Measuring Draft Performance

by Lowetide

I followed drafts as a kid. I remember the 1971 draft and I anticipated the 1973 Amateur Draft. I didn’t do a mock draft (that I recall) but would have if it had occurred to me. We are all limited by the power of our imaginations.

THE ATHLETIC!

The Athletic Edmonton features a fabulous cluster of stories (some linked below, some on the site). Great perspective from a ridiculous group of writers and analysts. Proud to be part of the group, here’s an incredible Offer!

  • New Lowetide: Joel Persson is ideally situated to win an opening night roster spot with the Oilers
  • New Daniel Nugent-Bowman: What the 2019-20 Oilers might look like without trade missteps.
  • Lowetide: Finding the best candidates for the final two spots on the Oilers skill lines in 2019-20.
  • Jonathan Willis: Projecting the Oilers’ opening night lineup, line combinations and more.
  • Lowetide: Does the James Neal acquisition impact Oilers’ prospects in 2019-20?
  • Lowetide: Oilers’ acquisition of James Neal could add badly needed scoring to the top two lines.
  • Daniel Nugent-Bowman: Ken Holland puts his stamp on the Oilers with first big move in Lucic-Neal trade
  • Jonathan Willis: Ken Holland ends an ugly situation for the Oilers by trading Milan Lucic for James Neal
  • Daniel Nugent-Bowman: Potential free-agent options for the Oilers in 2020
  • Jonathan Willis: Which Oilers defencemen can make an outlet pass?
  • Kent Wilson and Lowetide: Why the Flames and Oilers would (and wouldn’t) trade Sam Bennett for Jesse Puljujarvi
  • Lowetide: Looking ahead to Oilers training camp: 35 players for 23 jobs
  • New Jonathan Willis: What the Oilers’ 2020 cap situation suggests about Ken Holland’s master plan.
  • New Daniel Nugent-Bowman: With free agency all but over, Oilers’ Ken Holland has tough work ahead on the trade front
  • Jonathan Willis: Josh Archibald won’t fix the Oilers’ biggest problems, but he’ll help with some key issues.
  • Lowetide: Will the 2019-20 Bakersfield Condors be the Oilers’ best minor-league team ever?
  • Daniel Nugent-Bowman: The Oilers have a new amateur scouting director. What can we learn from Tyler Wright’s track record at the draft?
  • Lowetide: Projecting Darnell Nurse’s next contract and possible trades
  • Daniel Nugent-Bowman: A missing mom, aching feet and looking for Kevin Lowe: A week in the life of Oilers prospect Raphael Lavoie
  • Lowetide: What to do when Connor McDavid rests: The Oilers’ ideal No. 2 line for 2019-20
  • Jonathan Willis: How often do goalies like the Oilers’ Mike Smith rebound?
  • Daniel Nugent-Bowman: ‘He comes as advertised’: Philip Broberg’s skating makes him development camp standout for Oilers
  • Daniel Nugent-Bowman: Oilers plan to skew younger on defence could open the door for Evan Bouchard, Dmitri Samorukov
  • Lowetide: Oilers top 20 prospects summer 2019.

BEST DRAFTS, FIVE YEARS OUT

I rely heavily on games played, but if a team drafts three players who play a substantial role on a regular basis, that’s a strong draft. Edmonton has done it, just not enough since the turn of the century. Also, generational talents and impact forwards to the front of the line.

2015—Generational forward,  322 games. (Connor McDavid 287, Ethan Bear 18, Caleb Jones 17). McDavid’s draft is the only one in team history that can rival those magical drafts of 1979 and 1980, although 2015 probably doesn’t get 500 games in the first five years (one more year to go). It would be ridiculous to rank any draft year this century ahead of this one.

2010—Impact F, depth D, 425 games. (Taylor Hall 299, Martin Marincin 85, Tyler Pitlick 27, Brandon Davidson 12, Curtis Hamilton 1, Tyler Bunz 1). Taylor Hall was the first No. 1 overall pick in franchise history—incredible considering what has happened since. Brandon Davidson and Martin Marincin, along with Tyler Pitlick, teased and then emerged after the five year marker.

2011—Top 3F, Top 4D, Top 9F, 578 games. (Ryan Nugent Hopkins 313, Tobias Rieder 154, Oscar Klefbom 107, David Musil 4). A quality draft with a mighty head start, No. 1 overall and another player inside the first round. The Tobias Rieder pick was a strong moment for the scouts.

2014—Impact forward, 351 games. (Leon Draisaitl 351). MacT had two drafts and hit a home run when deciding on Draisaitl. The big man scored 50 goals in his draft+5 season, extremely rare in team history.

