost everyone applying the so-called “advanced stats” on this site (which are in fact exceptionally crude and simplistic) adopts an explicitly Newtonian stance and assume that it is somehow possible to describe things “objectively” and all results in a hockey game are the result of interactions between physical “parts” of the game (shots, saves, possession etc). The game for these people is strictly deterministic and causal in nature. These folks who are entirely focused on simplistic stuff like corsi have this delusion that they can watch the game like Cypher from the matrix watching his screen filled with flowing green numbers and saying “redhead, blond, brunette, I don’t even see the numbers anymore”. This is a fantasy because the universe is not mechanical and deterministic – it’s infinite, multidimensional, and interconnected – and so is this trivial little game called hockey. When the stats that people use on this site don’t provide the desired result – they derisively state that this is simply “luck”. Wow how intellectually weak can you get. These observers cannot measure or even perceive even a fraction of the system being observed – and when their model describing the system doesn’t produce the desired results – their only defense for this failure is to conjure up ghosts and witches under the heading of “luck”? It’s laughable really.

    How do you measure creativity, happiness, commitment, motivation, determination, cleverness, adaptability? Difficult right? The simplistic stats people simply ignore these factors or deride them as unimportant while fiddling with their little possession numbers. They are not unimportant – in fact they are critical – but they are devilishly difficult to measure. Eakins appears to me to be the poster boy for simplistic use of stats by focusing on a very narrow interpretation of what is required to win. His approach was the product of a ridiculously naive understanding of what he was attempting to study. Nelson appears to have a much better handle on many of the traditional skills under arguably worse conditions. This drives the simplistic stats people nuts and leads them to try justify Eakins pathetic failure and denigrate Nelson’s limited success. But there is no getting around a simple fact – Travis Eakins was a horrible coach because his team didn’t just lose a lot – they were one-dimensional zombies – and they also never won. Eakins has none of the skills required to teach, create, motivate and lead. We have two examples now where both Kruger and Nelson got better results than Eakins with poorer possession stats – so perhaps this may encourage the simplistic stats folks to re-evaluate their approach which is clearly so limited – and stop denigrating anything their little tools can’t measure as “luck”.

    Simple possession stats and the like have their role – but they are misused more often than not – and explain far less than their strident advocates suggest. In fact it can be argued that the misuse of these simplistic tools is killing the artistry and creativity in the game. Simpletons like Eakins produce what Igor Larionov has described as a “dullness” and I believe a lot of this stems from the inappropriate use of tools like possession stats. Just watch the 1987 Canada Cup highlights of the USSR team play and weep for what we’ve lost.

    http://www.theplayerstribune.com/miracle-on-ice-hockey-russia/

  25. Halfwise says:

    G Money:

    It’s kind of like alternative medicine that way – the 98% of fraudulent claims drown out the 2% of the legitimate ones, but you’re still best off being skeptical when a new claim comes up.

    Good analogy.

    If we had reliable stats on trends in the UFA premium that Edmonton has to pay, or the NMC incidence, we’d have an indicator of how players around the league view the team. That’s my soft proxy for how things are in the room.

  26. G Money says:

    Eh Team:
    Why are we saying Scrivens and Fasth have played better with Nelson as coach?Right now Scrivens is still at .897 save % and Fasth at .888%. Isn’t that were they were with Eakin as coach also.

    They were bad and are still bad.

    The goalies were so bad early that they still look bad even after improving (modestly).

    I think if you calculated the pre-post splits, you’ll find that Fasth has been better and Scrivens slightly better under Nelson. Not by a huge amount, but .890 -> .910 means half a goal less given up every game, and that adds up quickly.

  27. frjohnk says:

    G Money,

    Good discussion.

    Look forward to more of your corsi project.

  28. Jon K says:

    Halfwise: Let me get this right.
    If he trades Petry before other trades, “he jumped the gun.”
    If he trades Petry during the other trades, “he panicked, could have held out for more.”
    If he trades Petry after the other trades, “he waited too long.”
    If no deal gets made, “MacT is stubborn and overvalues his players”
    If MacT spends the rest of the season negotiating an extension, “he’s just trying to cover for his negligence”

    What is the scenario where it’s “MacT handled that pretty well”?

