Phoenix Coyotes: WTF?
by Hawerchuk on Mar 31, 2010 9:00 AM EDT
Email
PrintI didn't hold out much hope for the Phoenix Coyotes at the start of this season. But they're going to finish in a completely different dimension than I had them in:
| GF | GA | Pts | |
| Actual | 227 | 202 | 108 |
| Predicted | 222 | 254 | 80 |
All stats are from before last night's game and those predictions come from Tom Awad's Vukota system. Interestingly, the driver of Phoenix's success this season is blindingly obvious when you look at those numbers - Tom hit the offense dead-on, but he had team defense and goaltending pegged at more than eight wins worse. So what happened? First of all, Phoenix's two goaltenders have played out of their minds:
| Goalie | Career SV% | Sv% | Shots | Goals Saved |
| Labarbera | 0.906 | 0.928 | 404 | 8.9 |
| Bryzgalov | 0.914 | 0.921 | 1850 | 13 |
| Total | 0.913 | 0.922 | 2254 | 21.9 |
As Tyler Dellow noted, Labarbera has made Phoenix's record about three points better...And he has no value in the playoffs. Bryzgalov has played unexpectedly well and has a claim to be the second-best goaltender in the league given his even-strength save percentage. But that's not enough - if the goaltenders merely played at their career averages, Phoenix would have a 227-224 goal differential, which would put them in the low 90s points-wise.
But no one could have predicted that Phoenix would go 17-6 in overtime and the shootout. That kind of play is simply unsustainable for any modern NHL team. So let's give them average luck: a 12-11 record in extra frames. Then we're at 222-229 in the goals department.
Phoenix has had an extra benefit: timely scoring. They lead the league in winning percentage when scoring first, which is the same kind of talent as winning one-goal games: not repeatable. So while 222-229 might normally make you an 90-point team, in Phoenix's case, timely scoring got them to 97 points, with unexpectedly good goaltending carrying them the rest of the way to 108.
Still, Phoenix was a 90-point team and not an 80-point team as projected. So something else must have gone right. Well, Kurt Sauer went down at the beginning of the season, which pushed Ed Jovanovski into a shutdown role that he hasn't played in years. Jovo's been getting shelled, though less so than Sauer was last year, and together with Keith Yandle's step up, the defense has stayed afloat. But it's not what's driving a ten-point improvement. Nor is Phoenix's "core" of young players the driver - though Martin Hanzal has played better than he did last year, Mikkel Boedker, Kyle Turris, Peter Mueller and Viktor Tikhonov have produced essentially zero for the big-league team this season, which is less than they brought to the table last year.
No, the true changes are very subtle: a huge improvement in faceoff percentage, from last year's historic bad to this season's league-average; and some improvements around the edges with acquisitions like Scottie Upshall, Vern Fiddler and Radim Vrbata. Vrbata, in particular, has been great.
So while this Phoenix team should have beaten their projection by a number of wins, they are fundamentally not the 108-point team they're going to finish as. If they come up against Detroit in the first round of the playoffs, they should be a serious underdog.
50 comments
|
0 recs |
Do you like this story?
Comments
Pretty much a season-long WTF, which has to be a record.
Bettman's Nightmare: A Blog Where Hockey Aficionados Dismantle That Mighty Empire, One Balsillie at a Time
http://bettmansnightmare.blogspot.com/
by Bettman's Nightmare on Mar 31, 2010 9:49 AM EDT reply actions
I dunno Gabe. An expected goal differential of 222-254 would imply that Phoneix is expected to get dominated at even strength PLUS have poor special teams.
Now their special teams haven’t been all that by at even strength they hold territorial advantage. It’s not due to score effects either, they do it while the score is close.
Now that doesn’t explain a 100-point outing and you nail many of the reasons why they’ve been able to win at the pace they have and some would rightly point out that winning 8 or so games in a row is just not a reasonable expectation for any 8-game stretch, and the fact that the Coyotes did so (did they actually win 9 in a row? don’t remember) undeniably buoy them up in the standings.