2001—Top 3F, backup goalie, 471 games (Ales Hemsky-275, Jussi Markkanen-128, Ales Pisa-53, Kari Haakana-13, Doug Lynch-2). Hemsky represented the best skill forward drafted by Edmonton and successfully brought to the NHL since Ryan Smyth in 1994.

2002—Top 6F, Top 6D, 440 games. (Jarret Stoll-286, Matt Greene-151, Mikko Luoma-3). A little misleading in that Stoll was a draft re-entry and technically belongs to the 2000 group. Greene and Stoll had long and productive careers. Unlike the 2010’s, both men were second rounders, top pick Jesse Niinimaki washed out.

2008—Top 6F, 250 games. (Jordan Eberle 195, Teemu Hartikainen 52, Phil Cornet-2, Johan Motin-1)Eberle took two years in junior and then played well from his rookie year. Hartikainen looked like a player, but could not survive all the management chaos in his time with Edmonton. They didn’t have their second or third-round picks.

2013—Top 4D, 300 games. (Darnell Nurse 197, Anton Slepyshev 102, Bogdan Yakimov 1). It took some time to get Nurse rolling, but his most recent seasons suggest the club drafted a top-4 defender. It would have been fab if Slepyshev had also made it.

2007—Top 6F, 447 games. (Sam Gagner 366, Linus Omark 65, Alex Plante-10, Riley Nash-5, Milan Kytnar-1)So much promise—three picks inside the top 21 overall—and so little to show for it. Gagner is now over 800 NHL games, but did not cover his draft bet. Nash took forever to arrive but is now at 477 NHL games.

2005—Top 9F, 339 games. (Andrew Cogliano-246, Danny Syvret-49, Taylor Chorney-44). Cogliano emerged as a solid player and has been a consistent performer over many years now. Chris VandeVelde, not listed here, arrived after the five-year window closed.

That’s 10, with the odd man out being 2003. Kyle Brodziak couldn’t save it from the seventh round. The 2016-19 draft also not listed, too soon to know.

LOWDOWN WITH LOWETIDE

At 10 this morning, TSN1260, we have a fantastic group of guests. Aaron Bronsteter from TSN will talk UFC 240 this weekend, and the growing excitement for the big event. Andy McNamara from TSN4Downs drops in at 11 to talk NFL, and at 11:25 it’s Gerry Moddejonge from Post Media to tee up the Eskimos game tonight.

120 comments
0

You may also like

0 0 vote
Article Rating
120 Comments
Newest
Oldest Most Voted
Inline Feedbacks
View all comments
Tesla's Hair

Conflict Resolution. Pick a style and see which one works well for you. I pick Yield

Avoidance- read his last line or skip over, or avoid blogs where he hangs out completely, opt out of all two way hockey discussions

Competition- show me your stats, im smarter than you and ill show you

Compromise- let him continue posting, respect his opinion by agreeing to disagree

Yield – see him as a genius and see if he posts less inflamatory remarks

Collaboration- find a win plus win, take the guy out to lunch at his fav restaurant, hear some of his epic tales of love and life, come away a wiser petson with 1 less enemy to inflame you

treevojo

leadfarmer,

ArmchairGM,

Cassandra,

Usually I just skip down to the last paragraph of the BEARS posts to see which poster he is calling an idiot this time to give me a chuckle.

The rest of it I find unreadable nonsense.

Works for me.

Cassandra

We can add selecting on the dependent variable as another basic social science error that Ricki commits.

Coding certain D as “rovers” when no such position exists couldn’t be more binary.

ArmchairGM

leadfarmer: Don’t worry you will just be called an idiot
Followed by he’s been tracking bees for 20 years
Followed by a social justice warrior rant
Followed by the history of 2 ply toilet paper

I don’t get why LT allows him to treat his guests on here this way
The loudest voice in the room used to get pushed out by the smartest in the room in the past
But the smartest people have given up
Wood guy was getting called a plagiarist and rarely post and when he does he does not acknowledge this poster because of blatant attacks on him
Bruce will chime in one the outskirts but you can tell he doesn’t want to deal with this
I get called an idiot on a daily basis (I’m definitely not the smartest man in the room)
I’ve been coming here at least every day for 12 years
LT put me through residency, fellowship and first 5 years of practice
Today will be my last
Good luck to all

Sad to see you go, Leadfarmer. I used to get frustrated with Mr thebear’s posts too, until I stopped reading them – even if he is replying directly to me. This may not work / be an option for you, but it certainly has made a difference for me.

There are a few other posters who I sometimes skip over for one reason or another, but nobody that gets consistently ignored like Mr thebear. Although I wish him a full and speedy recovery.

