RE 15-16 CAM TALBOT: PHOTOGRAPH

Among the first issues facing Peter Chiarelli when he took over the Edmonton Oilers was the big hole in goal. He traded for Cam Talbot—the numbers rhymed and the price was right—and moved on to the next task at hand. Edmonton signed Talbot during his first season as the franchise goalie, and the plan is to move forward with him as the No. 1 man. Can Cam Talbot get this team to the playoffs? (Photograph)

RE 15-16: 50GP, 2.40 .920

ACTUAL 15-16: 56GP, 2.55 .917

  1. Your RE was pretty close. Yes, the even-strength number needs to be better, but a solid year.
  2. Even-strength number? Yes. He was .920 at evens, ranking him No. 20 among goalies who started 40 or more games. He needs to post a better number.
  3. Where was he overall? Talbot was .917, that ranked him No. 15 among goalies who started 40 or more games.
  4. Are you confident in him? I am, there are people I respect who are less convinced of Talbot. With new equipment coming in, one worries about a slow start.
  5. Oilers have had enough of those for a lifetime. Yes, agreed. Talbot was slow off the the mark last fall, he started 3-8-1 and had a save percentage of .897 in October and .864 in November.
  6. But he was money from there? By month, his save percentages were .934, .932, .914, .935 and then .889 in three April games.
  7. Are you satisfied with him as a 55-game starter? Yes. I would like to have seen the club sign a better backup, though.
  8. We can chat about that when we get to the next RE. No, this is the last one, I waited for the draft and free agency so we could wrap everything up here.
  9. What a rip job!!! What about the coach and GM and Hall and Nuge? They are all here, 27 REs including the final Taylor Hall item. Enjoyed doing it, love Blue Rodeo so this was fun.
  10. You are going to miss Hall. Not as much as the Oilers.
  11. Will Talbot play 60 games next season? I haven’t projected it, but suspect Edmonton will rely on him as much as they did in 2015-16.
  12. Not a fan of Gustavsson? I think PC probably decided to sign a mid-level free agent because he believes in Laurent Brossoit to take over the job as backup at some point this season.
  13. A sound piece of reasoning. Well, I think the Oilers make a lot of decisions that assume things that may or may not be true, and Gustavsson-Brossoit could be one.
  14. What do you mean? If Edmonton has goalie issues in October, the club will have to pay extra in trade to solve them. Why not aim a little higher on July 1? Goaling is damned important, you know.
  15. Edmonton rarely makes a move involving goalies in-season? Oh good grief. Bryz, Dubnyk, Fasth, Scrivens, where you been Mac? If Edmonton had taken a little more care, aimed a little higher, we could talk about a team with enough depth to overcome an injury, a slump, et cetera.
  16. God. Is this about balance again? YES. Exactly! You need to have more than one option for really important things in life. Seriously. It is a thing.
  17. I think you are worried about nothing. Said the Oilers fan.
  18. Oilers appear serious about finding real defensemen. Larsson helps for sure, and we can also light candles for good health. Every little bit helps.
  19. Are you seriously worried about Talbot? I like the goalie, but felt the same way about Dubnyk and Scrivens at the same time each summer. I tell you what: I will stop worrying if we are talking about Talbot as the starter again this time next year. If we are talking about another goalie, I reserve the right to worry till the good lord calls me home.
  20. Fine, but I think you are making to much of it. Hope so, because if this team gets off to another poor start and Edmonton spends yet one more winter mired deep in the second division, the goaling and lack of foresight in problem solving are going to be a daily discussion here.
  21. The schedule is better. Yes, actually that is a good point. Oilers may be able to put a fresh team and goalie on the ice more this coming year.
  22. Why this song? It is about chemistry, love at first sight and relationships that feel comfortable from the beginning. That is Talbot and the Oilers. Hope we see a happier ending than the song. Expect we will.

written by

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

141 Responses to "RE 15-16 CAM TALBOT: PHOTOGRAPH"

  1. Stelio Kontos says:

    Three headed monster? Might shoot myself if we go back to those days.

  2. striatic says:

    One of the good things about the Larsson trade is that as a shutdown guy he’s pretty strong and consistent. He’s out there a large enough portion of the game that it should make the goaltending a little bit easier to analyse. This is different than the Dubnyk/Scrivens situations where bad defence likely lead to variability of performance.

    We know Talbot did ok behind a bad defence, let’s see what he can do behind a better one.

  3. Bruce McCurdy says:

    Stelio Kontos:
    Three headed monster? Might shoot myself if we go back to those days.

    My favourite was the time we had Huey Dewey & Louie in the 2005-06 campaign. MacT used to bring Mike Morrison to do the shootout. Oilers were first in the league in fewest shots allowed and were somehow on the outside of the playoffs looking in until Lowe traded for Roli at the deadline.

    Such a weird group: Jussi Markkanen led the team with 15 wins, Mike Morrison had 10 (a bunch of them in the shootout), and Ty Conklin & Dwayne Roloson won 8 each. But once the three-headed monster was slain and a new champion declared, that crazy season very nearly had a happy ending.

  4. LadiesloveSmid says:

    interested to see how Ellis does in Bakersfield. Not a huge track record in college, but had a great year. The guy he was backing up for 2 seasons before had a pretty seamless if brief transition to the AHL.

  5. Bruce McCurdy says:

    Then there was the three-headed monster of Jacques Plante, Chris Worthy and Ken Brown that played for the 1974-75 WHA Oilers in the premier season of the Coliseum. You want to talk about an equal split of the workload: Brown led the team in games played, Worthy in minutes played, and Plante in decisions.

    It’s so close to mathematically impossible that I have never forgotten that weird congruence.

  6. G Money says:

    Sorry LT, will continue the threadjacks for one more thread before I head back to the woodwork. (Get it)

    More defenders, by request (will run continue to run these today, so smoke ’em if you got ’em). Some patterns starting to emerge. Top defenders typically get 50% of their time against top comp, second pairing about 45%, and then a steep dropoff to the third pairing, which can be anywhere from 20 to 30% max.

    Below, the ‘holy shit’ factor is on Lindholm (he’s good) and Hjalmarsson (who you can see by the splits clearly partners with Keith and plays ridiculously difficult comp).

    Demers is playing a fair bit of top comp but looks more like a really good second pair guy.

    Hamonic is a competent top pair guy, though his DFF numbers look strange – some of it is that he’s less good at defending the Rickibox as others have noted, but some if it might be a NYI rink bias effect, I will have to check my rink bias corrections to make sure they are properly embedded.

    Faulk looks better than I thought he would, Barrie looks worse.

    **** Hamonic ****
    Top Comp TOI 40442 49% CF (607, 620) 49.5% DFF (437.0, 436.0) 50.1%
    Muddle TOI 25784 31% CF (454, 438) 50.9% DFF (333.0, 346.0) 49.0%
    Dregs TOI 16395 20% CF (269, 235) 53.4% DFF (177.0, 203.0) 46.6%

    **** Lindholm ****
    Top Comp TOI 39312 49% CF (615, 541) 53.2% DFF (435.0, 380.0) 53.4%
    Muddle TOI 23396 29% CF (406, 280) 59.2% DFF (300.0, 185.0) 61.9%
    Dregs TOI 16831 21% CF (318, 184) 63.3% DFF (238.0, 128.0) 65.1%

    **** Demers ****
    Top Comp TOI 29875 51% CF (663, 759) 46.6% DFF (471.0, 544.0) 46.4%
    Muddle TOI 17786 30% CF (522, 426) 55.1% DFF (363.0, 272.0) 57.1%
    Dregs TOI 11471 19% CF (406, 259) 61.1% DFF (292.0, 185.0) 61.3%

    **** Barrie ****
    Top Comp TOI 39072 46% CF (537, 711) 43.0% DFF (401.0, 502.0) 44.4%
    Muddle TOI 29550 35% CF (445, 568) 43.9% DFF (308.0, 378.0) 44.9%
    Dregs TOI 16758 20% CF (269, 280) 49.0% DFF (173.0, 186.0) 48.1%

    **** Faulk ****
    Top Comp TOI 39095 58% CF (581, 599) 49.2% DFF (401.0, 514.0) 43.8%
    Muddle TOI 18717 28% CF (322, 267) 54.7% DFF (240.0, 220.0) 52.1%
    Dregs TOI 9610 14% CF (211, 151) 58.3% DFF (172.0, 125.0) 57.9%

    **** Z. Redmond ****
    Top Comp TOI 5649 22% CF (94, 117) 44.5% DFF (79.0, 91.0) 46.3%
    Muddle TOI 9590 38% CF (131, 163) 44.6% DFF (81.0, 104.0) 43.7%
    Dregs TOI 10145 40% CF (122, 122) 50.0% DFF (79.0, 74.0) 51.7%

    **** Hjalmarsson ****
    Top Comp TOI 56094 63% CF (767, 784) 49.5% DFF (574.0, 539.0) 51.6%
    Muddle TOI 21859 25% CF (375, 322) 53.8% DFF (272.0, 267.0) 50.4%
    Dregs TOI 10415 12% CF (202, 137) 59.6% DFF (153.0, 109.0) 58.3%

  7. Lloyd B. says:

    G Do you have yours ears up? I suspect this is going to be painful could you please run Petry through Taylor Haul 9000. In aggregate is great but if you could break out his Oiler and Montreal days that would be much appreciated. Understand if you are locked in the programming cubby.

    On a side note I found out today that I am being promoted at work to supervisor in charge of the lumber department. May need to change my name. Thinking of woody. Or woodguy lite. Hopefully it will be as lucrative for me as it has been for our esteemed colleague.

  8. G Money says:

    Lloyd B.,

    Petry … I haven’t built the ability to do splits by team as yet (it’s in the database, but I haven’t built it into the query system).

    So 2014 is aggregated for the two teams.

    But yes, it is painful. The 2014 numbers are poor, as I expect all of the Oiler numbers are (this will clarify once I add the Rel numbers). (I’m also not sure of the validity of using the measures for 2014, suspect ideally that should consist of using 2013/2014 numbers to generate the QoC ratings list).

    But the 2015 MTL numbers suggest a guy playing second/first pairing, but his results make him a perfectly capable top pair RH D man.

    And we spent years developing him and then gave him away for nothing. And the guy who did it still works for the Oilers in a hockey related capacity.

    Fuck.

    **** Petry 2015 ****
    Top Comp TOI 22783 46% CF (348, 328) 51.5% DFF (241.0, 214.0) 53.0%
    Muddle TOI 17654 36% CF (309, 225) 57.9% DFF (194.0, 186.0) 51.0%
    Dregs TOI 9246 19% CF (168, 130) 56.4% DFF (122.0, 88.0) 58.2%

    **** Petry 2014 ****
    Top Comp TOI 31409 39% CF (446, 609) 42.3% DFF (304.0, 412.0) 42.5%
    Muddle TOI 32085 40% CF (550, 570) 49.1% DFF (337.0, 417.0) 44.7%
    Dregs TOI 17603 22% CF (324, 303) 51.7% DFF (196.0, 202.0) 49.2%

  9. striatic says:

    G Money,

    Enlightening stats as always.

    Lindholm is a beast. I’m curious about Josh Manson, who shows well on the Vollman Player Usage Chart for the Anaheim D when compared with Lindholm. What’s the story there?

  10. striatic says:

    G Money:
    And we spent years developing him and then gave him away for nothing.And the guy who did it still works for the Oilers in a hockey related capacity.

    The worst part is that his ability was almost universally recognized by this blog and many others. It wasn’t some secret.

    This wasn’t some surprise when he did well in MTL, or some “told you so” moment for a handful of niche Petry supporters. Lots of people could see this unfolding from miles away, except the Edmonton Oilers.

  11. G Money says:

    striatic,

    Good question. They are partnered together so the numbers I suspect will be very similar (they are). Teasing out the effect of those two requires digging in to the WOWYs.

    I don’t have those yet (not for a while), but if we look at the standard WOWYs, the two together are a lights out 57%, apart Lindholm is 55% (primary partner Bieksa) and Manson is 50% (primary partner Fowler).