But for my money they’re not even the closest WTF story in the Western Conference, let alone the league.
Right – I think they’re a 90-95 point team (ie – territory near 50%) that fluked into an extra 6-8 wins. They got 14 points on goaltending and OT luck alone – I suppose I didn’t really need to tell the rest of the story.
Obviously many other WTFs – San Jose and Colorado are winning despite poor underlying numbers.
I think probably the biggest WTF is San Jose being poor territorially. I mean I don’t know who could have foreseen that, they were going to have issus with depth but they signed a couple of guys in Malhotra and Nichol who, while not world-beaters, aren’t sinkholes either.
Also VAN’s surge, their underliers at even strength while the score is close suggest a poor even strength team. I haven’t checked in detail but my first thought is the bottom of the roster giving it all away.
And Colorado, if they make it, would be categorically the worst playoff team. An argument could easily be made that they belong in 14th place in the West.
Phoenix’s upturn at even strength play is surprising but not so much so as San Jose because they’ve acquired some decent players in Aucoin, Lang, etc. etc. And Hanzal has become a man.
Gabe,
Why no detailed analysis of the ‘Yotes’ offensive numbers? Just because their Goals For ended up close to where you projected doesn’t necessarily mean it SHOULD’VE happened that way. In other words, you played the regression game on one side of the ledger – only fair to do the same on the other, right? ;)
by sunnymehta.com on Mar 31, 2010 12:33 PM EDT reply actions
I did the o-side in my head and decided it wasn’t that interesting compared to goaltending and timely scoring. Aside from the PP, do you think they’ve been unlucky offensively?
no idea – haven’t even looked at it. i was being lazy and hoping i could just read your analysis of it in this post. :)
by sunnymehta.com on Mar 31, 2010 2:01 PM EDT up reply actions
Not exactly sure how/if that can be quantified, but something has to be said about coaching… It might just be a huge coincidence, but Tippet has a decent coaching record and the moment he takes over the team, they out-perform expectations.
Doesn’t take away the lucky OT record or the goaltending though… but it must make a difference.
Dave Tippett: 569 games coached. Expected points = 506.5. Actual points = 507.
If he had these skills, he should have made Marty Turco into a good goaltending for the four seasons following the lockout.
Where are you getting the “expected points” numbers from for coaches?
by Scott Reynolds on Apr 1, 2010 12:47 PM EDT up reply actions
So you’re measuring “coaching” skill by whether or not a team outperforms its’ goal differential? If so, that seems odd. Why would that measure coaching?
by Scott Reynolds on Apr 1, 2010 5:48 PM EDT up reply actions
No way is Tippett “a huge coincidence.” The system he’s instituted has been a huge, huge factor in their success. The comment about ‘if he had those skills, he should have made Turco a good goalie in Dallas’ are off-base. A goaltending coach is much more vital to a goalie’s success than the head coach. Tippett is the biggest reason they are where they are. Followed by Bryzgalov. Then clutch scoring across the roster……and then Adrian Aucoin’s shootout snipes, haha.
Seriously, prove it. Let’s say they finish with 106 points, 26 over their projection. Bryzgalov, Labarbera, the SO and OT are worth 14 points. So we’re left with 92 points. Better faceoffs are worth 3 points because they were so bad last year. 89 points.
So you’ve got 9 points to distribute between Vrbata and Hanzal being better, “clutch” scoring (aka luck), and Tippett. Best case, Tippett is worth 2 wins more than Gretzky…But that says Gretzky sucked, not that Tippett is a genius.
In response to both this and R O’s post below, I will point out a few things that Tippett has done that are different from Gretzky that are making a difference to this team.
First and foremost, there is a system which did not exist when Gretzky was coach. The players are asked to do specific things and they are evaluated based on how they are performing in 10 game increments. This is certainly not revolutionary, but for 3 years Gretzky did none of this and, particularly for the young players, they were lost.