Revolved

leadfarmer,

Oh come on, Ricki’s quite warm and cuddly. His posts are just so big he doesn’t realize when he’s mauling other posters. I have time for a lot of his theories, though I think it’s important to remember that every idea is built on other people’s work and is only valuable when communicated properly, refined and accepted by the community.

Revolved

Scungilli Slushy,

Do you have any idea how freely available the puck tracking data will be? It would create incredible innovation in player valuation and strategy if it was freely available and open source.

leadfarmer

OilFire: Ricki:Tell me what you mean by “multivariable” that academia is just finally catching on to.

Please, humor me and write out a short (3 sentence or less) definition of this new idea.Also, please avoid describing it imprecisely by listing comparisons or examples if you can.

Don’t worry you will just be called an idiot
Followed by he’s been tracking bees for 20 years
Followed by a social justice warrior rant
Followed by the history of 2 ply toilet paper

I don’t get why LT allows him to treat his guests on here this way
The loudest voice in the room used to get pushed out by the smartest in the room in the past
But the smartest people have given up
Wood guy was getting called a plagiarist and rarely post and when he does he does not acknowledge this poster because of blatant attacks on him
Bruce will chime in one the outskirts but you can tell he doesn’t want to deal with this
I get called an idiot on a daily basis (I’m definitely not the smartest man in the room)
I’ve been coming here at least every day for 12 years
LT put me through residency, fellowship and first 5 years of practice
Today will be my last
Good luck to all

Ranford.85

I understand why it didn’t hit the mainstream media…

But after Jujhar signed his extension this summer, he talked about training and hanging out with RNH and Gagner. Burnaby Nuge got hitched this summer and JK was there, along with Jultz, Ebs and Klefbom.

Unicorns…
McD-Kass (skill, speed and actual protection… watch the highlights)
Drai-Chai (numbers don’t lie)
Jujhar-RNH (toughest zone starts, friends, PK/saw off the comp 5v5)
Nygard-Haas (speed, skill and experience/maturity)

Conclusion: If Nuge and Jujhar lined up against every teams best line and came out 50/50… Drai and McD could destroy on different lines.

Mr DeBakey

Rutger Hauer died this week.

https://www.theguardian.com/film/2019/jul/25/rutger-hauer-obituary

First time I saw him was at the Varscona in “Soldier of Orange.”

Ranford.85

This completely off topic.

I’m a lowtide/athletic junkie and growing up Cgy (born in Edm) has made my love for Oil an interesting journey.

Anyone have a stat on Lucic’s hits given/taken? I feel like Looch had a target on his back from day one of his contract with the Oilers. I do know he always at the top of the league for hits given….

I feel like the Flames are now that one team that everyone will want rough up. It’s hard to provide protection for the stars when you’re on the 3rd/4th line.

Ps. Anyone see that Kassian highlight package on the Oilers IG? Wow

Faustkarz

rickithebear,

Firstly no, PDO is a dumb statistic simply by the basis of having a personal shooting% 20 and on-ice save% 80 is dissimilar to a 5% shooting and 95% because the skater’s effect on his own shooting percentage is stronger than their affect on a goalie yet both = 1 mutatis mutandis

evga seems to be far more variant than just FZ Trap, having two stay at home D (2-1) and whether they block shots (0% closed shot theory). ie most common forward linemates, outlet pass ability success/completion rate, partner’s ability

Can you definitively say that in the NHL that two rovers playing with each other have NEVER outscored two 2-1 defensemen when playing on the same pairs, accounting for linemates, zone starts, xSV%, etc.?

I do agree with some or most of what you say but calling something binary or multi-variate without a legitimate algorithm which produces a definitive number is meaningless to me. It is no different than your idea of constructivist narrative building because YOU are choosing the 5 variables but with no definitive proof they are the 5 variables WE should be looking at.

You say evga is a function of 5 variables but at what rate do all 5 affect evga. Home ice quality I am sure affects evga but at an unmeasurable (probably could) rate. You can have infinite variables but which one’s are statistically significant as the “academics” say.. that is up to actual algorithmic analysis and generation and the results therein not by narrative building.

Scungilli Slushy

OriginalPouzar: Why do these experienced coaches switch to man to man and stick with it after it continues to lead to negative results?

Bull headed and attributing success to the wrong factors like having matured exceptional players.

The Oilers have exceptional players but the key ones aren’t mature as players yet, and also, the game has changed. It might have been a hack given nobody does it, if the situation suited. But it didn’t. That was the fail in judgement.

pts2pndr

v4ance:
When I did a draft analysis of the years from 2000-2010, the cutoff I used was 100 games/player to try find an average drafting result.From that data, if you got 5 players out of every two draft years, you were doing about the average.Elite drafting teams tended to be 3 players per draft but the Oilers always hovered around the middle or the lower end of the spectrum.