    Answering my own question, if it is Petry 5x$5M or Petry for Antti Raantta or Sobotka’s rights plus either a 2015 2nd or any 21 year old prospect with size, speed and a brain, I will say “MacT handled that pretty well”.

    Well, we don’t really know the reaction as the result has yet to occur.

    If Petry gets a $5×5 extension I’ll be ecstatic.

    If Petry gets dealt for a return similar to what we’ve seen for Sekera and the Franson package, then MacT will have handled it well.

    If Petry gets dealt for a B prospect or for 3rd and 5th round picks, then I think we could count the Petry scenario as a negative on MacT’s resume.

    I’ve been a MacT guy going way back. My comment is speculative and should really only be taken as being a sign of uneasiness as major buyers go elsewhere while Petry remains an unsigned asset a few days from the deadline.

    I mean, we saw this sort of general fear realized with the Smyth scenario. Lowe didn’t look to trade him until the absolute last minute and had to accept the collection of junk that NYI reportedly was offering to everyone at the deadline. The result? A few seasons of Nilsson who had the skill but lacked the will (which was the prior report on him, even from his Dad, an Oiler employee), a prospect with upside to be a 3rd/4th line tweener (which Pierre MacGuire correctly noted at the time and lambasted Lowe for), and a 15th overall pick (the best asset in the deal) that the Oilers wasted on a guy who couldn’t skate or defend. Oops.

  29. spoiler says:

    Woodguy:
    I’m so glad Gmoney posts as much as he does because it means I’m not compelled to and I get a lot more work done.

    On point every post in the thread, correct, well reasoned and very read able.

    I really appreciate your posts sir.

    Also,

    Eakins wrecked Godot!!

    I am finally reading this thread now, and I am thinking the exact same thing. I think a shout-out needs to go to FRJohnK too.

    I am also reminded of when MC79 threw his hands up and asked LT to ban the lazy idiots who were asking the same stupid questions he had been answering relentlessly for 5 years (from other previous lazy idiots)… I believe shortly thereafter he started his own blog.

    There should be some kind of stats primer so the convo doesn’t get de-railed by newbs and trolls into issues that were resolved long ago, like the thread today.

  30. Lois Lowe says:

    G Money killed it today. Good show old chap.

  31. wheatnoil says:

    spoiler:
    I am also reminded of when MC79 threw his hands up and asked LT to ban the lazy idiots who were asking the same stupid questions he had been answering relentlessly for 5 years (from other previous lazy idiots)… I believe shortly thereafter he started his own blog.

    There should be some kind of stats primer so the convo doesn’t get de-railed by newbs and trolls into issues that were resolved long ago, like the thread today.

    The only problem is that everyone starts somewhere. Years ago I read this blog and made occasional lazy, generic, uninformed comments. NOW, after reading through these debates and learning from mc79, Woodguy, G Money, FrJohnK and on the list goes, I now make lazy, generic, mildly informed comments!

    Good show G Money!

  32. Heinz 57 says:

    godot10:
    Making the fundamental mistake of confusing correlation for causation, and there is NO PROOF of causation.

    What you’re really arguing for here is that we’d all be better off living in a small world, because that’s where you are personally most comfortable living.

    95% of what smart people do in difficult jobs where other people are involved lies on the UNPROVEN side of the fence.

    ———

    Imagine you’re The Professor on Gilligan’s Island. One day a perfectly serviceable small engine washes up on shore. You figure you can use this engine to fabricate a small biplane out of bamboo and coconut hair and other materials native to the environment.

    With this plan in mind, the first thing you do is construct a wooden balance beam so as to select the best pilot. Ginger is clearly taller than Mary Ann, but also has a somewhat less voluptuous figure, so this will be narrowly contested.

    They both step onto their end of the plank, it wobbles for a second, then Ginger’s end dips down and impacts the damp earth with a soft thud. Mary Ann lets out a girlish squeal as she bounces a bit off her toes on the high side.

    Your first instinct is to say to Ginger “do that again” but since you’re the professor you think twice about the severity of the matter at hand, so instead you say to Ginger, “lose the gown”.

    Ginger is more than happy to comply with this request. She steps off to step out, and Mary Ann goes down with a thump, letting out another delightful squeal—about half as girlish—as Ginger slides out of her gown and hands it to Mrs Howell, who discretely holds it up to her own shoulders while pretending to sniff at the shocking label. Thurston doesn’t know where to look. He studiously avoids eye contact with Mrs Howell, without exactly looking at Ginger either.