Now I can’t break it down in technical terms because I didn’t play competitive ice hockey, but from my basic knowledge – it’s a strong one-man forecheck that comes in waves (i.e. when the puck moves, the forechecker will peel off onto one player and the next forward pushes in), the d-men start at the redline, work back towards the blue line and try and stuff the attack there without much backing into the zone. It’s a kind of trap, without actually being a trap. It generally does not change based on the score, although the forecheckers will be more aggressive down late (could lead to timely scoring benefits). When the other team has solid defense he plays a sort of man-to-man system with the forwards staying high above the circles covering d-men, the defense covering either deep forward just outside the posts and the centers pinching back forcing the puck. In the offensive zone it is a pure cycle on the strong wall with the players trying to work it out in front from the end wall or feed it in front to a cutting in d-man. The penalty kill is a large box with the defensemen deep and the forwards playing aggressive towards the side wall while giving the men at the point a bit more room. The powerplay just hasn’t had a solid quarterback all year, so it becomes disjointed. I’d call it an umbrella set up.
Because the players all know their role, they are rarely out of position…a far cry from last season when because the team had no system and Gretzky shuffled his lines continuously, players were constantly out of position.
Which gets me to point 2 – Tippett basically has kept the lines together as units. We have our Czech line (Vrbata – Hanzal – Prucha) which play against the other team’s top line. Now the lines have shuffled a bit since the trade deadline, but not too much (Doan – Lombardi/Fiddler – Pyatt) (Wolski – Fiddler/Lombardi – Stempniak) (Korpikoski – Winnik/Lang – Nokeleinen). The d-pairings also have been pretty consistent (Aucoin – Vandermeer) (Yandle – LePisto/Morris) (Michalek – Jovanovski). They aren’t the sexiest lines in the league, but they finally have had a chance to develop chemistry and it has paid dividends.
Tippett/Don Maloney brought in Sean Burke to be goalie coach this year and he spent a TON of time working with Bryz and LaBarbera on their positioning. The number of soft goals that Bryz has allowed this year are WAAAAYYY down from last year (no idea how you would quantify this). Even though he’s facing more shots, with his improved positioning he is making more saves than he did last year. This has been a coaching improvement more than anything else…
So yes, these players are playing better…and there has been some SO/OT luck in the mix, but I have no doubt that if Gretzky still coached this team (with the rest of his staff – Ulf Samuelson, Grant Fuhr, Doug Sulliman) as opposed to Tippett (wtih Ulf, Dave King and Sean Burke) they wouldn’t be having this type of improvement.
Support Coyotes Hockey - Five For Howling
(Oh, and go Philly teams as well!)
What I will grant you is that Gretzky was a poor coach. Tippett is average. For players who had been with Phoenix for a while, their projections were a little lower because they had played for Gretzky. But this wasn’t true for Pyatt, Vrbata, Upshall, Fiddler, Korpikoski, Nokkelainen, Aucoin, Vandermeer, Lombardi, Stempniak, Wolski or even most of the young guys. So maybe, maybe that was worth two points. Maybe.
I don’t think there’s any evidence that keeping line pairings together causes a team to do better. And for all of Bryzgalov’s improvement, he’s still “worse” than he was two years ago.
I wouldn’t overestimate the coach’s role on a team. Phoenix has goaltending (whether by luck or by skills), amazing OT performance (luck, for sure) along with a slightly better roster than they had last year. Tippett may be a more professional coach than Gretzky, but the projections didn’t really expect a coach who would drag down his players anyways.
I don’t know the math behind everything, so it’s hard to make a coherent argument. I would ask, who is a good coach? I mean, if Tippett is just average, then I’d like to know who is above-average. Mike Babcock? Lindy Ruff? Ken Hitchcock? I’m not trying to be difficult, but I think there is a possibility that you underestimate the role of the coaching…
Not that Tippett would tell you that…I know, I spoke to him – he gives all the credit to the players for bringing the type of extreme, never-say-die, pack-mentality effort that has allowed them to succeed. But Tippett instilled that mindset into them. It might be the type of “mumbo-jumbo” that R O and you don’t want to hear about, but it’s a factor on this team. I’ve been a STH for 3 years now and the roster (although improved) isn’t all that different in terms of talent….Tippett is getting more out of them than his predecessor…a lot more.