From other studies, the average career of an NHLer turned out to be ~250 games or about 3 seasons.This basically represents a 3rd line player or a top 6 player whose career was shortened due to injury.Top line guys or 1st pair defencemen usually get to about 500 games.

From a holistic team building perspective, getting a draft pick to play 100 games should be regarded as a success. I know some people refer to picks as magic beans and players who only do 100 games as replacement level players but you do want as many of these guys or lottery tickets as possible.There’s always roster churn as the bottom 6 guys cycle in and out as they are found to be not good enough to stick around or to be promoted to higher in the lineup.

When you can’t get enough picks or players to this minimal level, you end up overpaying for bottom 6 guys as UFAs which is doubly damaging with a cap limit.You need as many cheap homegrown guys as you can find so you can allocate the maximum amount to star players.“Pay for elite talent, scrimp on the complimentary or replacement players” should be the overriding mantra.Draft picks are a renewable resource… cap money is not

If we look at the situation of Zack Kassian, he really shouldn’t get over $2 million going forward.His success with McDavid is nice but away from McDavid, he’s like any other bottom 6 grinder.Don’t overpay for guys who are carried by elite talent.We should have learned that from Maroon, Chaisson, Rattie, etc. by now.Next year’s one year wonder could be Khaira or Maxisimov or Benson.We don’t know who will be on McDavid’s wing so don’t overpay for them if they’re not a guaranteed top line player.

That’s why the salary tiers broken down into 1st line, 2nd line, etc. would and should never work.On paper, it sounds nice but one year a guy can go on a heater onto the first line but if you pay him for that temporary bump, the next year you have an overpaid bottom 6 player and a salary cap anchor (AKA the Chiarelli method).Plus entitlement would be an issue as every 2nd liner who spends a few minutes elevated up the lineup would be wanting to be paid as a 1st liner, every 3rd liner elevated to 2nd will be wanting to be paid as a 2nd liner, etc. etc…

A quality GM should always grind hard in salary negotiations because you never know when you might need that extra money. It might lose you a few players due to hurt feelings or entitlement just like in the Petry situation but it might also allow you to sign Hall to a bargain $6 million.If you can shave some salary on 50% of your contracts, you have more leeway in case you do make mistakes or in case you want to fit in an opportunistic acquisition from another cap strapped team. Salary cap space is an asset and if you don’t try to maximize how you use it, you miss opportunities to improve or you do long term damage with cap anchors.

The Petry stupidity started the landslide of mistakes including the trading of Taylor Hall. Ego should not enter into the bargaining process. There is a fine line between hard bargaining and unfair bargaining. Cheap or unfair bargaining reaps its own rewards and they’re not pretty.

OriginalPouzar

Ryan:
From a recent thread, we were discussing man-to-man vs zone d.

Hitchcock and Mclellan both experienced coaches preferred the man-to-man d zone coverage.

There was a quote from an unnamed dman who cited the Blues struggling under Mike Yeo’s man-to-man d zone system before he was replaced but Berube who had implemented zone d coverage.

Funny how I found an article that suggested the Blues had success under Hitchcock with a zone d system then came unglued the following season when he implemented a man-to-man d zone system. He was fired and replaced by Mike Yeo who subsequently had success with a zone coverage system.

https://www.stltoday.com/sports/hockey/professional/blues-get-big-results-from-small-changes-on-defense/article_12cc2fad-a2f0-50ff-8407-0ffe3da10c09.html

So basically, Hitch has success with zone d, switches to man-to-man d, it doesn’t work his team struggles then he gets fired from the Blues.

Mike Yeo is hired, he reverts back to zone d, the team has a modicum of success. Next season, he switches to man-to-man d, the teams struggles, bottom of the league and he gets fired.

Berube comes in, implements zone d, team makes the playoffs and wins the cup.

By several accounts man-to-man d zone coverage doesn’t work post lockout without clutch and grab, so what’s going on here?

Why do these experienced coaches switch to man to man and stick with it after it continues to lead to negative results?

Scungilli Slushy

Ryan:
From a recent thread, we were discussing man-to-man vs zone d.

Hitchcock and Mclellan both experienced coaches preferred the man-to-man d zone coverage.

There was a quote from an unnamed dman who cited the Blues struggling under Mike Yeo’s man-to-man d zone system before he was replaced but Berube who had implemented zone d coverage.