    So I’d have to say I agree with the thought that Lindholm zooms Manson, not the other way around.

    **** Manson ****
    Top Comp TOI 33047 50% CF (487, 430) 53.1% DFF (349.0, 297.0) 54.0%
    Muddle TOI 20221 30% CF (333, 247) 57.4% DFF (254.0, 182.0) 58.3%
    Dregs TOI 13164 20% CF (247, 134) 64.8% DFF (189.0, 100.0) 65.3%

  12. blainer says:

    G Money:
    Sorry LT, will continue the threadjacks for one more thread before I head back to the woodwork.(Get it)

    More defenders, by request (will run continue to run these today, so smoke ’em if you got ’em).Some patterns starting to emerge.Top defenders typically get 50% of their time against top comp, second pairing about 45%, and then a steep dropoff to the third pairing, which can be anywhere from 20 to 30% max.

    Below, the ‘holy shit’ factor is on Lindholm (he’s good) and Hjalmarsson (who you can see by the splits clearly partners with Keith and plays ridiculously difficult comp).

    Demers is playing a fair bit of top comp but looks more like a really good second pair guy.

    Hamonic is a competent top pair guy, though his DFF numbers look strange – some of it is that he’s less good at defending the Rickibox as others have noted, but some if it might be a NYI rink bias effect, I will have to check my rink bias corrections to make sure they are properly embedded.

    Faulk looks better than I thought he would, Barrie looks worse.

    **** Hamonic ****
    Top Comp TOI 40442 49% CF (607, 620) 49.5% DFF (437.0, 436.0) 50.1%
    Muddle TOI 25784 31% CF (454, 438) 50.9% DFF (333.0, 346.0) 49.0%
    Dregs TOI 16395 20% CF (269, 235) 53.4% DFF (177.0, 203.0) 46.6%

    **** Lindholm ****
    Top Comp TOI 39312 49% CF (615, 541) 53.2% DFF (435.0, 380.0) 53.4%
    Muddle TOI 23396 29% CF (406, 280) 59.2% DFF (300.0, 185.0) 61.9%
    Dregs TOI 16831 21% CF (318, 184) 63.3% DFF (238.0, 128.0) 65.1%

    **** Demers ****
    Top Comp TOI 29875 51% CF (663, 759) 46.6% DFF (471.0, 544.0) 46.4%
    Muddle TOI 17786 30% CF (522, 426) 55.1% DFF (363.0, 272.0) 57.1%
    Dregs TOI 11471 19% CF (406, 259) 61.1% DFF (292.0, 185.0) 61.3%

    **** Barrie ****
    Top Comp TOI 39072 46% CF (537, 711) 43.0% DFF (401.0, 502.0) 44.4%
    Muddle TOI 29550 35% CF (445, 568) 43.9% DFF (308.0, 378.0) 44.9%
    Dregs TOI 16758 20% CF (269, 280) 49.0% DFF (173.0, 186.0) 48.1%

    **** Faulk ****
    Top Comp TOI 39095 58% CF (581, 599) 49.2% DFF (401.0, 514.0) 43.8%
    Muddle TOI 18717 28% CF (322, 267) 54.7% DFF (240.0, 220.0) 52.1%
    Dregs TOI 9610 14% CF (211, 151) 58.3% DFF (172.0, 125.0) 57.9%

    **** Z. Redmond ****
    Top Comp TOI 5649 22% CF (94, 117) 44.5% DFF (79.0, 91.0) 46.3%
    Muddle TOI 9590 38% CF (131, 163) 44.6% DFF (81.0, 104.0) 43.7%
    Dregs TOI 10145 40% CF (122, 122) 50.0% DFF (79.0, 74.0) 51.7%

    **** Hjalmarsson ****
    Top Comp TOI 56094 63% CF (767, 784) 49.5% DFF (574.0, 539.0) 51.6%
    Muddle TOI 21859 25% CF (375, 322) 53.8% DFF (272.0, 267.0) 50.4%
    Dregs TOI 10415 12% CF (202, 137) 59.6% DFF (153.0, 109.0) 58.3%

    Jeebus G awesome work by you and WG. Been very busy with the in laws in town from Kelowna but sneaking a peek every now and again.

    Did I miss Larsson’s numbers ?

    I can only imagine the time you guys have put into this.. very impressive indeed. just great work.

  13. haters says:

    Quick LT, post another article before this takes off in dramatic effect like the last one!!

    Man it was fun watching Riki Wg And G argue. Imagine if they put their egos aside and worked together! The Woodguy/Rikibear/Gmoney blog !! Where dumb fans come to feel dumber !!

    I like how WG was comparing himself to Einstien and junk yesterday after my harmless comment..

    I read everything Oilers and appreciate everything I read.., well staples is kinda nuts but almost everything is good. The problem I have with blogs set on a mostly analytical view point is I’ve always said hockey has way to many variables to get a crystal clear vision on a player via analytics. Context and actually watching the players is a must. That’s why LT’s site is my favorite because it blends both worlds so well.

  14. LadiesloveSmid says:

    G Money,

    Can you run any of Vatanen, Hamilton, Trouba, Jones, and Ristolainen?

    edit: Petrovic as well

  15. striatic says:

    G Money: I don’t have those yet (not for a while), but if we look at the standard WOWYs, the two together are a lights out 57%, apart Lindholm is 55% (primary partner Bieksa) and Manson is 50% (primary partner Fowler).

    That 50% for Manson, at first glance anyway, is interesting. Fowler is not great by these metrics and would be a drag.

    Manson might be getting a boost from Lindholm but Lindholm does drop without Manson, and I think Manson’s big drop off might be related to Fowler as much as it has to do with not having Lindholm.

    Anyway, Manson is an interesting player and worth keeping an eye on I think.

    Dammit Anaheim. We need you to be bad, not good.

  16. Lloyd B. says:

    G Money,

    Thanks G. Those numbers are exactly what I was afraid of. MacT wouldn’t keep Petry because “you can’t have all your defencemen making $4Million”. Then promptly traded away our true 1st comp RHD and gave Shultz his money. That MacT is likely still influencing decisions is truly frightful.

    One hope is that when Chiarelli sees these numbers tonight he plugs his ears.

    Thanks to you and Woodguy for the work and commitment to your project. I don’t have first clue about running any numbers but greatly appreciate the work you and others do here for our edification.

    P.S. Ignore the bear. He doesn’t have a cave to hibernate in. 🙂

  17. blainer says:

    Lt one more quick comment on the goaltending.

    I am also not sold on the tending either. I will say this I do feel a little better this year as long as Talbot stays healthy.

    IMO LB took a step back with his cup of coffee at the end of the year. Now I do not agree with Asia much but he was bang on with his point about not counting on LB when the wheels came off both goalers last year.

    Count me in as worried about the goalers and really hope that LB or whoever plays well in the A can bring it to the big league in case of injury.

    As I say EVERY year the goaltending is the most important position in the game and can make or break your season. Calgary experienced it both ways last year and the year before.

  18. godot10 says:

    Gmoney…

    Andy Greene, please. To compare to Larsson.

  19. Lowetide says:

    G: Thanks for using sorry properly, appreciated! 🙂 And don’t be, I am enjoying the information very much, along with the others.

  20. OF17 says:

    G, are you somewhat worried that separating players into bands ignores relative quality of team? Top opposition isn’t the same quality across the league, and bringing RelCorsi into what makes a player top opposition means good players on bad teams are overvalued and good players on good teams are undervalued.

    I’m a bit worried by the “middling” classification as well. Most teams have a significant drop-off between the quality of their 2nd and 3rd lines, which would imply that it’s harder to drive CF% against 2nd line opposition than it is against 3rd line opposition. As I understand it, the middling grouping doesn’t distinguish between the two.

    Your work is clearly a step forward for QC metrics, but I’m just curious to what extent these queries have been factored in.

  21. TeeVee says:

    Def Leppard?

    *edit* NM, this is 15/16

  22. Woodguy says:

    G Money,

    Thanks G!

    **** Petry 2015 ****
    Top Comp TOI 22783 46% CF (348, 328) 51.5% DFF (241.0, 214.0) 53.0%
    Muddle TOI 17654 36% CF (309, 225) 57.9% DFF (194.0, 186.0) 51.0%
    Dregs TOI 9246 19% CF (168, 130) 56.4% DFF (122.0, 88.0) 58.2%

    **** Petry 2014 ****
    Top Comp TOI 31409 39% CF (446, 609) 42.3% DFF (304.0, 412.0) 42.5%
    Muddle TOI 32085 40% CF (550, 570) 49.1% DFF (337.0, 417.0) 44.7%
    Dregs TOI 17603 22% CF (324, 303) 51.7% DFF (196.0, 202.0) 49.2%

    This is a great object lesson for everyone about why Relative (Rel) stats are best.

    Petry looks meh-poor in Edmonton, and then very good in MTL using the raw data.

    The real truth lies in these numbers, but Rel and then WOWY to suss out which other players are driving the other metrics (i.e. When Petry played with Hall the CF was great, with Lander, not so much)

  23. Woodguy says:

    G Money,

    **** Hjalmarsson ****
    Top Comp TOI 56094 63% CF (767, 784) 49.5% DFF (574.0, 539.0) 51.6%
    Muddle TOI 21859 25% CF (375, 322) 53.8% DFF (272.0, 267.0) 50.4%
    Dregs TOI 10415 12% CF (202, 137) 59.6% DFF (153.0, 109.0) 58.3%

    Here’s something interesting to know.

    I’m working on my “is Larsson being zoomed by Greene? post.

    There are two other RHD in the NHL who:

    1) Had very, very tough zone starts (although Larsson were the toughest, by far)
    2) Played mostly tough comp
    3) Managed to still lower the RelCA/60
    4) Managed to have a + Relative xGF%

    Larsson, Hjarlsmasson and Tanev.

    Both are probably good comps for Larsson and I think that Dreamy-Larsson might be a top pair that could be as effective as Keith-Hjarlm very soon.

    All for $8.2MM combined for the next 5 years.

  24. Woodguy says:

    G Money,

    **** Barrie ****
    Top Comp TOI 39072 46% CF (537, 711) 43.0% DFF (401.0, 502.0) 44.4%
    Muddle TOI 29550 35% CF (445, 568) 43.9% DFF (308.0, 378.0) 44.9%
    Dregs TOI 16758 20% CF (269, 280) 49.0% DFF (173.0, 186.0) 48.1%

    This is much better than looks.

    Remember that COL is a 44.2% CF team (therefore vs ALL comp) and a 43.08% xGF team (correlates closely to DFF)

  25. Woodguy says:

    G Money,

    **** Faulk ****
    Top Comp TOI 39095 58% CF (581, 599) 49.2% DFF (401.0, 514.0) 43.8%
    Muddle TOI 18717 28% CF (322, 267) 54.7% DFF (240.0, 220.0) 52.1%
    Dregs TOI 9610 14% CF (211, 151) 58.3% DFF (172.0, 125.0) 57.9%

    Faulks CF% being 49.2 and DFF being 43.8% is something I saw throughout the data on him.

    Good possession numbers, but his box protection would anger every bear.

  26. Woodguy says:

    G Money,

    **** Hamonic ****
    Top Comp TOI 40442 49% CF (607, 620) 49.5% DFF (437.0, 436.0) 50.1%
    Muddle TOI 25784 31% CF (454, 438) 50.9% DFF (333.0, 346.0) 49.0%
    Dregs TOI 16395 20% CF (269, 235) 53.4% DFF (177.0, 203.0) 46.6%

    This exceeds what I saw in terms of xCF/60 and xGA/60

    Strange that he falls off the pace against the lower comps.

    Nice player vs toughs.

    I wonder if their vaunted but really shitty 4th line had anything to do with that?