Support Coyotes Hockey - Five For Howling
(Oh, and go Philly teams as well!)
Not that Tippett would tell you that…I know, I spoke to him – he gives all the credit to the players for bringing the type of extreme, never-say-die, pack-mentality effort that has allowed them to succeed. But Tippett instilled that mindset into them. It might be the type of "mumbo-jumbo" that R O and you don’t want to hear about, but it’s a factor on this team.
To be extreme, never-say-die, and have a pack mentality… compared to the actual ability to play hockey well in some or all of its aspects, it just doesn’t matter. If you believe it matters, I have oceanfront property in your neighborhood to sell you.
And while I have read anecdotes about Gretzky’s strange defense matching tactics… it’s just an out-and-out lie that Gretz didn’t have a “system”. All coaches at this level have systems for offensive zone pressure, defensive zone coverage, forechecks (which undeniably change according to the score, it’s laughable that you think otherwise… and some coaches just decide to leave a lot of decision-making power in the players’ hands anyways), PPs, PKs, everything.
And looking back at the usage of players under both regimes… the young Hanzal playing brutally tough minutes was a constant. And that’s a choice, because young rookies overwhelmingly fail in this role. Hanzal looks to have been good at it from Day One, it was obvious for both Gretz and Tippett to play him to his strengths.
It’s funny you praised Tippett for playing Hanzal against top competition and didn’t mention that Gretz did the same. This is one of the principle ways that coaches affect results, by who they match up against the opposition’s best. And you couldn’t have missed it watching PHX for three years, I mean I couldn’t miss it watching five games.
And somehow Tippett is still responsible for five to ten wins of improvement?
My ass. That’s bullshit, and disingenious bullshit at that.
And this whole business about “speaking to Tippett”, that’s the cherry on the cake.
And it’s funny, I’m no PHX fan but I think I give more credit to the PHX players for being generally good players than you ever will. Are you a friend of the Tippett family?
by R O on Apr 1, 2010 4:59 PM EDT up reply actions
Look, you can believe whatever you want…I didn’t come to pick a fight with anyone here or what anyone was saying, and I’m certainly not a personal friend of the Tippett family. I did get to speak to him at a team event, and when I talked to him about the difference between last year and this year, he credited the players for it. So I don’t know what you want me to say there…
I certainly know that Gretzky played Hanzal against top competition since his first game in the league. It’s obviously his strength and I’d be happy to credit Gretzky for it. You asked who plays against top competition, and I answered…I wasn’t implying that particular factor was something that Tippett was so better attuned to than Gretzky.
That being said, for everything else you are arguing, I don’t know what to tell you. Tippett does try to get the team to play the same type of game regardless of the score…I’ve heard him say it during his press conferences, and it’s apparent by their play on their ice. So I’m sorry you think it’s laughable I think that – I’ve been watching Phoenix games for a lot longer than you…
As for the argument that Gretz didn’t have a system…he didn’t. Derek Morris and a variety of other players that left the team said it when they left (http://www.puckupdate.com/2009/09/10/gretzky-my-system-is-no-system/) So yeah, there was a system…it just changed every day…with a team full of rookies and younger players. It wasn’t exactly a formula for successful coaching.
Again, maybe none of this makes Tippett a great coach, or anything more than average…but who is a great coach, is still my question?
As for your last point, it’s laughable beyond ludicrous. I do praise the heck out of the players and think that the Coyotes roster is a heckuva lot better than almost any pundit or outside critic would ever say. I think there’s a ton of talent there. But there’s been a lot of talent on this roster in past years too and it never stepped up and performed. So either the players just got better overnight, or they are playing better under a coaching system that works for them. I don’t understand why that’s so difficult to understand.
Last year’s Penguins team was floundering before Dan Bylsma was brought in. With a roster that was remarkably the same, that team started to win games and won a cup…are you telling me that none of it had to do with Bylsma?
Support Coyotes Hockey - Five For Howling
(Oh, and go Philly teams as well!)