Funny how I found an article that suggested the Blues had success under Hitchcock with a zone d system then came unglued the following season when he implemented a man-to-man d zone system. He was fired and replaced by Mike Yeo who subsequently had success with a zone coverage system.

https://www.stltoday.com/sports/hockey/professional/blues-get-big-results-from-small-changes-on-defense/article_12cc2fad-a2f0-50ff-8407-0ffe3da10c09.html

So basically, Hitch has success with zone d, switches to man-to-man d, it doesn’t work his team struggles then he gets fired from the Blues.

Mike Yeo is hired, he reverts back to zone d, the team has a modicum of success. Next season, he switches to man-to-man d, the teams struggles, bottom of the league and he gets fired.

Berube comes in, implements zone d, team makes the playoffs and wins the cup.

By several accounts man-to-man d zone coverage doesn’t work post lockout without clutch and grab, so what’s going on here?

Calls from inside the house is my guess. Not getting the hoped results, looking for the answers, and coming up with the wrong problem as the issue to begin with.

Thus my comment about data as the answer to problems. It only lights up the room, you still have to find what you’re looking for with correct decisions.

It’s human condition that wreaks havoc across the board, not just hockey.

OilFire

rickithebear: I look forward to having my perception changed on the multivariable options on the multivariable intelligence diffrent brain functions required for full intelligence. that academia is now addressing.

Ricki: Tell me what you mean by “multivariable” that academia is just finally catching on to.

Please, humor me and write out a short (3 sentence or less) definition of this new idea. Also, please avoid describing it imprecisely by listing comparisons or examples if you can.

Ryan

From a recent thread, we were discussing man-to-man vs zone d.

Hitchcock and Mclellan both experienced coaches preferred the man-to-man d zone coverage.

There was a quote from an unnamed dman who cited the Blues struggling under Mike Yeo’s man-to-man d zone system before he was replaced but Berube who had implemented zone d coverage.

Funny how I found an article that suggested the Blues had success under Hitchcock with a zone d system then came unglued the following season when he implemented a man-to-man d zone system. He was fired and replaced by Mike Yeo who subsequently had success with a zone coverage system.

https://www.stltoday.com/sports/hockey/professional/blues-get-big-results-from-small-changes-on-defense/article_12cc2fad-a2f0-50ff-8407-0ffe3da10c09.html

So basically, Hitch has success with zone d, switches to man-to-man d, it doesn’t work his team struggles then he gets fired from the Blues.

Mike Yeo is hired, he reverts back to zone d, the team has a modicum of success. Next season, he switches to man-to-man d, the teams struggles, bottom of the league and he gets fired.

Berube comes in, implements zone d, team makes the playoffs and wins the cup.

By several accounts man-to-man d zone coverage doesn’t work post lockout without clutch and grab, so what’s going on here?

rickithebear

Cassandra: Not only that but he has no idea how hockey is played.His “theorem” of the structural “advantage” of centers in scoring because they play in the middle of the ice might reflect table hockey but it doesn’t reflect actual hockey.In actual hockey there are no positions once you leave the defensive zone. The data is meaningless.

Moreover, I bet he is calculating his average per position based on how it is commonly listed which, as we all know, does not reflect where players actual line up on a shift by shift basis.

Finally, just the other day he admitted he had no idea what was meant by replacement player.Which means he does not understand the concept of marginal gains.

I understand War as it relates to a perfect binary sport like Baseball.

The large factor influences are understood by everyone so marginal gains are the way you try to beat the other team on a seasonal basis.

Then they build a playoff system were you do not have to face a 5 man rotation base.
Throws all analytics out the window.
The teams with the 2 most elite pitchers has a high prob of moving thru in a series.

But to ignore 100% to 700% influences for marginal gains in hockey is a fools game.
Until everyone is playing like robots and have eliminated large margin factors.
What does a marginal gain tell you when large factors have scued any result.

Of coarse there is F1, F2, F3 Analysis.
But a move to a center based attack is still part of the equation in all even play.

Yup I do not know as much as you who has tracked the sides of attack of each player.

Not likening the level of vitriol I am directing over the last while.

Large factor influences have to be played same way by all.
To have marginal gains matter.

It is not baseball!

JimmyV1965

defmn: I hope that is not referring to what I posted because that is not at all what I wrote.

My apologies. I misread your post. This is what you said;

And I would add that if you want Russell to facilitate a trade at the deadline the best way to do that is to make it clear that staying with the Oilers will sewer any chance he has at a next contract because press box dmen do not command a lot of money come UFA season the following summer.

I just assumed this was an ultimatum by the GM. I’m still not sure what you meant then.