    “Skipper, can you put your toe on Ginger’s end so that Ginger can step back on,” you suggest, which you hope will also help the Skipper break free of his frozen tongue loll.

    Skipper, quick to oblige, reaches out his toe of his crusty Dockers. Mary Ann flies upward—way off balance—her arms instinctively fly out parallel to her pony tails and the entire assemblage—bronzed wings and knotted canards—does a quick double backwards helicopter, while she “eeps” ever so enticingly, before finally regaining her balance.

    “Careful Skipper, you don’t know your own strength,” you say. Skipper turns bright red while Mary Ann glowers down at Skipper’s black flat-top then up at the midget coconut tree providing the operation’s skimpy shade.

    “Oooo, if he does that again, I’m going to fling up my arms and pluck that pair of low-hanging coconuts so I can bonk his bald spot” she thinks to herself, without pausing to consider that for the hard-headed Skipper, any swell is a good swell.

    Ginger steps back onto her end of the balance beam, now wearing an under-wire white two-piece, a diamond bracelet, and beach-casual white stilettos.

    “Wire!” you think to yourself. “What a precious resource! Just what I need to hinge the control surface.”

    You make a mental note for later, but somehow you’re not quite sure how to frame the question. You suspect Ginger has never struggled with shear-stressed hinge stiction. She might not fully apprehend your difficult engineering challenge. If she kicks up a fuss, this particular requisition won’t be popular with the women, and it won’t be popular with the men, either. How about we just leave this technical refinement out of the prototype aircraft, and see how things go? That’s probably the best plan.

    “You can let it go, now, Skipper” you say. Skipper, who is still feeling the heat on the back of his neck through the fronds of the skimpy coconut tree, retracts his foot with exaggerated gingerness.

    This time—now opposing Ginger minus her gown—it’s Mary Ann who slowly descends, while Ginger’s underwire ascends thereabouts to Gilligan’s eye level. Mary Ann, who has got the hang of this now, curtsies slightly on impact, and moves to step off.

    “Not yet, Mary Ann!” cries Gilligan, holding forth his arms like a limp hammock. “Wait for Ginger.”

    “Ginger isn’t going to step down from up there, Gilligan, dressed like that” says Mrs Howell tartly.

    “Quite right,” says Mr Howell who has miraculously fetched his bamboo easy chair, which he wedges under the beam. “With the plank thus secured, she can walk down.”

    “Excellent physics,” you say to Mr Howell. “I’m impressed!”

    “You’ve made me a ramp! How charming!” says Ginger, as she descends with a series of short, practiced cross-steps, gracefully evading her own gown which Mrs Howell abruptly thrusts toward her with her long, white gloves.

    “Looks like we’ve chosen our test pilot,” you say.

    “By a pair of thick and flouncy pony tails,” you add, catching the first hint of Mary Ann’s darkening expression. Mary Ann softens slightly. “How do you keep your hair so lustrous under such primitive conditions?” you continue earnestly, exuding a keen, rarefied intelligence as you contemplate how her sly roborant softener might bear on the vexing problem of aeronautical hinge stiction.

    Having now finished her descent down the improv gazeway to her own satisfaction, Ginger extracts her gown from Mrs Howell—along with Mrs Howell’s elbow-length left glove, whose fingers seem to have snagged on a sequin.

    “Lovey! You’ve lost your glove!” shrieks Mr Howell, who hasn’t actually seen Mrs Howell’s bare forearm since the day they opened their first joint checking account—years and years, even, before the interminable, three-hour cruise.

    Ginger slips her thumbs under the straps, holds her arms up, and allows the gown to unfurl. Mrs Howell’s long glove also unfurls, clinging to a short-seamed prominence by the tip of the index white fingers—like one of Mr Howell’s silk scarves dangling precariously from a blazer pocket in the shape of a disembodied hand.

    “Eep!” go Ginger and Mrs Howell collectively, without the least trace of enticement.

    “I’ll get it!” says Gilligan, who’s previously outstretched arms are still in prime position to provide prompt assistance.

    “No!” yell both Howells, grappling at Gilligan from behind with three arms. Mrs Howell’s bare arm had quickly been tucked discretely behind her beach boa, where it’s now become tangled and useless. If this frustrates Mrs Howell, she doesn’t let on.