  27. Woodguy says:

    G Money,

    **** Demers ****
    Top Comp TOI 29875 51% CF (663, 759) 46.6% DFF (471.0, 544.0) 46.4%
    Muddle TOI 17786 30% CF (522, 426) 55.1% DFF (363.0, 272.0) 57.1%
    Dregs TOI 11471 19% CF (406, 259) 61.1% DFF (292.0, 185.0) 61.3%

    This was my biggest fear with Demers.

    I bet that 46.4% DFF vs toughs is also a negative Rel.

    Nice player, but not quite what was needed I think.

    Similar to Sekera, but RH

  28. Woodguy says:

    G Money,

    **** Z. Redmond ****
    Top Comp TOI 5649 22% CF (94, 117) 44.5% DFF (79.0, 91.0) 46.3%
    Muddle TOI 9590 38% CF (131, 163) 44.6% DFF (81.0, 104.0) 43.7%
    Dregs TOI 10145 40% CF (122, 122) 50.0% DFF (79.0, 74.0) 51.7%

    Given what his Rel would be, this is really sexy for a 3rd pairing guy.

    2 years under $1MM/yr and can play power play.

    If Mustache Pete signs him for 3RD there isn’t the urgency for Barrie to fill the 1RDPP spot.

    Opportunity missed.

  29. LadiesloveSmid says:

    Woodguy,

    Petry had to play a ton with Ference in 2014 too, about half his ES ice time. Full 3% CF better without him. Marincin-Petry was a quality pairing coming out of 13-14, had good numbers in a short timeframe the next year but MacT/Eakins had fallen out of love with both.

  30. Drew says:

    haters:
    Quick LT, post another article before this takes off in dramatic effect like the last one!!

    Man it was fun watching Riki Wg And G argue. Imagine if they put their egos aside and worked together! The Woodguy/Rikibear/Gmoney blog !! Where dumb fans come to feel dumber !!

    I like how WG was comparing himself to Einstien and junk yesterday after my harmless comment..

    I read everything Oilers and appreciate everything I read.., well staples is kinda nuts but almost everything is good. The problem I have with blogs set on a mostly analytical view point is I’ve always said hockey has way to many variables to get a crystal clear vision on a player via analytics. Context and actually watching the players is a must. That’s why LT’s site is my favorite because it blends both worlds so well.

    I am by no means a long time commenter here but I have been a Lowetide acolyte since his days on Hockeyfutures. There is no cooperation with the Bear, they are generally solitary and cranky when around others

    Oh, as well the tone in some of your comments can be interpreted as being very negative and demeaning. if you are trying to be light hearted you might want to note it somehow. (trying to be positive not attacking)

  31. Drew says:

    Woodguy:
    G Money,

    Thanks G!

    **** Petry 2015 ****
    Top Comp TOI 22783 46% CF (348, 328) 51.5% DFF (241.0, 214.0) 53.0%
    Muddle TOI 17654 36% CF (309, 225) 57.9% DFF (194.0, 186.0) 51.0%
    Dregs TOI 9246 19% CF (168, 130) 56.4% DFF (122.0, 88.0) 58.2%


    **** Petry 2014 ****
    Top Comp TOI 31409 39% CF (446, 609) 42.3% DFF (304.0, 412.0) 42.5%
    Muddle TOI 32085 40% CF (550, 570) 49.1% DFF (337.0, 417.0) 44.7%
    Dregs TOI 17603 22% CF (324, 303) 51.7% DFF (196.0, 202.0) 49.2%

    This is a great object lesson for everyone about why Relative (Rel) stats are best.

    Petry looks meh-poor in Edmonton, and then very good in MTL using the raw data.

    The real truth lies in these numbers, but Rel and then WOWY to suss out which other players are driving the other metrics (i.e. When Petry played with Hall the CF was great, with Lander, not so much)

    So this fits the get good keep good players mantra and the let the bad players go. Who knew???

    Great stuff guys, i am so not worthy!!!

  32. Woodguy says:

    haters,

    I like how WG was comparing himself to Einstien and junk yesterday after my harmless comment..

    4 things:

    1) It wasn’t a harmless comment. It was quite derisive.

    2) I didn’t compare myself to Einstein. I brought up two well know works that are very descriptive of a certain subject and then inserted your flippant remarks to show how derisive you were.

    The authors of those works were Stephen Hawking and Edward Gibbon. While Einstein and Hawking were both physicists, they were not the same person.

    3) The argument with RTB was not a clash of egos but one of ideas, origins and who owns public information.

    I can’t wait for your next derisive post that mis-describes something new.

  33. Genjutsu says:

    Woodguy:
    G Money,

    **** Hjalmarsson ****
    Top Comp TOI 56094 63% CF (767, 784) 49.5% DFF (574.0, 539.0) 51.6%
    Muddle TOI 21859 25% CF (375, 322) 53.8% DFF (272.0, 267.0) 50.4%
    Dregs TOI 10415 12% CF (202, 137) 59.6% DFF (153.0, 109.0) 58.3%

    Here’s something interesting to know.

    I’m working on my “is Larsson being zoomed by Greene?post.

    There are two other RHD in the NHL who:

    1) Had very, very tough zone starts (although Larsson were the toughest, by far)
    2) Played mostly tough comp
    3) Managed to still lower the RelCA/60
    4) Managed to have a + Relative xGF%

    Larsson, Hjarlsmasson and Tanev.

    Both are probably good comps for Larsson and I think that Dreamy-Larsson might be a top pair that could be as effective as Keith-Hjarlm very soon.

    All for $8.2MM combined for the next 5 years.

    As much as the Hall trade was a kick in the nuts, and it really really was, if this comes to pass I’ll be OK with it.

    If this happens we should be looking at playoffs and cup(s)?

  34. frjohnk says:

    Adam Larsson:

    Top Comp TOI 46577 CF 42.3 DFF 44.5
    Muddle TOI 24173 CF 45.8 DFF 53.2
    Dregs TOI 14914 CF 48.9 DFF 51.8

    **** Eric Gryba ****
    Top Comp TOI 20108 41% CF (306, 340) 47.4% DFF (217.0, 235.0) 48.0%
    Muddle TOI 16326 34% CF (246, 248) 49.8% DFF (161.0, 193.0) 45.5%
    Dregs TOI 12256 25% CF (187, 168) 52.7% DFF (118.0, 116.0) 50.5%

    Gryba looks pretty decent while Larsson’s numbers are raising the pucker factor for me.
    But team effects certainly need to be considered for Larsson as well.

  35. Woodguy says:

    LadiesloveSmid:
    Woodguy,

    Petry had to play a ton with Ference in 2014 too, about half his ES ice time. Full 3% CF better without him. Marincin-Petry was a quality pairing coming out of 13-14, had good numbers in a short timeframe the next year but MacT/Eakins had fallen out of love with both.

    Excellent point.

    WOWYs and Rel’s are needed to flesh out the truth.

  36. stephen sheps says:

    G Money,

    Since you’re actually in the thread today – sent you an email…

  37. Genjutsu says:

    Woodguy:
    haters,

    I like how WG was comparing himself to Einstien and junk yesterday after my harmless comment..

    4 things:

    1) It wasn’t a harmless comment.It was quite derisive.

    2) I didn’t compare myself to Einstein.I brought up two well know worksthat are very descriptive of a certain subject and then inserted your flippant remarks to show how derisive you were.

    The authors of those works were Stephen Hawking and Edward Gibbon.While Einstein and Hawking were both physicists, they were not the same person.

    3) The argument with RTB was not a clash of egos but one of ideas, origins and who owns public information.

    I can’t wait for your next derisive post that mis-describes something new.

    Yeah I didn’t really get his comment in the first place.

    On be half of all hockey fans: thanks to you and G and Mr Bear for all the hard work you’ve put forth.

    It’s helped me understand more and I’d really appreciate it if y’all keep doing it.

  38. Lowetide says:

    Woodguy:
    haters,

    I like how WG was comparing himself to Einstien and junk yesterday after my harmless comment..

    4 things:

    1) It wasn’t a harmless comment.It was quite derisive.

    2) I didn’t compare myself to Einstein.I brought up two well know worksthat are very descriptive of a certain subject and then inserted your flippant remarks to show how derisive you were.

    The authors of those works were Stephen Hawking and Edward Gibbon.While Einstein and Hawking were both physicists, they were not the same person.

    3) The argument with RTB was not a clash of egos but one of ideas, origins and who owns public information.

    I can’t wait for your next derisive post that mis-describes something new.

    Phil Einstein? Great guy. Poor sense of humor but the man can juggle.

  39. Woodguy says:

    frjohnk,

    Really impressed with Gryba. If Mongo Taylor Haul 9000 was fully functional I’d dive into that one deep to see if it was real.

    Adam Larsson:

    Top Comp TOI 46577 CF 42.3 DFF 44.5
    Muddle TOI 24173 CF 45.8 DFF 53.2
    Dregs TOI 14914 CF 48.9 DFF 51.8

    NJD as a team: CF% 46.12 (all comps) xGF% 47.77%

    Remember that Larsson has the toughest DZFO% in the NHL at 44.4% (Greene next at 43.61) next toughest is Beauchemin at 37.66.

    Tanev was at 34.4 and Hjarlm at 34.5

    When I examined Larsson he was where CF% went to die, but the CA/60 and xGA/60 was amazing considering the ZS, comp and how bad Greene was away from him this year (there are partner issues there, but Greene didn’t overcome them)

    Not enough to make you unpucker, but the value of Larsson lies on one side of the puck and these metrics are showing both.

    I would feel better if that DFF was a couple points higher though.

  40. Woodguy says:

    Lowetide: Phil Einstein? Great guy. Poor sense of humor but the man can juggle.

    Makes a hell of a bagel though.

  41. Woodguy says:

    JDï™:
    Woodguy,

    https://www.youtube.com/watch?v=mGsC_LO3oFY

    Love that show.

    Is this in reference to RTB?

  42. frjohnk says:

    If I understand it right, a teams 3rd pairing should be able to handle the “dregs” and “muddle” well/OK and should be limited to facing the “top comp” as they will most likely get their heads kicked in. ( I like DFF% more than CF%, but still acknowledge CF% somewhat )

    We have these 4 guys who fit the bill for the 3rd pairing
    **** Darnell Nurse ****
    Top Comp TOI 30354 43% CF (386, 551) 41.2% DFF (280.0, 393.0) 41.6%
    Muddle TOI 25345 36% CF (373, 399) 48.3% DFF (262.0, 274.0) 48.9%
    Dregs TOI 15341 22% CF (258, 247) 51.1% DFF (175.0, 186.0) 48.4%

    **** Jordan Oesterle ****
    Top Comp TOI 6900 38% CF (93, 120) 43.7% DFF (78.0, 109.0) 41.8%
    Muddle TOI 7476 42% CF (106, 103) 50.7% DFF (83.0, 75.0) 52.6%
    Dregs TOI 3626 20% CF (46, 44) 51.1% DFF (38.0, 35.0) 51.8%

    **** Mark Fayne ****
    Top Comp TOI 29172 49% CF (674, 906) 42.7% DFF (474.0, 629.0) 43.0%
    Muddle TOI 18599 31% CF (586, 570) 50.7% DFF (382.0, 394.0) 49.2%
    Dregs TOI 11696 20% CF (399, 336) 54.3% DFF (275.0, 247.0) 52.7%

    **** Griffin Reinhart ****
    Top Comp TOI 10622 39% CF (176, 251) 41.2% DFF (142.0, 179.0) 44.3%
    Muddle TOI 10663 39% CF (175, 169) 50.9% DFF (135.0, 122.0) 52.5%
    Dregs TOI 6201 23% CF (92, 83) 52.6% DFF (79.0, 53.0) 60.0%

    That leaves 4D to fill the top 2 pairings

    **** Andrej Sekera ****
    Top Comp TOI 39771 49% CF (924, 1108) 45.5% DFF (649.0, 767.0) 45.8%
    Muddle TOI 25021 31% CF (728, 741) 49.6% DFF (505.0, 530.0) 48.8%
    Dregs TOI 15739 20% CF (504, 407) 55.3% DFF (360.0, 285.0) 55.9%

    **** Oscar Klefbom ****
    Top Comp TOI 15378 53% CF (215, 245) 46.7% DFF (163.0, 176.0) 48.1%
    Muddle TOI 9399 32% CF (144, 124) 53.7% DFF (105.0, 84.0) 55.4%
    Dregs TOI 4422 15% CF (77, 58) 57.0% DFF (40.0, 34.0) 53.4%

    **** Brandon Davidson ****
    Top Comp TOI 20277 43% CF (307, 309) 49.8% DFF (196.0, 202.0) 49.2%
    Muddle TOI 17487 37% CF (259, 261) 49.8% DFF (174.0, 171.0) 50.4%
    Dregs TOI 9600 20% CF (158, 100) 61.2% DFF (100.0, 65.0) 60.7%

    *****Adam Larsson*****
    Top Comp TOI 46577 CF 42.3 DFF 44.5
    Muddle TOI 24173 CF 45.8 DFF 53.2
    Dregs TOI 14914 CF 48.9 DFF 51.8

    Top pairing Dmen should be able to handle “top comp” OK, do well against “muddle” and destroy the “dregs”
    Davidson and Klefbom are fitting the bill with those numbers, but the sample size is not huge.