Scungilli Slushy

Russell will have a significant role on the D Corp unless traded. Maybe Holland and Tip are cagey enough for us to see our first pump and dump in decades!

rickithebear

Faustkarz:
Ricki, for the WOWY of those defensive pairs, wouldn’t you need more variables to make it less binary such as FWD’s on the ice, OZS%, EliteComp%, evgf/60, on-ice sv%, xGF%

isn’t evga/60 a binary statistic? seems to more suit a narrative than to shine a light.

Wouldn’t a multivariable analysis be multi-variable? Not just evga/60

Damn straight.

The fact that you did not list that evga is a function of
Fwd NZ trap, GA structure 2-1 or 1-1, and seperste Dpair and goalie performance.
Tells me you are being facetious.
Likely a fan of a PDO.

We can identify the 3-2-1 dmen and 3-1-1-1 dmen.
We know Nurse, Klefbom, Jones, Bear are 3-1-1-1
1-1 breakdown of 2-1 structure is 100% rovers cause until 2-1 structure is recovered.

We can look at a Corsi faced rates from our forwards.
All are big minute Dmen faced 56-59 CA/60
Not normal large variance that can occur.

We can adjust for faceoff zone start
Russell 45.2%
Larsson 46.2%
Nurse 46.9%
Sekera 46.9%
Gravel 48.4%
Klefbom 49.0%
Garrison 50.4%
Benning 50.7%
Manning 53.0%

We can adjust for bench change zone start with without puck.
On the fly starts/60 are indicated but not differentiated by with or without.
With a faceoff to on fly ratio for each dman
But no database I found differentiates with or without.

What we know from WOWY
Larsson & Russell are still able generate avg evga with rovers Klefbom & Nurse.
But Benning & others are unable to deal with rovers.

OriginalPouzar

defmn: Good point. I think there is usually a market for veteran dmen at the the trade deadlne though. But you are right. That would make it more difficult.

Also, the general market at that time is for players on expiring contracts, not those with a full year of (bloated) term left.

defmn

OriginalPouzar: At the same time, how is it possible to get a clean (or anywhere near clean) disposition of Russ next off-season without him playing games?

Good point. I think there is usually a market for veteran dmen at the the trade deadlne though. But you are right. That would make it more difficult.

Scungilli Slushy

Many love the story of young people with limited experience coming in hot with fresh big ideas and blowing everything up.

The thing is it rarely happens in established industry. The NHL isn’t some old dusty relic where everyone is a dinosaur and is 50 years out of date.

Even Vic Ferrari said he thought GMs knew what they were doing. He works for one now.

Yes this is a math blog as the host says, but it remains that data is a lens on the game, not the frame.

When teams are wrong it isn’t a data problem necessarily, all teams now do that, it’s a concept problem.

Chiarelli isn’t a dense man, he’s just not right in his philosophy. McLellan isn’t a bad coach, he’s made errors in managing people and strategy like Hitch. They can change, if they don’t then they’re dinosaurs.

For Dubas and Chayka, show me consistent high level success and I’ll buy the hype. We don’t know yet just like drafted players.

The game of hockey isn’t as big as trying to analyze world wide industry and national trends in business. A person like Holland or any correct thinking individual with decades of experience should be able to read the lay of the land pretty quickly and know who’s getting things done.

All of you who are good at your jobs can do this, why shouldn’t they be able? Data can give the fine grain insights into a team after the big picture is set. It’s a tool to delve deeper and refine. But it’s a tool, not the hand that wields it.

My instinct is those (which are few) that try to run teams by analytics as the primary decision maker won’t ultimately succeed, that isn’t Billy Bean at all.

Scungilli Slushy

NHL teams are not going be able to manage crunching puck tracking data. They won’t want to. Somebody’s going to make a pile doing that.

Cassandra

leadfarmer:
Cassandra,

Well he doesn’t get that the reasons most top scorers are centers is not positional on the ice but because most top players are played at center at a young age to get them more involved at every point on the ice.
One of the ways coaches have used to get an offensive boost is opposite of bear theory
Example of this is Giroux moving to wing

Exactly. There exists an incredibly obvious alternative hypothesis to his “facts” and not only does he not investigate it, it doesn’t even occur to him.

And his conclusion is so, so, wrong. It is more likely that it is harder for centers to score goals because they spend more of the game as F3, because they spend more of the game lower in the defensive zone.

Scungilli Slushy

Russell 3LD with a competent partner would be just fine and a solid fill in at 2LD.

The problem is everyone is LD, so it’s either rookies or him at RD.

Maybe we see a balance miracle still.

Players not being ideal is different than useless given no better option.

Ben

My current anxiety about next season mostly involves Persson making the team.

