Wednesday, July 31, 2013

The Oilers and reasonable expectations

I don't follow many people on twitter - as of today I only follow 37 people.  But I do keep track of a number of other people on twitter, particularly those in the sabermetrics and hockey analytics community.  A good # of those people, including one person who I follow (Jonathan Willis), are Oilers fans.  This shouldn't be a surprise to anyone - the Oilogosphere (as I believe it's called) is heavily active online and features some of the brighter analytical thinkers out there for Hockey. 

That said, I do think there's a pattern out there in the Oiler blogosphere, at least in certain parts, where people who should know better have higher than reasonable expectations for the team to begin the season. 

What touched this off by the way, though a similar idea has been in my head for a bit, was Willis taking issue with Rob Vollman labeling the Oilers as likely 30th next year.  Now I don't agree a lot with Vollman's methodology, and I don't know what he used here.  But let's begin by saying such a prediction isn't TERRIBLY unreasonable.  Again, last year's Oilers were:

28th in Fenwick Close (a miserable 44.48%) and 29th in Corsi tied (at 44.5% per HockeyAnalysis).  This was not a good team, despite the solid goaltending of Devan Dubynk and the 3 #1 overall picks. 

Who's returning from that team?  (Relying upon this: http://oilersnation.com/2013/7/8/the-edmonton-oilers-today-tomorrow/)
First Line: Hall-RNH-Eberle - Completely the same, although RNH may miss some time.
2nd  Line: Yakupov-Gagner
3rd   Line: Hemsky
4th   Line: Smyth, Jones
Extras: Brown
D: Smid-Petry
D: Schultz-Schultz
D: Potter

So you have a decent amount of turnover, but a large amount of players from the awful Oilers team of last year ARE coming back.  Of course, saying that is missing part of the picture.  RNH should be better if he recovers fine from surgery, as should Hall ( a scary thought).  Same for Yakupov of course.  On the other hand, that 4th line is still bad.  And well, the D was pretty awful, and 5 of those guys are coming back.  And only Justin Schultz should really be expected to be a little better. 

So again, despite the presence of 3 top picks, this isn't exactly a strong core.  How about the additions?  Well on D, the Oilers clearly have made a bunch additions:

Andrew Ference is probably the most high profile, and as noted on Copper & Blue, he's just not...very good.  Some of the negative parts of his relative corsi undoubtedly comes from the fact that the #1 D on Boston was a guy named Chara, who he didn't play with.*  On the other hand, the last two years, his relative corsi is more than a little bit bad - it's pretty darn bad.  He's not exactly being buried in the D Zone either (Neutral Zone Starts).   In the small sample of the playoffs last year, he was easily Boston's worst D Man in scoring chances.

*Tyler Dellow has argued that in addition, Ference's time with Chara was handicapped by playing off-hand, further limiting his relative corsi.  Of course, the negatives of playing off-hand are at this point purely anecdotal (SOMEBODY RESEARCH THIS) and I can point to plenty of examples of off-hand players performing quite excellently.  

Denis Grebeshkov is another addition - a D man who went to Russia after a few years in the NHL.  He had two good years for the Oilers under MacTavish, and then had a bad year where he was eventually shipped to Nashville.  From there he went to the KHL, where his career is interesting.   After a few years of high usage suggesting being a top KHL D man, Grebeshkov's TOI dropped with his team SKA until this year, when he was traded to Yugra.  In Yugra, his minutes returned.  I'd guess he was probably considered a borderline top pair KHL D man in Russia, but it's hard to tell from that kind of time.  As for the NHL - it's hard to say - is the the 2009-2010 guy or the 2008-2009 guy?  The latter guy won't help the Oilers too much.  Grebeshkov is also going to be 30, so he's not going to get much better. 

The other addition is Anton Belov, another KHL D man.  Belov played a ton of KHL minutes, and has a pretty praising scouting report up on Oilers Nation.  On the other hand, Derek Zona's recently tweeted out D pairs suggested he was the current 6th DMan or fighting for that spot with Nick Schultz.  There's a lot of uncertainty there, and he could be good.



Finally, Oskar Klefbom is the wild card.  Pronman projects him as a #2 D Man.  If he was to play, we'd probably figure he's more likely to be a #4 type guy right now.  But that's still an improvement.  Of course, it's hard to see where he cracks the lineup right away. 