    Sekera and Larsson seem to show that they are more apt to handle “muddle” and the “dregs” than they can “top comp” but I also believe New Jersey’s system, Larsson’s deployment are hindering Larsson’s numbers.

    Overall, Im pretty pumped by these numbers.

    By eye, I had felt Nurse had done OK when not going against the toughs and tried to measure it with numbers but was at a loss on trying to prove my point as I couldn’t find numbers that matched what I saw. These numbers help big time.

    I had felt Fayne was not a top 4 Dman, these numbers suggest that.

    I have held the believe that Davidson is not unlike Brodie on how he handles the puck and decision making. Could possibly become “Brodie-lite”. Figured Davidson could become a top 3 Dman by November if he took a step forward and brought more offense. His numbers from last year would suggest close to a number 3. Still small sample size but Davidson is looking like a gem. Love this guy.

    Klefbom. Are we surprised? He stays healthy and takes a step forward, are we looking at a Dman entering close into the realm of a number 1?

  43. Drew says:

    Lowetide: Phil Einstein? Great guy. Poor sense of humor but the man can juggle.

    Einstein’s mother said, “take the rest of the day off”
    you’re welcome if your see what i did there.

  44. Water Fire says:

    Woodguy:
    frjohnk,

    Really impressed with Gryba.If Mongo Taylor Haul was fully functional I’d dive into that one deep to see if it was real.

    Adam Larsson:


    Top Comp TOI 46577 CF 42.3 DFF 44.5
    Muddle TOI 24173 CF 45.8 DFF 53.2
    Dregs TOI 14914 CF 48.9 DFF 51.8

    NJD as a team:CF% 46.12 (all comps)xGF% 47.77%

    Remember that Larsson has the toughest DZFO% in the NHL at 44.4% (Greene next at 43.61) next toughest is Beauchemin at 37.66.

    Tanev was at 34.4 and Hjarlm at 34.5

    When I examined Larsson he was where CF% went to die, but the CA/60 and xGA/60 was amazing considering the ZS, comp and how bad Greene was away from him this year (there are partner issues there, but Greene didn’t overcome them)

    Not enough to make you unpucker, but the value of Larsson lies on one side of the puck and these metrics are showing both.

    You’se guys work has reinforced my thought that strong two way D are the key as opposed to offensive. I don’t think many players can outscore a weak D game. And it seems the best teams have a top D that kills it at both ends

  45. G Money says:

    haters: Imagine if they put their egos aside and worked together!

    Pffft. I’m simply way too smart, good-looking, generous, athletic, and humble to have time for an ego.

  46. Drew says:

    G Money: Pffft.I’m simply way too smart, good-looking, generous, athletic, and humble to have time for an ego.

    i heard that G Money does all this work to compensate for the fact he is terrible in the corners and has poor eye glow/60.

  47. Little Poteet says:

    Woodguy and Gmoney,
    I remember there being discussions here and on Twitter about high event vs low event defenseman. I think the agreed upon definition was lots of shots both ways to be a high event guy. Seeing your data brings up a few questions about event frequency in evaluating players. I’d love to know your thoughts on the following:

    1) does frequency, or amount of shots going both ways matter? Is it better to be a dman who (against any particular level of competition) is on the ice for a) 10 shots for and 8 against or b) 20 for and 16 against in the same number of minutes played, assuming the danger level of the shots that a and b faced is the same? Which guy would you want on your team?

    2) is it possible to know if the number of events per minute played is due to a player at all or is a result of structure dictated by the coach?

    3) does having a high event dman effect the numbers of the forwards they play with vs a low event dman in a meaningful way?

  48. Woodguy says:

    In regards to the Actual Thread:

    Here’s the top 60 goalies who have played at least 1000 min (all situations) over the last 2 years sorted by all situations SV%

    Player Sv%
    CAREY.PRICE 93.36
    ANDREW.HAMMOND 92.74
    CALVIN.PICKARD 92.67
    ANTTI.RAANTA 92.6
    CORY.SCHNEIDER 92.46
    COREY.CRAWFORD 92.4
    DEVAN.DUBNYK 92.33
    BRIAN.ELLIOTT 92.31
    BRADEN.HOLTBY 92.25
    STEVE.MASON 92.24
    SCOTT.DARLING 92.19
    BEN.BISHOP 92.13
    ROBERTO.LUONGO 92.13
    HENRIK.LUNDQVIST 92.09
    MARC-ANDRE.FLEURY 92.05
    CAM.TALBOT 92.04****Oiler
    RETO.BERRA 92.02**** waived last season – traded for Grimaldi this season
    PETR.MRAZEK 91.97
    THOMAS.GREISS 91.97
    JOONAS.KORPISALO 91.95
    TUUKKA.RASK 91.92
    CRAIG.ANDERSON 91.87
    MICHAL.NEUVIRTH 91.85
    CONNOR.HELLEBUYCK 91.8
    JONATHAN.QUICK 91.79
    PHILIPP.GRUBAUER 91.79
    JOHN.GIBSON 91.78
    SEMYON.VARLAMOV 91.75
    JAKE.ALLEN 91.71
    FREDERIK.ANDERSEN 91.63
    JAROSLAV.HALAK 91.57
    MARTIN.JONES 91.57
    PEKKA.RINNE 91.53
    JAMES.REIMER 91.45
    RYAN.MILLER 91.39
    JACOB.MARKSTROM 91.38
    SERGEI.BOBROVSKY 91.38
    ONDREJ.PAVELEC 91.37
    ROBIN.LEHNER 91.37
    ANDREI.VASILEVSKIY 91.33
    LINUS.ULLMARK 91.33
    EDDIE.LACK 91.24
    LOUIS.DOMINGUE 91.16
    CHAD.JOHNSON 91.13*****was UFA
    MICHAEL.HUTCHINSON 91.07
    ANTTI.NIEMI 91.06
    JONATHAN.BERNIER 91.05
    KARRI.RAMO 91.04 ***still unsigned
    CARTER.HUTTON 91.03 **** was UFA
    CAM.WARD 90.98
    DARCY.KUEMPER 90.95
    JONAS.GUSTAVSSON 90.9***new Oiler
    KEITH.KINKAID 90.88
    JIMMY.HOWARD 90.85
    MIKE.SMITH 90.82
    AL.MONTOYA 90.76 **** was UFA
    JHONAS.ENROTH 90.74 ****still unsigned
    CURTIS.MCELHINNEY 90.63
    JONAS.HILLER 90.54

    The day Berra was waived last year (Feb 28) was the day before the trade deadline and Brossoit was on the roster as the back up.

    Perhaps they should have picked him up then? There are questions about his “work ethic” : http://bsndenver.com/avalanche-waive-reto-berra/

    Johnson looks a bit better. Ramo and Hutton look to be a *bit* better, but its so close that it probably within variance of the same number. (edited to pump Johnson’s tires a bit)

    Maybe Pete didn’t do as bad as I thought here?

    5v5 the order is different and we know 5v5 is more repeatable, but I used all situations here as LT did in his main post.

  49. Drew says:

    Little Poteet:
    Woodguy and Gmoney,
    I remember there being discussions here and on Twitter about high event vs low event defenseman. I think the agreed upon definition was lots of shots both ways to be a high event guy. Seeing your data brings up a few questions about event frequency in evaluating players. I’d love to know your thoughts on the following:

    1) does frequency, or amount of shots going both ways matter? Is it better to be a dman who (against any particular level of competition) is on the ice for a) 10 shots for and 8 against or b) 20 for and 16 against in the same number of minutes played, assuming the danger level of the shots that a and b faced is the same? Which guy would you want on your team?

    2) is it possible to know if the number of events per minute played is due to a player at all or is a result of structure dictated by the coach?

    3) does having a high event dman effect the numbers of the forwards they play with vs a low event dman in a meaningful way?

    me just being a dummy would suggest that for 1) the more you out shoot /60 wood be better. All else being equal of course.

  50. Woodguy says:

    frjohnk,

    Klefbom. Are we surprised? He stays healthy and takes a step forward, are we looking at a Dman entering close into the realm of a number 1?

    Absolutely.

    I talked with Gmoney today for 1:15min.

    15 min about the project and the other hour was Gmoney going through a whole laundry list of reasons why the Oilers are actually a threat to make the playoffs this year.

    Klef-Larsson was one of the key reasons.

  51. Woodguy says:

    G Money: Pffft.I’m simply way too smart, good-looking, generous, athletic, and humble to have time for an ego.

    if you’re still around and taking requests:

    Gio
    Brodie
    Hamilton

    Please.

  52. Lloyd B. says:

    G Money,

    I’m stealing this ! Please don’t sue me. I promise I won’t make any money from it.

  53. G Money says:

    OF17,

    My good sir (or maam) this is an excellent set of questions. For the sake of others who might still be getting their heads around shot metrics, please indulge me while I take a moment to review the fundamentals before answering! (I’ll split up the posts so as not to aggravate my walloftextitis).

  54. G Money says:

    Drew: i heard that G Money does all this work to compensate for the fact he is terrible in the corners and has poor eye glow/60.

    Lots of grit though – I keyboard with my elbows high!

  55. frjohnk says:

    Woodguy: I talked with Gmoney today for 1:15min.
    15 min about the project and the other hour was Gmoney going through a whole laundry list of reasons why the Oilers are actually a threat to make the playoffs this year.
    Klef-Larsson was one of the key reasons.

    Im not betting against GMoney.

    Can the season start tomorrow?

  56. LadiesloveSmid says:

    Woodguy,

    in a tough spot where they’ve gotta find a goalie that’s OK with 2 other goalies already on 1-ways. Wouldn’t say Chiarelli made out too poorly in that case. Just need Talbot to be a quality 55-60 game guy this year with no twin baby blips.

  57. JDï™ says:

    Woodguy,

    Yeah, sorry – just getting caught up on last night’s action.

    Great song too.

  58. mustang says:

    Woodguy:
    G Money,

    **** Faulk ****
    Top Comp TOI 39095 58% CF (581, 599) 49.2% DFF (401.0, 514.0) 43.8%
    Muddle TOI 18717 28% CF (322, 267) 54.7% DFF (240.0, 220.0) 52.1%
    Dregs TOI 9610 14% CF (211, 151) 58.3% DFF (172.0, 125.0) 57.9%

    Faulks CF% being 49.2 and DFF being 43.8% is something I saw throughout the data on him.

    Good possession numbers, but his box protection would anger every bear.

    WG and G you two are doing a fabulous job!

    Not only would Mr.Bear be angry, a few other humans/creatures might be as well.
    Net front protection/shot suppression will be paramount with the slimmer dialed down goalie equip.

    Larsson might come in handy

  59. Water Fire says:

    With these TH9 stats, IF Davidson stays the course and is emerging, given those first comp numbers this may be possible and good

    Davidson Larsson
    Sekera Klefbom
    Nurse Fayne

    I like this group especially if a more mobile player with passing skills can take over for Fayne at some point.