Gene: “Hey wait…Persson…that sounds just like the word ‘person’!” I can just swap the player name right in there where the word would go! Quinn’s gonna shit his pants!”.

leadfarmer

Cassandra,

Well he doesn’t get that the reasons most top scorers are centers is not positional on the ice but because most top players are played at center at a young age to get them more involved at every point on the ice.
One of the ways coaches have used to get an offensive boost is opposite of bear theory
Example of this is Giroux moving to wing

Cassandra

leadfarmer:
Ricki is king of binary thinking
Evgf for forwards 1 or 0
GA for defensemen 1 or 0
The most binary analysis on the blogosphere

Not only that but he has no idea how hockey is played. His “theorem” of the structural “advantage” of centers in scoring because they play in the middle of the ice might reflect table hockey but it doesn’t reflect actual hockey. In actual hockey there are no positions once you leave the defensive zone. The data is meaningless.

Moreover, I bet he is calculating his average per position based on how it is commonly listed which, as we all know, does not reflect where players actual line up on a shift by shift basis.

Finally, just the other day he admitted he had no idea what was meant by replacement player. Which means he does not understand the concept of marginal gains.

Bulging Twine

Professor Q: Mahe help guide a strong analytical future for the Oilers.

well done

Bulging Twine

Revolved:
The thing about puck and player tracking is that it is going to push even well developed analytics departments to have to change their approach. The kind of big data that such tracking will create is going to require rooms full of programmers that are capable of creating algorithms for processing it in different ways. These people are in such great demand from many different sectors that it will be difficult for NHL teams that are not already connected to such networks to recruit them and give them the proper guidance.

Yes those teams without good analytic departments are behind the 8 ball. They’ll have to hire people to hire people.

Where as the teams that have analytic departments already established are already onto the next step of anticipating and preparing for the new data.

Professor Q

Bulging Twine:
Do we know what the Oil have for analytics?There is nothing on their website.
Shayna from The Athletic has Justin Mahe as Manager of Hockey Analysis.Do we know anything about Justin Mahe?When he was hired?Anything?

Mahe help guide a strong analytical future for the Oilers.

leadfarmer

Ricki is king of binary thinking
Evgf for forwards 1 or 0
GA for defensemen 1 or 0
The most binary analysis on the blogosphere

GMB3

Bag of Pucks: The saving grace is the vast majority of hockey GMs are still pretty old school on analytics. Chayka and Dubas are outliers, not the norm.

So most of the guys Holland is competing against approach it from the same perspective and skillset.

Innovation works quickest when circumstances force it. When an NHL team starts to win Cups and attributes much of their competitive advantage to analytics, that’s when the rest will get onboard. I suspect the puck tracking data will be key to this as to date, the data has largely been proxies. With puck tracking, they’re going to have directly correlative datasets and it’s much easier to draw insights from that. Even still, building effective decision models off that big data isn’t easy and that talent is expensive and in demand by everyone.

We’ve been talking about it for years, but the reality is it’s very very early days for analytics in hockey. Most of the clubs are dabblers not adopters. And outside ARZ and possibly FLA, there’s very few instances where data is driving the decision models and organizational culture.

I believe Carolina and Vegas belong in here as well. Seems lots of the moves Carolina has made have met the approval of the analytics community on twitter in the least

Revolved

The thing about puck and player tracking is that it is going to push even well developed analytics departments to have to change their approach. The kind of big data that such tracking will create is going to require rooms full of programmers that are capable of creating algorithms for processing it in different ways. These people are in such great demand from many different sectors that it will be difficult for NHL teams that are not already connected to such networks to recruit them and give them the proper guidance.

Revolved

godot10,

I agree that we probably need to start running two rookies in the starting six on opening night. I like the idea of dressing Russell as 7D in order to reduce the minutes played by those rookies, though. I think we can get by with 11F if we run the three centres with more balanced minutes and spotting them with the bottom two wingers.

Bulging Twine

Bulging Twine:
Do we know what the Oil have for analytics?There is nothing on their website.
Shayna from The Athletic has Justin Mahe as Manager of Hockey Analysis.Do we know anything about Justin Mahe?When he was hired?Anything?

I did find this from 2016
https://www.nhl.com/oilers/news/oilers-announce-changes-to-hockey-operations/c-890460

Bulging Twine

Do we know what the Oil have for analytics? There is nothing on their website.
Shayna from The Athletic has Justin Mahe as Manager of Hockey Analysis. Do we know anything about Justin Mahe? When he was hired? Anything?

OriginalPouzar

defmn: This is correct imo. And I would add that if you want Russell to facilitate a trade at the deadline the best way to do that is to make it clear that staying with the Oilers will sewer any chance he has at a next contract because press box dmen do not command a lot of money come UFA season the following summer.

Nice to have Russell in case of injuries or slumps but he is not part of the future and he is not going to cost or win enough games this year that he shouldn’t be thought of as the spare guy.