So the Oilers will add potentially 3 guys whose abilities are highly unclear but could range from good 2nd pair to meh 3rd pair at best and a meh player in Andrew Ference.  Other than Ference - a bad signing - these are actually smart moves for the Oilers.  But the ranger of expectations is high here.  It's entirely possible that Grebeshkov and Belov bust, and then you're left again with the same bad Oilers D.  So it's not like you can't see the D of the Oilers leading to another poor finish. 

How about the forwards?

The Oilers have added 2 "major" additions at Forward.  The first, a swap of Parjaavi for David Perron, seems like an improvement (although Parjaavi could've improved as well).  He's a solid possession forward (although some argument can be made others were driving the bus) and adds some points.  The 2nd line last year was dreadful for the Oilers, despite the supposed high quality of Gagner, but at the very least Yakupov should get better and Perron should be an improvement at wing.  Is that enough? Hard to tell. 

The other is Boyd Gordon, a defensive center to replace the failure of Eric Belanger.  Of course, Belanger was once thought to be a pretty good defensive center whose signing was celebrated at Copper and Blue.  Meanwhile, it's unclear how well such guys are at consistently maintaining good performance in defensive minutes.  Look at Gordon - in 12-13 his D possession #s are terrific.  In 11-12 they're poor.  In 10-11 they're good in easier minutes.  In 09-10 they're horrible.  The odds of Gordon being another Belanger seem awfully high.

A few others were added of course.  Jesse Joensuu was an enigma on the Isles, and now he's one for the Oilers, for example. 

CONCLUSION:

The Oilers this offseason, with the exception of Ference (and the signing of LaBarbara to be the backup, an excellent signing), went for the unknown, taking a # of gambles on D improvements from the KHL, but didn't improve really at forward.  When your additions are like this however, your range of outcomes is wide.  A bottom 5 finish is very plausible.  So is a middle of the pack finish. 

But I've forgotten one element.  Coaching. 

Willis argues, as have several other Oilers fans, that Kreuger was the major problem and the reason for the step back last year, and that his failure to line match and flawed strategies were responsible for the gigantic step back.  I'm far from convinced. 

For one, line matching competition wise doesn't seem to be very influential on results - coaches simply have little ability to get players out against specific opponents (particularly on the road), resulting in each player playing similar loads of competition.  This is why the gaps in QOC metrics are at most 4 shots, and usually within 2 if not lower.  That's not much.  ZONE matching does seem more effective, but I've seen nothing to suggest Eakins is going to make this extreme.  So perhaps we get a marginal benefit here by getting line 1 out in the O Zone more often (on the other hand, this isn't going to help the extremely poor other lines). 

For another the suggestion that the players' step back was due to Krueger seems to imply that a coach's impact is HUGE: able to take a just sub average team into a bottom 3 team.  This seems incredibly unlikely, and would make certain coaches dramatically underpaid. 

Yes both C&B and Dellow have used video to show that certain plays seem particularly poorly designed, or that the strategies used were unconventional (C&B at one point argued there was a defensive strategy by Krueger which no one in the NHL used).  On the other hand, we don't have data showing that such poor looking plays were the results of the coach's strategies, as opposed to the players' being poor.  Or that certain plays were destined to get poor results.  We simply do NOT have the data to conclude that Krueger's coaching was to blame, despite the results.    (How to separate coaching from roster talent is a difficult one). 

Finally, as hinted previously, like the players, it's not a guaranty that Eakins is a good coach at all, or even better than Krueger.  Yes the Marlies had good results under him.  That's not a guaranty of anyone being a good coach.  Kreuger may not have been as high profile for instance, but he was once considered a pretty good coach in waiting. 

Jonathan Willis stated on twitter that, and i'm quoting to avoid misquoting here:
Shot differential collapsed in 2012-13 despite an improved roster, and it seems a safe bet to me that Dallas Eakins will reverse that trend.
The problem is that the only proof of 12-13 being an improved roster is well, the presupposition that those players were better.  That seems unclear at best - they certainly performed a lot worse!  Why is that on the coach and not the players!   Moreover, Eakins is not a safe bet - he's an unknown.  Like the rest of the Oilers moves this offseason. 

The Oilers certainly could be in the middle of the pack next year.  Or they could be in the bottom 5 yet again.  Nothing is certain about this team, and thus a prediction of utter atrocity is not exactly unreasonable. 

No comments:

Post a Comment