    “No!” yelps Ginger simultaneously, recoiling so violently with the snagged garments, she nearly knocks the Skipper backwards onto his giant, rumpled, rum-boil.

    ———

    All this careful optimization of the power-to-weight ratio ultimately goes for naught, when it proves impossible for Ginger to fit herself into the tiny cockpit. She’s plenty agile enough to sit on the edge of the wing and pivot her legs under the coconut canopy, but her legs are simply too long, and after numerous attempts—with copious advice from all quarters—even her most sensible stilettos have made a complete hash of several critical flight surfaces.

    Besides, no-one was quite sure she had exactly the Right Stuff in the first place.

    The obvious replacement is Mary Ann—and she’s probably the most gymnastically inclined person on the entire island—only no-one really wants to see Mary Ann pilot an untried aircraft—secretly, not even Ginger, though she tries not to show it—and as Gilligan is quite obviously nearly the same size—especially as perceived in proximity to the Skipper—that quickly settles matters.

    No-one even bothers to suggest that Gilligan empty the pockets of his cargo pants, though he does loosen his belt after being served a particularly rich breakfast on the morning of the first test flight. The ladies can’t help but cook when they’re nervous.

    “Got to keep up your strength, little Buddy,” says the Skipper, who is so pleased himself with the breakfast bounty that he fails to perceive the portly omen.

    ———

    You—The Professor—are completely lost in contemplation on the vexing question of wing flex. Surely a stiffer wing will provide a more stable flight model, but adding stiffening materials also subtracts from thrust-to-weight ratio. This is bad, not even counting the loss of Ginger, or how much the-loss-of-Ginger just ate for breakfast.

    Do you press forward with soft wings and hope that Gilligan has the Right Stuff to achieve controlled flight? Or do you reduce wing flex and run the risk that the craft fails to gain lift before running out of island?

    With controlled flight, it will be easier to evaluate the aircraft, and thence make further improvements. But if it never rises out of ground effect, this proves nothing.

    Whatever is a thinking man supposed to do?

    If only you had brought Applications of Principia Mathematica to Aeronautical Engineering Using Commonly Available Biomaterials along for the cruise.

    Then you remember the Skipper’s theorem.

    I guess in some matters you just have to cross your fingers and hope for the best, like the Skipper says.

  33. gogliano says:

    Dark Asia:
    An amusing little discussion but the line that “wins do not determine the better team” is entirely laughable on multiple counts. First off – the entire purpose of playing the game is to “win” as determined by the “score” which identifies the “better team”. If you reject this then you reject the fundamental purpose of the entire exercise. Winning the game does indeed determine the better team on any given night – unless you want to keep score using corsi or possession or whatever – but that’s not how we agree to keep score. No attempts to redefine winning will change that. A “poor” team (definition debatable) can beat a “good team” (definition debatable) on any given night due to the complexity and unpredictability of the system – but the better team will “win” on the scoreboard more often. If this were not true then there would be no point in playing game would there?

    Almost everyone applying the so-called “advanced stats” on this site (which are in fact exceptionally crude and simplistic) adopts an explicitly Newtonian stance and assume that it is somehow possible to describe things “objectively” and all results in a hockey game are the result of interactions between physical “parts” of the game (shots, saves, possession etc). The game for these people is strictly deterministic and causal in nature. These folks who are entirely focused on simplistic stuff like corsi have this delusion that they can watch the game like Cypher from the matrix watching his screen filled with flowing green numbers and saying “redhead, blond, brunette, I don’t even see the numbers anymore”. This is a fantasy because the universe is not mechanical and deterministic – it’s infinite, multidimensional, and interconnected – and so is this trivial little game called hockey. When the stats that people use on this site don’t provide the desired result – they derisively state that this is simply “luck”. Wow how intellectually weak can you get. These observers cannot measure or even perceive even a fraction of the system being observed – and when their model describing the system doesn’t produce the desired results – their only defense for this failure is to conjure up ghosts and witches under the heading of “luck”?It’s laughable really.