At the same time, how is it possible to get a clean (or anywhere near clean) disposition of Russ next off-season without him playing games?

OriginalPouzar

GordieHoweHatTrick:
I have to keep pinching myself. It wasn’t just a dream. Lucic contract is gone!

It is indeed gone and it is am amazing and wonderful thing.

Of course, its currently been replaced by another currently very ugly contract, however that contract is not nearly as ugly and not nearly as “clingy” (which is very important) and it actually has a chance to spruce itself up to a half pretty level.

OriginalPouzar

Durag: Is there any site that breaks down EV and PP SH%? Just by eye, it seemed like Leon shot damn near 50% on the PP. I wonder if the EV SH% is more sustainable.

20% at 5 on5

23.44% at 5 on 4

https://www.naturalstattrick.com/playerteams.php?fromseason=20182019&thruseason=20182019&stype=2&sit=5v4&score=all&stdoi=std&rate=n&team=EDM&pos=C&loc=B&toi=0&gpfilt=none&fd=&td=&tgp=410&lines=single&draftteam=ALL

OriginalPouzar

KingerOilredux: – That is all true, and feel good stories.But the purpose of drafting is ultimately to get players into the NHL, not just to outperform their drafted slot.Great the 7th rounder “outperformed”.He’s a “victory on some level, but doesn’t matter.

– But we need to draft guys who play in the NHL at above replacement level to move the needle.We haven’t done that.Things won’t change untill we do

– The problem with just taking games played, or doing better than expected is that it doesn’t measure what impacts resultsDrafting 1 Brodziak and 3 duds in 7th round than 4 Kemps, (I’m assuming Kemp doesn’t make the NHL, for my illustration)

I don’t disagree – the main purpose of the draft is to acquire NHL players and NHL players that impact the lineup but there are secondary and tertiary purposes as well. There needs to be players competing for NHL jobs, even if they don’t make it, there needs to be players to fill out the AHL and ECHL teams and its important the AHL team have success.

If Phil Kemp signs a contract with the Oilers and plays 150 games in Bakersfield over the course of a few years and compete for a job with the Oilers and maybe get a cup of coffee – that have value to the organization.

OriginalPouzar

DarthTu: The idea of a Nurse Larsson “top” or shutdown pairing to play against the opposition top lines has really grown on me.

I guess the Klefbom pairing would really be down to who’s performing better at any given time, be it Persson or Benning.

If Persson shits the bed then we always have the option of bringing Russell back into the lineup and bumping Benning or Jones up to RD with Klef.

I’m completely excluding thoughts of Bear or Bouchard jumping from the AHL for now.

Completely agree on the last line – as at the end of last year Lagesson and Jones are both well ahead of Bear as far as NHL readiness – hopefully Bear has had his deficiencies (defending speed on puck retrieval and off the rush and general board battles) and has been working on related areas this off-season.

Its imperative that the organization is risk-adverse with its most important prospect and allow Bouchard to get some regular top 4 non-sheltered minutes in the AHL even after he puts up 6 points in 4 preseason games and makes a few tantalizing passes. Unlike other years, there are actual legit options to allow that to happen.

Bulging Twine

Sean Tierney Retweeted

Michael Russo
@RussoHockey
·
Jul 22
With the #mnwild working toward refilling their analytics department, I’m hearing GM Paul Fenton has hired Hockey Canada’s Mat Sells. He did advanced pre-scout and analytics

defmn

JimmyV1965: What I was really objecting to was the notion that the GM should give him some kind of ultimatum; agree to be traded or your in the press box. Silly idea on so many levels.

I hope that is not referring to what I posted because that is not at all what I wrote.

defmn

JimmyV1965: Like him or hate him, Russell is our fourth best dman right now. He will be deployed that way by the coach until other dmen pass him on the depth chart. The GM might want to trade the player yesterday, but it won’t change the way the coaching staff deploys him.Russel will be press boxed if and only if the coach thinks he’s the seventh best dman on the team; not some ultimatum by the GM.

I don’t like or hate Russel. 😉

I disagree that on ice deployment is decided by who is the best player though. I think on ice deployment should be decided by the best fit.

The idea isn’t to assemble the best players. The idea is to assemble the best team. Those two aren’t exactly the same thing imo.

OriginalPouzar

I don’t imagine Russell not being in the lineup on game 1, subject to injury.

Trust me, I’m more excited about Jones and Persson and their puck moving and think that Lagesson could absolutely handle the 3LD (and would prove to be a plus 3LD in little time) but I think Tippett will have his veteran in the lineup even with his verbal re: time defending and moving the puck as 5 man units.