  60. Little Poteet says:

    Drew: me just being a dummy would suggest that for 1) the more you out shoot /60 wood be better.

    But you are also giving up more the other way, so I guess the real question is, is a 20ft wrister in the slot “worth” the same if it’s against you as it is for you?
    At first you say, well yes, and if that’s the case then you want the high event guy because you get more absolute number of chances even though relative to your opponent the ratio of chances for and against is the same.
    If a shot is not the same value for as it is against then you become new jersey, because save percentage skews the value of each individual shot. A shot for new jersey is worth more than a shot against because they get so few for and their goalie has a high save percentage. They are ok with playing low event because due to save percentage every shot they get is worth more than one against.
    Is this flawed thinking?

  61. Lloyd B. says:

    Woodguy:
    frjohnk,

    Klefbom. Are we surprised? He stays healthy and takes a step forward, are we looking at a Dman entering close into the realm of a number 1?

    Absolutely.

    I talked with Gmoney today for 1:15min.

    15 min about the project and the other hour was Gmoney going through a whole laundry list of reasons why the Oilers are actually a threat to make the playoffs this year.

    Klef-Larsson was one of the key reasons.

    On the surface the Oilers lost the trade. But..IF.. Klefbom Larrson makes a solid first pairing D and that makes the team better, so much better that the playoffs are a RE this year, does Chiarelli still lose the trade? I’m thinking maybe not. Especially at their combined cap hit.

    This may well become the **Oiler model** that others will follow once we win the cup. Without the ten years of nasty.

  62. G Money says:

    OF17,

    Fundamental 1: All shot metrics show the balance of a five man unit vs a five man unit.

    Some folks misinterpret this to mean that shot metrics can’t be used at an individual level. This unfortunately displays a profound lack of understanding of statistical methods, the purpose of which ultimately is to pull specific information out of large broad datasets. (It’s OK when the person saying that is some doofus posting online, it’s very unfortunate when it’s someone like Staples who has such a large platform to scream his innumeracy).

    In fact, when you look at the ebb and flow of an individual, you find that despite the fact that teams on paper always have four lines and three d pairs, the actual on-ice variation in teammates for an individual player, especially a forward, is quite high.

    That means that as soon as you start getting reasonable sample sizes (call it minimum 20 games), the metric despite being a five man metric starts to reflect individual performance.

    That’s why it is exceedingly rare to have two teammates with exactly the same measure, even if they’re linemates or d partners.

  63. Drew says:

    Little Poteet: But you are also giving up more the other way,so I guess the real question is, is a 20ft wrister in the slot “worth”the same if it’s against you as it is for you?
    At first you say, well yes, and if that’s the case then you want the high event guy because you get more absolute number of chances even though relative to your opponent the ratio of chances for and against is the same.
    If a shot is not the same value for as it is against then you become new jersey, because save percentage skews the value of each individual shot. A shot for new jersey is worth more than a shot against because they get so few for and their goalie has a high save percentage. They are ok with playing low event because due to save percentage every shot they get is worth more than one against.
    Is this flawed thinking?

    i added with all else being equal, 1 is good, 4 is better, 10 is wonderful, being; out shooting/60, out chancing/60, outscoring/60

  64. G Money says:

    OF17,

    Fundamental 2:

    There are two things that ‘break’ this rosy situation: pairs, and quality of opposition.

    By pairs, this obviously fits d pairs, who tend to play a lot together.

    But it also fits forwards, where there tends to be more ‘play’ in one linemate on a line than others, so you tend to see time in pairs stack up much faster than time in three man units. (I track all pairs, lines, and five man units, and you can see very clearly how fast the pairs build up and how slow the lines and five-mans build, even over the course of an entire season).

    Any time you get those pairs happening, they throw off the metrics somewhat, because the numbers for those two individuals do reflect NOT the five man pairing performance, but the performance of the pair. The two players end up with a similar number reflecting an averaged performance.

  65. Little Poteet says:

    Little Poteet:
    Woodguy and Gmoney,
    I remember there being discussions here and on Twitter about high event vs low event defenseman. I think the agreed upon definition was lots of shots both ways to be a high event guy. Seeing your data brings up a few questions about event frequency in evaluating players. I’d love to know your thoughts on the following:

    1) does frequency, or amount of shots going both ways matter? Is it better to be a dman who (against any particular level of competition) is on the ice for a) 10 shots for and 8 against or b) 20 for and 16 against in the same number of minutes played, assuming the danger level of the shots that a and b faced is the same? Which guy would you want on your team?

    2) is it possible to know if the number of events per minute played is due to a player at all or is a result of structure dictated by the coach?

    3) does having a high event dman effect the numbers of the forwards they play with vs a low event dman in a meaningful way?

    And if shot rate is entirely coaching based then can you project a low event guy into a high event system? My concern with this is Fayne

  66. G Money says:

    OF17,

    Fundamental 3:

    We have two tools to deal with this state of affairs.

    For the teammate situation, we use WOWYs – we look at how the two players do apart from each other, and tends to give a better picture. (In fact, when assessing d men, I think it should be considered a mandatory tool).

    There’s lots of work to do even so, because who the other players are when the two players are apart makes a big difference too (as we saw above with the Lindholm/Manson example, or what Woodguy found when he looked at the Jagr effect).

    But at least it is a reasonably effective tool.

  67. Woodguy says:

    Little Poteet,

    1) does frequency, or amount of shots going both ways matter? Is it better to be a dman who (against any particular level of competition) is on the ice for a) 10 shots for and 8 against or b) 20 for and 16 against in the same number of minutes played, assuming the danger level of the shots that a and b faced is the same? Which guy would you want on your team?

    Yes it does matter. I have a post half done for a few months now on this subject.

    Think of it this way.

    2 Dmen on the same team with same ONSH% of 7 and same ONSV% of .910 This team will have a PDO of .980

    Dman 1 has 30 shots for and 30 shots against per 60. 50% shot share.
    Dman 2 has 15 shots for and 15 shots against per 60. 50% shot share.

    So they are identical in every way except frequency and they play on a lower PDO team.

    Dman 1 will get 2.1 goals for per 60 and 2.7 goals against for a -0.6/60 goal share
    Dman 2 will get 1.05 goals for per 60 and 1.35 goals against for a -0.3/60 goal share

    Even though their shot metrics are identical as a ratio and their SH% and SV% are identical, the team will be outscored by twice as many goals when Dman 1 is on the ice due to frequency of events and shitty goalering.

    This is basically what Faulk it, except he effects SV% in a negative way too.

    If the PDO is 100 they are even.

    If the PDO is above 100 then the higher event player will outscore the lower event player.


    2) is it possible to know if the number of events per minute played is due to a player at all or is a result of structure dictated by the coach?

    Its both. All players on a team play for the same coach and you can look at differences in frequencies among players and see real differences.

    That’s why Relative metrics are important. They should how different player do in the same system and mostly with the same players (except Dpartners for the most part)

    A good place to look at this data is stats.hockeyanalysis.com

    Here is a link to Sekera’s page from this year to get you going: http://stats.hockeyanalysis.com/showplayer.php?pid=1724&withagainst=true&season=2015-16&sit=5v5

    This page show goals (GF and GA) and shot attempts (CF and CA) as well as Zone Face offs at the far end


    3) does having a high event dman effect the numbers of the forwards they play with vs a low event dman in a meaningful way?

    Don’t think so. Their TOI will dictate that, not the number of events per minute.

    The most high event Dman is Klingberg. Here’s his page from last year:

    http://stats.hockeyanalysis.com/showplayer.php?pid=2028&withagainst=true&season=2015-16&sit=5v5

    Then find Demers, click on him, then click on 14/15 and compare the two with the same forwards.

  68. Ryan says:

    I think this is great stuff with Woodguy and Gmoney’s collaboration.

    I’ll lay aside any fiduciary claim to my ideas lol… 🙂

    With that proviso, could Gmoney segment the players into 3 buckets based on Zone start? I think it would be really cool to do this analysis to see how each defensemen’s cff% and Dff% varied by offensive, neutral, or dzone start.

    Also,time permitting, how about Danger Fenwick against per sixty based on zone starts?

    Last request, how does Alex Petrovic stack up on the analysis you’re already doing?

  69. Woodguy says:

    Little Poteet: And if shot rate is entirely coaching based then can you project a low event guy into a high event system? My concern with this is Fayne

    The base is coach driven then the players will be layered on top of it.

    Fayne will always be Fayne, even if its “new and fit Fayne” its Fayne.

  70. Woodguy says:

    Lloyd B.: On the surface the Oilers lost the trade. But..IF.. Klefbom Larrson makes a solid first pairing D and that makes the team better, so much better that the playoffs are a RE this year, does Chiarelli still lose the trade?I’m thinking maybe not.Especially at their combined cap hit.

    This may well become the **Oiler model** that others will follow once we win the cup.Without the ten years of nasty.

    The Oilers lost the Hall trade.

    Lost it solidly.

    That’s not a reflection on Larsson’s value, which is good, but of Hall’s which is elite.

  71. G Money says:

    OF17,

    Fundamental 4:

    But oh those QoCs!!

    The only tool for sussing out opposition effects is QoC. Which as I outlined in my introductory post yesterday, is a terrible metric. It’s based on a poor measure of quality, and then it runs all of that poorly measured info together into five man units, which completely obliterates the specific one-player effect you’re trying to find!

    Suppose you have two players, one with a CF% of 49% and one with a CF% of 51%. The first faces a QoC of 0.4 and the second of 0.2. OK, so you should make an allowance for the first guy for facing tougher comp.

    But how much?

    Is that 49% a good number at 0.4? Or is 50% the right number.

    You have a vague idea of the quality of competition, but no idea of what to do with it as far as assessing the differing metrics for the two players.

  72. G Money says:

    OF17,

    Answer 1:

    Which leads us to today, and the work that we’ve been talking through.

    WOWYs are pretty good for splitting out teammate effects, because teammate effects tend to break down into teammate pairs.

    But QoC is blunt saw where you need a scalpel.

    The scalpel is the idea of splitting out the levels of competition using a higher grade measure than CFRel or TOI, standardizing those levels of competition (so that when Player A faces ‘top comp’ it is as closely as possible the same ‘top comp’ that another player faces, EVEN on another team), and then looking at how an individual does against those levels.

    So we’ve taken a massive leap forward (IMAWGHO – in my and WGs humble opinion) in now being able to suss out those damn competition effects.

    Is it perfect?

    Nope. You’ve highlighted some of the challenges.

  73. Lloyd B. says:

    G Money,

    If we agree there are typically 4 lines … to pick up on a previous thought on the thread should you consider 4 buckets instead of 3? Elite, solid ( proven, way up, way down), middle mush( regular NHL players), dregs?

    Perhaps that is why your data is showing more Elite than you would like. A second tier of solid might explain why your numbers are showing 90 elite instead of 60.

    Would this reduce elite players to the 60 you and WG are setting as the right amount? The solid bucket might even reduce the elite to 30. Per team it would work out to 1 elite, 3-5 solid, 13 middling, 6 dregs.

    Or my head could be straight up a dark stinky place.

  74. G Money says:

    OF17: I’m a bit worried by the “middling” classification as well. Most teams have a significant drop-off between the quality of their 2nd and 3rd lines, which would imply that it’s harder to drive CF% against 2nd line opposition than it is against 3rd line opposition. As I understand it, the middling grouping doesn’t distinguish between the two.

    Answer 2:

    So I hope the previous post answered your first question. By standardizing QoC levels so they mean the same thing (as closely as you can considering the natural variations in schedule), ‘Top Comp’ is now directly comparable.

    Your second question is about the ‘muddy middle’, and whether we’re losing resolution by just lumping it into three categories rather than more.

    The first point I’d make is that we always need to be cognizant of splitting the data too finely. The finer the data splits, the smaller the sample sizes, the less valid the conclusions. So there is a tradeoff there.