    How do you measure creativity, happiness, commitment, motivation, determination, cleverness, adaptability? Difficult right? The simplistic stats people simply ignore these factors or deride them as unimportant while fiddling with their little possession numbers. They are not unimportant – in fact they are critical – but they are devilishly difficult to measure. Eakins appears to me to be the poster boy for simplistic use of stats by focusing on a very narrow interpretation of what is required to win. His approach was the product of a ridiculously naive understanding of what he was attempting to study. Nelson appears to have a much better handle on many of the traditional skills under arguably worse conditions. This drives the simplistic stats people nuts and leads them to try justify Eakins pathetic failure and denigrate Nelson’s limited success. But there is no getting around a simple fact – Travis Eakins was a horrible coach because his team didn’t just lose a lot – they were one-dimensional zombies – and they also never won. Eakins has none of the skills required to teach, create, motivate and lead. We have two examples now where both Kruger and Nelson got better results than Eakins with poorer possession stats – so perhaps this may encourage the simplistic stats folks to re-evaluate their approach which is clearly so limited – and stop denigrating anything their little tools can’t measure as “luck”.

    Simple possession stats and the like have their role – but they are misused more often than not – and explain far less than their strident advocates suggest. In fact it can be argued that the misuse of these simplistic tools is killing the artistry and creativity in the game. Simpletons like Eakins produce what Igor Larionov has described as a “dullness” and I believe a lot of this stems from the inappropriate use of tools like possession stats. Just watch the 1987 Canada Cup highlights of the USSR team play and weep for what we’ve lost.

    http://www.theplayerstribune.com/miracle-on-ice-hockey-russia/

    This is basically my position and well articulated. It gets mistaken for an ignorance or distrust of numbers around here — see above for numerous examples — but that’s a pretty facile misinterpretation.

    Eakins was the epitome of a coach making a team into his own leveled image of the world.

  34. Woodguy says:

    Dark Asia,

    More strawmen in your post than a Brian Burke tirade.

    Well done.

  35. Hockeyman 99 says:

    Dark Asia,

    🙂

  36. Dark Asia says:

    I’ll see your strawman and raise the suggestion that your succinct reply is poisoning the well worthy of Kevin Lowe. But if you have any worthwhile comments I’m all ears.

    Woodguy:
    Dark Asia,

    More strawmen in your post than a Brian Burke tirade.

    Well done.

  37. SwedishPoster says:

    G Money: Sorry, my friend, but this is demonstrably false.As last night’s game showed, where the better team kicked the teeth in of the shittier team, and the shittier team won.

    Scrivens won the game for the Oilers who were by far the shitter team on the night – just as he has lost the game for the Oilers when they were the better team on the night.

    Skill and execution win out, but only over the long run.

    And that in the end is why we look at shot metrics and the like, because they are actually the best indicators we have of long run skill rather than short run variation.

    Due to time zones I didn’t see this until now. We’ll have to agree to disagree I think. I’d argue it isn’t luck that has Scrivens posting a great game, it’s preparation and execution. The ability to do it more often is what seperates a good goalie from a mediocre one. It’s not luck that has the d-men giving him free sight on the shots and clearing rebounds, it’s execution, focus. It takes just a tiny lapse and you’ve got a goal against. It’s not luck.
    The two goals forward comes from good hockey plays and skill it’s not luck.
    The difficulty lies in executing night in and night out and if you don’t carry the play you leave very little room for mistakes. Eventually you make mistakes and if the puck is in your own end they are more costly. Has nothing to do with luck. Everything to do with probability but probability you can control at every given second.

    The more skilled squad usually wins in the long run. Agreed. But it’s not luck or even random variation that has the worse team winning from time to time. It’s how well they execute during key moments of the game.

  38. wheatnoil says:

    SwedishPoster,

    I may be wrong, and often am, but I think you guys are arguing the same point but from a different vantage. I believe you’re making the point that it’s not “luck”, it’s focus, execution, skill and other factors. However, you acknowledge that you don’t always get the same outcomes every game. That kind of is the definition of variance. Variance is just how spread out the numbers are in statistics. Your argument is against “luck”, which implies a causal (or lack of causal) relationship with the numbers. Variance just means that the numbers aren’t always the same and that they’re spread out.

    Essentially, whether the fluctuation in percentages that all players and goalies experience is “luck” or “skill/execution/focus” is an arguable point (though I would postulate that it’s both). Whether it’s “variance” is inarguable.

« Older Comments

Leave a Reply

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