Coaches have virtually zero impact on a team’s performance. I think Lemaire and Hitchcock can do a little something with nothing, but other than that, the spread between average and very good is tiny.
will do…I look forward to it.
Support Coyotes Hockey - Five For Howling
(Oh, and go Philly teams as well!)
fair enough. I don’t agree, but I can’t prove anything and in the end I don’t completely disagree with your conclusions. Bryzgalov and Labarbera have been huge, and they’ve been lucky to win games in OT/SO. I just don’t think that if you had a monkey in a suit behind the Coyotes bench this year, they’d be doing this well…
Support Coyotes Hockey - Five For Howling
(Oh, and go Philly teams as well!)
Or better yet, just tell us what Tippett’s systems are. How does he line up his players when the opposition has solid possession at EV? How about when his team has solid possession? Forecheck? Does it depend on the score? Who does he have playing tough minutes? Who is he sheltering? What PP configuration does he use most often? PK? How does this affect how good or bad his players look on the ice?
I bet you don’t even know the answer to even one of these questions.
There’s a place for talking out your ass, unfortunately that place is the toilet.
R O, no need to completely trash AlexAZ, all he was he was if Tippett has made a significant difference on the Yotes record this year which is a legitimate statement.
For example, if the Flames applied a different system then what Brent Sutter has implemented (ie. sending out the 4th line after icing calls), would they have a better record?
I’m not sure I’d call it a legitimate statement. He said Tippett was most responsible for Phoenix’s improvement. But a reasonable estimate says:
Bryzgalov: 5 points
Labarbera: 3 points
OT/SO Luck: 5 points
Faceoffs: 3 points
Vrbata playing well again: 3 points
We have 7 more points of improvement to hand out. Nothing for Upshall? Fiddler? Hanzal?
At best Tippett is worth 4 points, which is less than Bryzgalov. More likely Tippett is actually worth zero points (he does not hurt his team.)
I think you have to run through some kind of analysis to make a claim about Tippett’s ability. An unjustified opinion doesn’t win you many fans!
Also, I can’t believe that in six months of blogging, I’ve only used the word shit twice and fuck zero times.
Yeah, obviously Tippett doesn’t have everything to do with it, he was pointing out that he has been a part of it. I would say that’s legit, and that he’s doing a better job then Gretzky. Also, there are many more factors that would play into PHX record this year then what you just stated.
Also, there are many more factors that would play into PHX record this year then what you just stated.
Such as?
And be honest now, only things that happened on the ice. None of this feel-good BS about leadership and confidence.
Do tell, what factors play into the PHX record? List them all.
I wasn’t talking about "leadership " or “chemistry” did i state that anywhere RO?
What I meant to say maybe their PP has cost them a 1 point (I’m pulling that number out of the air). Or Jovo’s improved play from last year has got them an addition 0.5 points (again that number is out of the air). Make sense?
How much of the goalie’s improvement could possibly be due to coaching, i.e. by reducing shot quality and/or improved team defense in front of the goalie?
Just throwing that out there, I have no idea, but attributing it all to the goalie may or may not be correct either. Especially with both playing ‘over their heads’ there may be other underlying factors.
I don’t know that Bryzgalov has “played out of his mind” or even “unexpectedly well”. His save percentage the last 4 years have been 0.910, 0.907, 0.921, and 0.906. His even strength save percentages have been 0.919, 0.907, 0.928, and 0.912. This year’s numbers of 0.920 and 0.924 are within his expected performance range.
by DoctorMyBrainHurts on Mar 31, 2010 8:40 PM EDT reply actions
Take a look at EV:
2009-10: .928
2008-09: .918
2007-08: .934
2006-07: .907
2005-06: .919
Average: .921
.007 is a big difference. And it’s mostly luck. When a goalie plays so much better than his established talent level – so much so that it’s the main reason his team has overperformed – I call that “playing over his head.”
btw, I'm going to flip this around and blow your mind
2009-10: .072
2008-09: .082
2007-08: .066
2006-07: .093
2005-06: .081
Average: .079
Bryzgalov is playing 10% better than his recent performance. That’s seriously huge for a goalie. It’s more than the difference between Martin Brodeur and Ty Conklin.