    The second is that we spent some time looking at how the distributions of the various metrics we use (p/60, CF%Rel, DFA/60, etc) vary across the league by player. Turns out that these things tend to look very Gaussian (“normal”) in large samples.

    That means that we can apply standard statistical thinking and start considering slicing the data using e.g. standard deviation. And that’s kind of where we’re getting to – the Top and Dregs are (sort of) approaching +1SD and -1SD levels for players.

    You can variegate the sample in the middle too (and in fact, my programs all split them into six categories of 16.7% each), but what we find is that the differences in those middle bunch do in fact bunch together.

    So at the moment (I reserve the right to have the data change my mind!) I believe we’re OK in just using the three levels, but hopefully time will tell.

    Questions answered?

  75. kinger_OIL says:

    haters,

    Hater’s says: “Man it was fun watching Riki Wg And G argue. Imagine if they put their egos aside and worked together!”

    – Kinger says: I’m not sure if it was fun, I didn’t like the hostility. It was to me the cools kids in high-school “beating up” the outlier. I read Ricki as being very respectful and engaged, while the cool kids and thier friends making fun of Ricki: I am simplifying to be sure.

    – I’m proud of the passion and the intellect on this blog and being able to be a participant. I feel bad about the bickering over different ways of communicating and sharing information.

  76. G Money says:

    Lloyd B.,

    I think? I answered your question in my reply to OF.

  77. G Money says:

    Ryan,

    I will run all the additional requested d men including Petrovic in a moment.

    Your question on zone starts is an interesting one!

    For most players, the effect is quite small in the overall flow of the game (CF% impact typically < 1%).

    There's also the matter of separating zone starts from true zone starts, where the coach sends a player to start a shift in the o or d zone. Most zone starts are mid-shift, which means the player ‘earned’ the o or d zone faceoff. They are symptom rather than cause.

    That said, one of the players for which this effect is significant is Larsson. Figures. Dammit.

    I do track zone starts, mid and true, so providing a zone start adjustment may be in the cards down the road.

    We might also be able to answer the question as to whether zone starts are more relevant when facing top comp vs dregs. I would think they are – a d zone faceoff against Thornton & co. is vastly more frightening and likely to be damaging than is a d zone faceoff against Tanner Glass.

  78. Woodguy says:

    Lloyd B.:
    G Money,

    If we agree there are typically 4 lines …to pick up on a previous thought on the thread should you consider 4 buckets instead of 3? Elite, solid ( proven, way up, way down), middle mush( regular NHL players), dregs?

    Perhaps that is why your data is showing more Elite than you would like. A second tier of solid might explain why your numbers are showing 90 elite instead of 60.

    Would this reduce elite players to the 60 you and WG are setting as the right amount?The solid bucket might even reduce the elite to 30. Per team it would work out to 1 elite, 3-5 solid, 13 middling, 6 dregs.

    Or my head could be straight up a dark stinky place.

    See Gmoney’s answer right below your post.

    The “muddle middle” is within 1 SD of the mean on either side of the distribution of most metrics.

    Many of those players will be on the positive side in one metric, and then the negative side in another depending on the metric, so they end up looking very much similar to each other.

    Think of it like a big bowl of jello with whipped cream on top (elite players) and non-gelled liquid at the bottom (dregs)

    There really isn’t much difference in the actual jello in the bowl, but the whipped cream and liquid are quite different.

    Everyone is conditioned to think of “4 lines” so four groupings of players, but the results show us that’s not really how talent is distributed in the NHL.

  79. Water Fire says:

    Larsson may be the lynchpin of multiple Oiler cups, but Chiarelli still lost the Hall trade.

    If both players were at the same level of establishment, it would then be a gamble as to who developed more.

    Because Hall is already elite in every way except defensive play, offence has the highest value because it is the most rare skill, and Larsson is not clearly established and has not put points up yet, the Oilers return is well below what everyone expected.

    The argument that the ‘market’ was expensive might hold true if the deal was for a clear point getting #1. But it wasn’t,that line is cover. I believe it is cover for the fact that Hall was being moved no matter what which I say from the verbal from many sources especially Rishaugs response to the GM who said he had a better package for Hall which was that if he didn’t know that Hall was available he had his head in the sand since January.

    The trade was a poor return but the Oilers may still be fine. However if they continue losing trades they will start to lose overall talent and fail to improve. That has to be the last loss for a long while

  80. Woodguy says:

    kinger_OIL,

    read Ricki as being very respectful and engaged

    When someone says they invented stuff you’ve been working on and figuring out on your own its not respectful.

    I agree that RTB was genuinely happy that we are doing this, but it still wasn’t a great way to start the conversation.

    Especially when he tells you to not profit from it and this project is a sinkhole of very valuable time and not nearly as valuable money with zero prospects or designs of monetizing the end product.

    He swung first.

  81. West says:

    Oscar gave us 60 games in 2014-15, and 30 games in 2015-16, big minutes (averaged just under 22 min/game, both seasons).

    How many is Dreamy going to treat us to this season?

    I’m quietly hoping for a big number, but you know, don’t want to jinx it.

  82. Lloyd B. says:

    Woodguy,

    G answered while I was typing. I was too quick on the draw. My wife would agree. 🙂 I’m going back to reading and not thinking.

  83. DRFNsuperstar says:

    Water Fire:
    With these TH9 stats, IF Davidson stays the course and is emerging, given those first comp numbers this may be possible and good

    Davidson Larsson
    Sekera Klefbom
    Nurse Fayne

    I like this group especially if a more mobile player with passing skills can take over for Fayne at some point.

    Nice idea, two things:
    1. The pairings are probably going to depend who Todd wants facing the tuffs. Sekera and Larsson have a history of doing it fairly well. Klefbom missed a ton of last season and throwing him to the wolves might not be the best.
    2. Nurse struggled to get the puck going the other direction last year, that means he needs to play against 3rd and 4th lines. I don’t know if sticking him with a partner on his offhand is the best idea considering what we know about handiness and the impact on possession

    So I would suggest:

    Sekera-Larsson (Tuffs)
    Klefbom-Davidson (o-zone starts)
    Nurse-Fayne (easy comp)

  84. Little Poteet says:

    Woodguy,

    Thanks for taking the time to thoughtfully answer, it is appreciated. I’m just learning all this. Seems like PDO is a big deal, but I understand it is almost luck based, in that it isn’t generally repeated season to season? Or is it? If taken in a big enough sample is PDO actually indicative of the talent level of a team?

  85. striatic says:

    Woodguy: The Oilers lost the Hall trade.

    Lost it solidly.

    That’s not a reflection on Larsson’s value, which is good, but of Hall’s which is elite.

    I think people underestimate how critical Hall was to the Oilers offence. Best producer on a team with low offensive production. It is painful to think about how much lower a team with basement level Goals For could get without Hall. The painful myth of the offensive powerhouse oilers.

    There’s an argument that I agree with, that the reasons the Oilers are so lousy on offence is because they don’t have the players on D to get the clean breakouts and point pressure necessary to activate their forward talent. Fine. How does Larsson help with that?

    Also, I’m really tired of the “doing the same thing over and over again and expecting different results is insanity” statement, as if it wasn’t possible to do anything worse. Sometimes persistence and patience is important and does lead to different outcomes eventually. There’s an element of truth to the statement but trading Taylor Hall had to be the worst conclusion to draw from it.

  86. jm363561 says:

    LadiesloveSmid:
    G Money,

    Can you run any of Vatanen, Hamilton, Trouba, Jones, and Ristolainen?

    edit: Petrovic as well

    Can you run Davidcon Left side vs Right side?

  87. Aron_S says:

    G Money,

    Just wanted to say thanks for sharing all of this information today (and as someone else already mentioned, hat tip to Woodguy). This stuff looks great and I look forward to seeing where this goes.

    Also, if you’re still taking requests, I’m didn’t notice anyone else asking about him, but can you run Wisniewski? I’m not sure if the sample will work or not since he didn’t play last year, but I’d be keen to see what how his results look, even if they’re a year out.

    Thanks gents, and thanks LT for hosting this discussion.

  88. jm363561 says:

    Woodguy: The base is coach driven then the players will be layered on top of it.

    Fayne will always be Fayne, even if its “new and fit Fayne” its Fayne.

    Does anyone know why we had the “old and unfit” version of Fayne previously?

  89. Woodguy says:

    Lloyd B.:
    Woodguy,

    G answered while I was typing. I was too quick on the draw. My wife would agree. I’m going back to reading and not thinking.

    No!

    Post more!!

  90. Woodguy says:

    jm363561: Does anyone know why we had the “old and unfit” version of Fayne previously?

    Warranty issues?

  91. G Money says:

    Aron_S,

    You bet.

    jm363561,

    Ha ha, ya bugger! I can’t do WOWYs or combos as yet, but that will be the next major milestone, once we get a refined QoC measure locked in.

  92. kinger_OIL says:

    Woodguy,

    WG says: “He swung first.”

    – You have a daughter roughly the age as my son. You know better. This is ongoing bad-blood

    – Your work and collaboration with G is great, congrats: loved it: it takes intuition and assigns numbers, uncovers a lot of gems

    – Picking on others who don’t share the same language, and in this case a bona-fide wonderful poster and person: I didn’t enjoy. Clearly neither did he, as he isn’t posting.

    – Anyway carry on, its not like I am some moral authority on how to interact: I just felt strongly.

    – Be humble and respect differences. We are all just friends on this wonderful blog with various levels of insight, knowledge and access to information to disseminate and interpret.

  93. Woodguy says:

    Little Poteet:
    Woodguy,

    Thanks for taking the time to thoughtfully answer, it is appreciated. I’m just learning all this. Seems like PDO is a big deal, but I understand it is almost luck based, in that it isn’t generally repeated season to season? Or is it? If taken in a big enough sample is PDO actually indicative of the talent level of a team?

    PDO’s luck side is usually heavy to the SH%.

    Goalies have true talents (good and bad) and can drive some PDO on their own.

    Like most NHL level talents, most are in the mushy middle because most who makes it to the NHL is elite compared to those who don’t.

    There can be luck on the goalie side too, but some of it is talent.

    On the team level SH%, lots of luck and some aggregation of talent, but its thin.

    Teams with shit goaltending usually change goalies and that will cause the PDO to fluctuate.

    CAR is interesting as they’ve had bad goaltending for a long time and they just renewed Ward.

    They’ve remained a low PDO team for years and the high event nature of Faulk actually hurts them because of it.

    Here the last 4 years of PDO:

    NY Rangers 101.2
    Tampa Bay 100.9
    Boston 100.8
    Colorado 100.8
    Anaheim 100.7
    Washington 100.7
    Pittsburgh 100.7
    Montreal 100.6
    Minnesota 100.5
    Ottawa 100.4
    Chicago 100.3
    Columbus 100.3
    St. Louis 100.3
    Toronto 100.1
    Philadelphia 100.1
    Dallas 100
    Detroit 99.9
    Winnipeg 99.8
    Los Angeles 99.8
    Nashville 99.8
    Vancouver 99.6
    San Jose 99.6
    Arizona 99.5
    NY Islanders 99.4
    New Jersey 99.3
    Calgary 99.3
    Florida 99.2
    Buffalo 99.2
    Edmonton 98.7
    Carolina 98.3

    Here is last year only:

    NY Rangers 102.4
    Florida 102.1
    Ottawa 101.2
    Washington 101.1
    Tampa Bay 100.8
    Minnesota 100.8
    NY Islanders 100.7
    Pittsburgh 100.6
    Arizona 100.6
    Colorado 100.6
    Philadelphia 100.5
    St. Louis 100.2
    San Jose 100.1
    Boston 100.1
    Los Angeles 99.8
    Winnipeg 99.7
    Columbus 99.7
    New Jersey 99.6
    Chicago 99.6
    Detroit 99.6
    Dallas 99.5
    Nashville 99.5
    Vancouver 99.4
    Montreal 99.3
    Calgary 99.2
    Buffalo 99.1
    Anaheim 98.9
    Edmonton 98.7
    Toronto 98.3
    Carolina 98.3

    Source: http://stats.hockeyanalysis.com/teamstats.php?disp=1&db=201516&sit=5v5&sort=PDO&sortdir=DESC

  94. Ari says:

    WG’s and GM’s work goes a long way to explain the Oilers past decade. There’s a significant difference in the play of Dregs and 3rd pair dmen. I guess that’s why they were always out wayyy before Christmas.