Yeah Gabe, I’m thinking for around league average of 21 EV Shots against per game, and allowing for 70 or so starts, an EV SV% uptick of 0.007 results in 10.3 goals prevented or 3.4 points.
That’s kind of a lot. Especially when you’re a reasonably good goaltender to begin with, improving on your performance to the tune of another win and a half is crazy. You’d deserve, what, a $3.4M bonus?
Add in some puck luck on the Penalty Kill and 5 points is not out of the question. And you might not want to bet your house on whether or not that particular goalie will repeat his performance next year, but you could probably bet ten houses on the top and bottom 5 starting goalies regressing and you’d likely break even on houses.
I get a mean even strength save percentage of 0.916 with a 95%CI of 0.902 to 0.934. 0.924 is less than 1 SD above his mean. This isn’t even his best year. His numbers were obviously better in 2007-2008
by DoctorMyBrainHurts on Apr 1, 2010 12:28 AM EDT up reply actions
All this really suggests is that Bryzgalov has a decent midpoint (approximate “true talent level”) for his ability to stop pucks at 5on5. But his year to year results are highly influenced by luck.
The same can be said for most goalies in this league. It’s not something you can reasonably plan for, to have your goalie let in only 7% of the shots he faces at even strength. Unless you also reasonably plan for your goalie sieving out for 9 to 10% of the shots he faces at even strength.
And when the difference is translated to wins… the discrepancy is even more stark. It doesn’t pass the sniff test, that an already decent goalie who probably provides anywhere from 1 to 2 wins above replacement for his team, can provide yet another 1 to 2 wins above expectations due to factors under his control.
It’s luck.
Also, standard deviation is only a relevant concept if you think SV% is normally distributed (I don’t think that’s possible given the nature of the concept of puck stopping rate).
by R O on Apr 1, 2010 1:10 AM EDT up reply actions
We don’t know his (or anyone’s) TRUE EVEN STRENGTH SAVE PERCENTAGE (TESSP)(sorry for the caps, but I need to use some formatting). We have observed even strength save percentages (oessp). These individual observations estimate his TESSP. We can calculate an average (mean) oessp. This also estimates his TESSP. In fact, it’s our best estimate of TESSP. But it’s still just an estimate. Our observed values of oessp are likely normally distributed with a mean of TESSP . And even if they aren’t truly normal, the are almost certainly normal enough, as the technique is pretty robust.
I would not call small variations around his expected value good or bad luck as much as I would just call it statistical noise. And I would call +/- 1 SD a small variation.
As far as this translating into wins, no question that the observed wins can’t be accounted for solely by his save percentage. There are other, “confounding”, factors.
by DoctorMyBrainHurts on Apr 1, 2010 10:51 AM EDT up reply actions
Tyler did a nice post showing that save percentage regressed something like 50% to the mean. So your true talent 95% CI is 908 to 924. And I still think that overestimates an individual goalie’s talent variation (as opposed to leaguewide talent variation.) So there’s maybe a 1-in-50 chance that Bryzgalov’s deviation is talent-related.
And his year two years ago. Insane.
You lost me there.
I agree that his even strength save pecentage or QASP will regress to a mean. His mean. It will regress to the league mean if he’s an average goalie.
I don’t think there’s a talent variation from year to year. The numbers vary because they are observations, samples from an underlying distribution.
I’m not sure where you came up with 1-in -50 either. He’s performed at this level in 2 of the last 5 years.
by DoctorMyBrainHurts on Apr 1, 2010 11:02 AM EDT up reply actions
What I’m saying is that 50% of the observed variation in save percentage is talent, and 50% of the observed variation is luck. So when you take all of the save percentages in the league and find the standard deviation of them, you are overestimating the role of true talent by a factor of 2.
So Bryzgalov is really 2 s.d. out from his mean this year, and actually, I believe more than that because an individual’s talent variation is less than what you found by looking at all goalies (I assume that’s what you did.) So his performance is luck-driven and not talent-driven.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