  95. Woodguy says:

    kinger_OIL,

    – Picking on others who don’t share the same language, and in this case a bona-fide wonderful poster and person: I didn’t enjoy. Clearly neither did he, as he isn’t posting.

    – Anyway carry on, its not like I am some moral authority on how to interact: I just felt strongly.

    – Be humble and respect differences. We are all just friends on this wonderful blog with various levels of insight, knowledge and access to information to disseminate and interpret.

    I wasn’t picking on him, I was reacting.

    Maybe a bigger man could have ignored it, but I’m not there yet I guess.

    Also,

    I’ve engaged RTB more than 99% of the posters on this blog trying to figure out what he was trying to get at and valued his opinion.

    Painting RTB as a poor victim here when he repeatedly spewed profanities at us when we respectfully pointed out flaws in his system (like using ONSV% to grade Dmen) or simply disagreed with him is framing the issue in a way that I disagree with.

    Also,

    Pretty sure RTB’s first language is english, he just has issues with writing for communication.

    I’ve tried real hard over the years to bridge that gap and I think you can acknowledge that as you’ve been here.

  96. G Money says:

    Additional defensemen requested.

    One thing that is clear is that with this many different players on this many different teams, I gotta get my ass in gear and add CFRel/DFFRel, as it is critically important to place these guys in the context of their teams.

    **** Alex Petrovic ****
    Top Comp TOI 25348 44% CF (342, 399) 46.2% DFF (274.0, 292.0) 48.4%
    Muddle TOI 19300 33% CF (330, 299) 52.5% DFF (225.0, 194.0) 53.7%
    Dregs TOI 13008 23% CF (237, 214) 52.5% DFF (184.0, 164.0) 52.9%

    **** Mark Giordano ****
    Top Comp TOI 49336 57% CF (773, 783) 49.7% DFF (547.0, 511.0) 51.7%
    Muddle TOI 24367 28% CF (498, 506) 49.6% DFF (337.0, 394.0) 46.1%
    Dregs TOI 12976 15% CF (285, 281) 50.4% DFF (227.0, 200.0) 53.2%

    **** TJ Brodie ****
    Top Comp TOI 42881 55% CF (966, 1073) 47.4% DFF (684.0, 724.0) 48.6%
    Muddle TOI 24182 31% CF (752, 725) 50.9% DFF (541.0, 559.0) 49.2%
    Dregs TOI 11407 15% CF (383, 358) 51.7% DFF (283.0, 242.0) 53.9%

    **** Dougie Hamilton ****
    Top Comp TOI 33227 43% CF (472, 593) 44.3% DFF (315.0, 401.0) 44.0%
    Muddle TOI 27067 35% CF (462, 451) 50.6% DFF (337.0, 358.0) 48.5%
    Dregs TOI 16872 22% CF (290, 231) 55.7% DFF (203.0, 150.0) 57.5%

    **** Sami Vatanen ****
    Top Comp TOI 28884 45% CF (453, 457) 49.8% DFF (340.0, 326.0) 51.0%
    Muddle TOI 20330 32% CF (319, 321) 49.8% DFF (227.0, 212.0) 51.6%
    Dregs TOI 14729 23% CF (234, 203) 53.5% DFF (179.0, 142.0) 55.7%

    **** Jacob Trouba ****
    Top Comp TOI 38342 46% CF (567, 605) 48.4% DFF (407.0, 386.0) 51.3%
    Muddle TOI 28121 34% CF (455, 392) 53.7% DFF (328.0, 283.0) 53.7%
    Dregs TOI 16811 20% CF (286, 222) 56.3% DFF (214.0, 152.0) 58.5%

    **** Seth Jones ****
    Top Comp TOI 30214 38% CF (517, 493) 51.2% DFF (375.0, 364.0) 50.8%
    Muddle TOI 27648 35% CF (452, 427) 51.4% DFF (360.0, 312.0) 53.5%
    Dregs TOI 21586 27% CF (378, 282) 57.3% DFF (253.0, 190.0) 57.0%

    **** Rasmus Ristolainen ****
    Top Comp TOI 47903 54% CF (630, 809) 43.8% DFF (400.0, 529.0) 43.0%
    Muddle TOI 26590 30% CF (392, 460) 46.0% DFF (285.0, 302.0) 48.6%
    Dregs TOI 13727 16% CF (203, 207) 49.5% DFF (139.0, 120.0) 53.8%

    **** Andy Greene ****
    Top Comp TOI 48254 57% CF (477, 683) 41.1% DFF (351.0, 484.0) 42.1%
    Muddle TOI 22812 27% CF (254, 321) 44.2% DFF (224.0, 194.0) 53.6%
    Dregs TOI 14266 17% CF (182, 188) 49.2% DFF (146.0, 139.0) 51.1%

  97. Woodguy says:

    Ari:
    WG’s and GM’s work goes a long way to explain the Oilers past decade.There’s a significant difference in the play of Dregs and 3rd pair dmen.I guess that’s why they were always out wayyy before Christmas.

    This is some interesting reading on the same subject.

    http://becauseoilers.blogspot.ca/2016/04/playoff-teams-in-nhl-average-60-of-dman.html

  98. Woodguy says:

    G Money,

    GET BACK TO WORK!!!

  99. Woodguy says:

    G Money,

    Also,

    Thank you.

  100. G Money says:

    Just running Severson and Wiz.

  101. Woodguy says:

    G Money,

    **** Sami Vatanen ****
    Top Comp TOI 28884 45% CF (453, 457) 49.8% DFF (340.0, 326.0) 51.0%
    Muddle TOI 20330 32% CF (319, 321) 49.8% DFF (227.0, 212.0) 51.6%
    Dregs TOI 14729 23% CF (234, 203) 53.5% DFF (179.0, 142.0) 55.7%

    Wow.

    I think I really missed on him.

    His CF% WOWY with Dmen higher up the roster did not look this good.

    Yay for new information!

  102. Woodguy says:

    **** Jacob Trouba ****
    Top Comp TOI 38342 46% CF (567, 605) 48.4% DFF (407.0, 386.0) 51.3%
    Muddle TOI 28121 34% CF (455, 392) 53.7% DFF (328.0, 283.0) 53.7%
    Dregs TOI 16811 20% CF (286, 222) 56.3% DFF (214.0, 152.0) 58.5%

    Trade Nurse for him (Trouba is big and hits too) and you’re done fixing the D,

    Just play 4 forwards on the PP

  103. Woodguy says:

    G Money:
    Just running Severson and Wiz.

    Yes please!

  104. Woodguy says:

    **** Rasmus Ristolainen ****
    Top Comp TOI 47903 54% CF (630, 809) 43.8% DFF (400.0, 529.0) 43.0%
    Muddle TOI 26590 30% CF (392, 460) 46.0% DFF (285.0, 302.0) 48.6%
    Dregs TOI 13727 16% CF (203, 207) 49.5% DFF (139.0, 120.0) 53.8%

    Playing the toughs with Josh Gorges as a 19 and 20 year old.

    Also know as “The Oiler Way”

    Man.

  105. Woodguy says:

    G Money:
    Woodguy,

    Just installed one of these: http://www.hcamag.com/files/image/Human%20Capital/toilet_desk.jpg

    Perfect.

    Send me the bill along with some of your wife’s scorn for ignoring the family.

  106. Water Fire says:

    DRFNsuperstar: Nice idea, two things:
    1. The pairings are probably going to depend who Todd wants facing the tuffs. Sekera and Larsson have a history of doing it fairly well. Klefbom missed a ton of last season and throwing him to the wolves might not be the best.
    2. Nurse struggled to get the puck going the other direction last year, that means he needs to play against 3rd and 4th lines. I don’t know if sticking him with a partner on his offhand is the best idea considering what we know about handiness and the impact on possession

    So I would suggest:

    Sekera-Larsson (Tuffs)
    Klefbom-Davidson (o-zone starts)
    Nurse-Fayne (easy comp)

    Davidson had the best numbers against tough comp on the team according to TH9 stats, in a small sample. If he keeps on my thought was play him and Larsson against toughs and moreD zone and give Sek Klef more O zone.

    Easier on Klef and Saks first comp numbers weren’t great although he had to carry Fayne. This makes sense to me.

    I don’t get your Nurse comment because he is on his correct side and so is Fayne, and they are 3rd pair which is where they should be. That is a great 3rd pair IMO. A solid D group if healthy especially if they get a chance to get a rhythm going

  107. kinger_OIL says:

    Woodguy,

    WG says: “Pretty sure RTB’s first language is english, he just has issues with writing for communication”

    Kinger says: language = syntax. But you knew that, your just being a “dink” as per WG! (joke)

    – Anyway you are good people, just be kind. This isn’t a race to win

    – G: congrats as well: quantifying intuition is tremedous and uncovers a lot of of mis-perceptions.

    – Over and out.

  108. Ari says:

    Woodguy,

    Thanks for the link. Read it a while ago. Thumbs up!

  109. square_wheels says:

    G Money,

    Ya quit your BS’ing and produce !!

    Top shelf stuff G and Woodguy. BTW I missed what you’re calling this new QoC on Mexican supplements.

    Regardless will continue to endure the random power outages that Taylor Haul causes as long as you keep running it at 230am.

  110. Water Fire says:

    Woodguy:
    **** Jacob Trouba ****
    Top Comp TOI 38342 46% CF (567, 605) 48.4% DFF (407.0, 386.0) 51.3%
    Muddle TOI 28121 34% CF (455, 392) 53.7% DFF (328.0, 283.0) 53.7%
    Dregs TOI 16811 20% CF (286, 222) 56.3% DFF (214.0, 152.0) 58.5%

    Trade Nurse for him (Trouba is big and hits too) and you’re done fixing the D,

    Just play 4 forwards on the PP

    This.

    Also I’m sure Trouba and others can become useful on PP if given the time. Not everyone needs to like the D who play like forwards

    Edit – until Klef is a known commodity I can’t see Nurse going but I wonder if a first and Bear or Jones does it to save cap. Or Fayne retained

  111. LadiesloveSmid says:

    G Money,

    Petrovic does seem to run a bit like Fayne but seems to have more offence, at least based on his output this season.

    Rel% will be pretty important for people like me with little grip on team’s collective CF% and especially DFF%. Vatanen looks pretty solid, wonder if his top comp is with his usual weak partner in Stoner or if he’s bumped up to Lindholm. Build me a WOWY database now GMoney, earn your keep

  112. Mustard Tiger says:

    Water Fire: This.

    Also I’m sure Trouba and others can become useful on PP if given the time. Not everyone needs to like the D who play like forwards

    100%
    We need defencemen on the power play that will walk the line and shoot the puck on net.
    Watching the Oilers, I can’t help but wonder if the power play would be better sending 2 d-men out there rather than the one they normally send. Keep it simple, and also less short-handed chances.

  113. G Money says:

    Does the EV time on Severson seem low? That’s about 15 EV minutes a game. I guess that’s about right for a guy playing second pairing.

    **** Damon Severson ****
    Top Comp TOI 25622 39% CF (320, 408) 44.0% DFF (236.0, 315.0) 42.9%
    Muddle TOI 23110 35% CF (295, 286) 50.8% DFF (242.0, 221.0) 52.3%
    Dregs TOI 17411 26% CF (231, 182) 55.9% DFF (170.0, 152.0) 52.8%

    **** James Wisniewski ****
    I’m getting 15 secs vs Dregs, none vs Top, and 32 secs vs Muddy Middle, and no Corsi events happening, so not much going on for Mr Wiz in 2015.

    **** James Wisniewski 2014 ****
    Top Comp TOI 25932 40% CF (396, 457) 46.4% DFF (284.0, 339.0) 45.6%
    Muddle TOI 22949 36% CF (356, 379) 48.4% DFF (257.0, 291.0) 46.9%
    Dregs TOI 15207 24% CF (246, 191) 56.3% DFF (191.0, 159.0) 54.6%

    (take 2014 stats with a grain of salt)

  114. G Money says:

    Woodguy,

    Will have to pay special attention to him to make sure there isn’t a bug in there somewhere too!

    The TOI does line up nicely though, and the calculation logic is 100% reused from my existing code, so I’m reasonably confident in these numbers (which is why I’m OK putting them out there).

    But the old ‘alpha version oops’ egg-on-face risk is always there!

  115. Lloyd B. says:

    Woodguy,

    Thanks for the detailed response on three buckets versus four. Explains it all very clearly.

    I’ve often been accused of looking for snakes in the grass. There usually are. I can’t play in the stats work but hopefully can contribute to getting you math guys to think about a different angle. Even once would be cool.

    Of course, you guys usually have “been there done that”. Maybe once. just once I will get to contribute to a meaningful change. #Causesnakes.

  116. G Money says:

    LadiesloveSmid,

    Geez, ya give ’em an inch they take a WOWY!

  117. Woodguy says:

    Mustard Tiger: 100%
    We need defencemen on the power play that will walk the line and shoot the puck on net.
    Watching the Oilers, I can’t help but wonder if the power play would be better sending 2 d-men out there rather than the one they normally send. Keep it simple, and also less short-handed chances.

    Trouba does have a cannon for a shot …..

  118. Woodguy says:

    G Money:
    Woodguy,

    Will have to pay special attention to him to make sure there isn’t a bug in there somewhere too!

    The TOI does line up nicely though, and the calculation logic is 100% reused from my existing code, so I’m reasonably confident in these numbers (which is why I’m OK putting them out there).

    But the old ‘alpha version oops’ egg-on-face risk is always there!

    I’ll look closely at him.

  119. Mustard Tiger says:

    Woodguy,

    I meant in general, but I would like me some Trouba on the Oilers.

  120. Woodguy says:

    G Money,

    **** Damon Severson ****
    Top Comp TOI 25622 39% CF (320, 408) 44.0% DFF (236.0, 315.0) 42.9%
    Muddle TOI 23110 35% CF (295, 286) 50.8% DFF (242.0, 221.0) 52.3%
    Dregs TOI 17411 26% CF (231, 182) 55.9% DFF (170.0, 152.0) 52.8%

    Flip side of Larsson is Severson.

    Away from Greene (and therefore away from the toughs) he gets some of the easiest zone starts in the NHL.

  121. Woodguy says:

    Mustard Tiger:
    Woodguy,

    I meant in general, but I would like me some Trouba on the Oilers.

    Me too.

    Really started to like him when I’ve been grinding through Dman data as he showed up in a lot of really good spots.

    And now he shows up really nice here too

  122. Mustard Tiger says:

    Hits hard, shoots hard, numbers back it up? Yes please.
    Woodguy,

  123. Lloyd B. says:

    Woodguy,

    I’m following most of this new stuff and no doubt G explained it. But the TOI numbers are confusing me. I get these are two year numbers but in this example, it’s showing 25,000 minutes of top comp, 23,000 second comp and 17,400 dreg comp. That is 65,000 minutes. Not possible. Are these TOI in seconds?

  124. Woodguy says:

    Lloyd B.:
    Woodguy,

    I’m following most of this new stuff and no doubt G explained it. But the TOI numbers are confusing me. I get these are two year numbers but in this example,it’s showing 25,000 minutes of top comp, 23,000 second comp and 17,400 dreg comp. That is 65,000 minutes. Not possible. Are these TOI in seconds?

    Yes, seconds.

  125. G Money says:

    Lloyd B.,

    Correct. These are 5v5 toi records in seconds.

    Also, the stats are just for a single season (the calculation of the list of players for ‘top comp’ and ‘dregs’ records are based on two years of data, to minimize the good year/bad year effect).

  126. Tire Fire says:

    Woodguy: I’ve engaged RTB more than 99% of the posters on this blog trying to figure out what he was trying to get at and valued his opinion.
    Painting RTB as a poor victim here when he repeatedly spewed profanities at us when we respectfully pointed out flaws in his system (like using ONSV% to grade Dmen) or simply disagreed with him is framing the issue in a way that I disagree with.

    You’ve had the patience of a saint with RTB over the last few years. Nobody ’round these parts has managed to call people stupid and swear at them with quite the frequency that Ricki manages to slip past the ref.

    He seems to have some sort of magic “get out of jail” card that works to dull the impact and fog the memories of the community; even those as fresh as this morning’s ruckus.

  127. Tire Fire says:

    Oh, and big ups to G Money and Woodguy for the hard work.

    Really interested to see how well this works when you’ve gotten it finalized.

  128. Bruce McCurdy says:

    Woodguy: Really impressed with Gryba. If Mongo Taylor Haul 9000 was fully functional I’d dive into that one deep to see if it was real.

    I’ve liked Gryba since he was in Ottawa. Thought he was fine for Edmonton this past year, a break-even player who was strong on the PK and delivered some attributes the team was short of.

    Anyway, his name came up repeatedly in a post I published earlier today about Larsson, using GA/60 REL as the key stat. Gryba looked good by this measure over the most recent season and over the past four seasons, and also looked good on the PK.

    Woodguy: Faulks CF% being 49.2 and DFF being 43.8% is something I saw throughout the data on him.

    Good possession numbers, but his box protection would anger every bear.

    The other guy whose name came up repeatedly in my post today was Faulk, and not in a good way. In this respect my method seemed to corroborate your earlier work on this player. Is he the Justin Schultz of the east?

  129. AsiaOil says:

    GM and WG simply tearing it up today – real impressive guys and thanks.

    Probably a good thread to update the bet GMoney and I made last year with the year 1 numbers:

    a. that within 2 years (end of 2016-2017 season), Scrivens is back to being an above average to elite NHL goalie (call it .917 or above all sits, or .922 EV, adj sv%)

    2015-2016 result = .906 AS / .926 ES

    b. that within 2 years (end of 2016-2017 season), Cam Talbot is an elite NHL goalie (.920 AS and .925 EV, adj)

    2015-2016 result = .917 AS / .920 ES

    c. that by end of next season (2015-2016), Jhonas Enroth is established as a legitimate NHL starter (similar to Scrivens’ projected numbers)

    2015-2016 result = .917 AS / .937 ES

    d. that by end of next season (2015-2016), Karri Ramo is established as a backup and not a starter (.915 or lower all sits)

    2015-2016 result = .909 AS / .917 ES

    It’s a mixed bag with G man doing a bit better than I expected. Interesting thing though – 3 of the 4 guys are unsigned UFA at this point – and it’s entirely possible that both Scrivens and Ramo are left without jobs this year. Not many seats left at the dance. Enroth is a quality backup – had a good year in that role – but only 13 starts as he was a clear backup to Quick behind a defensively solid team. I think he ends up in SJS. Scrivens back to TOR – not many teams go to the well twice with goalies. My take last summer was that Ramo and Scrivens were Mendoza line or lower. Talbot was unproven and unknown given the lack of record, and Enroth was a quality backup. I continue to feel that way but Talbot has a more well defined record now and looks like an average starter by eye and number to me.

    As for the Oilers goaltending situation. There is what you want and what you can get. With LB almost ready and holding a one way contract in his hands – few quality backups will be interested in EDM – so Gus is probably the best you can get at this point.

  130. Barcs says:

    Woodguy:
    G Money,

    **** Hjalmarsson ****
    Top Comp TOI 56094 63% CF (767, 784) 49.5% DFF (574.0, 539.0) 51.6%
    Muddle TOI 21859 25% CF (375, 322) 53.8% DFF (272.0, 267.0) 50.4%
    Dregs TOI 10415 12% CF (202, 137) 59.6% DFF (153.0, 109.0) 58.3%

    Here’s something interesting to know.

    I’m working on my “is Larsson being zoomed by Greene?post.

    There are two other RHD in the NHL who:

    1) Had very, very tough zone starts (although Larsson were the toughest, by far)
    2) Played mostly tough comp
    3) Managed to still lower the RelCA/60
    4) Managed to have a + Relative xGF%

    Larsson, Hjarlsmasson and Tanev.

    Both are probably good comps for Larsson and I think that Dreamy-Larsson might be a top pair that could be as effective as Keith-Hjarlm very soon.

    All for $8.2MM combined for the next 5 years.

    A small addendum to this great thread:

    N. Hjalmarsson (CHI) is left handed.

    If he’s playing RD with Keith and producing those numbers, that’s astounding.

  131. Water Fire says:

    Bruce McCurdy: I’ve liked Gryba since he was in Ottawa. Thought he was fine for Edmonton this past year, a break-even player who was strong on the PK and delivered some attributes the team was short of.

    Anyway, his name came up repeatedly in a post I published earlier today about Larsson, using GA/60 REL as the key stat. Gryba looked good by this measure over the most recent season and over the past four seasons, and also looked good on the PK.

    The other guy whose name came up repeatedly in my post today was Faulk, and not in a good way. In this respect my method seemed to corroborate your earlier work on this player. Is he the Justin Schultz of the east?

    Yes other than he’s strong and scores a lot of points ha ha

  132. Water Fire says:

    Barcs: A small addendum to this great thread:

    N. Hjalmarsson (CHI) is left handed.

    If he’s playing RD with Keith and producing those numbers, that’s astounding.

    I think maybe it’s because GMWG data shows Keith is a monster . Harms is good but methinks he’s the second fiddle

  133. OF17 says:

    GMoney,

    Great answers, thanks for taking the time. The fact that the starts of top opposition and of the dregs roughly correspond to a standard deviation away from the mean is quite encouraging for the three-tiered system. Really great work you’ve been doing, and even though it’s not even complete yet, we’re already seeing the use of it in recent threads. Thank you for it!

  134. theres oil in virginia says:

    kinger_OIL,

    I agree with you that it wasn’t “fun”, but I disagree with how you’ve framed it. RTB made some ridiculous and aggressive statements and was wholly unprovoked. I thought that both Woodguy’s and GMoney’s responses were quite muted. (Leadfarmer could also have responded poorly and chose not to.) Certainly worse thoughts ran through my head than what they posted in response. RTB is usually a lot of fun, but that episode was ridiculous and uncalled-for.

  135. mustang says:

    theres oil in virginia:
    kinger_OIL,

    I agree with you that it wasn’t “fun”, but I disagree with how you’ve framed it.RTB made some ridiculous and aggressive statements and was wholly unprovoked.I thought that both Woodguy’s and GMoney’s responses were quite muted.(Leadfarmer could also have responded poorly and chose not to.)Certainly worse thoughts ran through my head than what they posted in response.RTB is usually a lot of fun, but that episode was ridiculous and uncalled-for.

    WG and G, have mostly been gracious with the bear man and to all posters really. I think the bear man gets fired up when the peanut gallery starts taking shots. Really it comes down communication and the lack there of.

  136. Ryan says:

    G Money,

    Petrovic looks like a great spare part to pry loose.

  137. Ryan says:

    Woodguy: The Oilers lost the Hall trade.

    Lost it solidly.

    That’s not a reflection on Larsson’s value, which is good, but of Hall’s which is elite.

    “My first thought: Really? Wow.

    My second thought: Just for Adam Larsson? Wow.

    My third thought: Is Lou Lamoriello secretly running the Devils? Wow.

    The big news today is that the New Jersey Devils acquired Taylor Hall from the Edmonton Oilers in exchange for Adam Larsson. Yes, this is not a joke. Yes, this happened; it was even posted at NHL.com. The first overall pick from the 2010 NHL Draft is now a Devil.”

    Pretty much says it all right there.

    http://www.allaboutthejersey.com/2016/6/29/12062174/breaking-down-taylor-hall-adam-larsson-trade-new-jersey-devils

Leave a Reply

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