Source: Winnipeg Free Press
In my last post, I focused on the longest losing streaks each coach has had behind the bench this year. The purpose was to find out why Nelson’s losing streak was somewhat dismissed, while Eakins losing streak received a far greater backlash. Eakins’ 11-game skid had some decent underlying numbers at 5-on-5, but had some sketchy goaltending and a weak powerplay and failed to have any positive results. Nelson just finished off a 7-game streak, where they were absolutely lights-out on the powerplay, but had some troubling underlying stats at 5-on-5. My guess is that the success of the powerplay, and the point production of the young guns like Eberle, RNH and Yakupov gave the perception that Nelson was doing a better job.
I received a comment that suggested that individual players have benefited from the coaching change and their production has been better under Nelson. At first glance, it appears to be true. Eberle and RNH in particular have been outstanding over the past few weeks, with a large chunk of their points coming on the powerplay. Make no mistake, Nelson has done very well with the man advantage, something he was known for at the AHL level, and deserves full credit for its success. My take, however, is that 5-on-5 play is much more important, so I decided to take a look at the productivity of players at even strength under the two different coaches.
Please note, I exclude the five games that MacTavish was behind the bench in all of my comparisons involving Nelson. I’ve included in the list below the players who played under both coaches. (Source: War on Ice)
| Eakins | Nelson | ||||||||
| Name |
Pos. |
Games |
P60 | CF% | TOI/Gm | Games | P60 | CF% | TOI/Gm |
| Ryan.Nugent-Hopkins | C | 29 | 1.97 | 51.83 | 15.79 | 37 | 1.73 | 50.57 | 15.00 |
| Nail.Yakupov | LR | 31 | 0.93 | 48.90 | 12.47 | 36 | 1.51 | 43.48 | 13.26 |
| Taylor.Hall | L | 25 | 1.65 | 53.15 | 16.04 | 14 | 2.31 | 50.38 | 14.86 |
| Jordan.Eberle | R | 30 | 1.82 | 53.97 | 15.41 | 37 | 2.24 | 50.29 | 15.23 |
| Benoit.Pouliot | L | 20 | 2.19 | 51.47 | 10.95 | 29 | 1.44 | 51.23 | 12.96 |
| Boyd.Gordon | C | 27 | 1.11 | 46.68 | 10.02 | 37 | 0.63 | 42.02 | 10.33 |
| Luke.Gazdic | L | 10 | 0.00 | 45.37 | 7.03 | 19 | 1.33 | 46.30 | 7.13 |
| Matt.Hendricks | LR | 27 | 0.68 | 47.54 | 9.80 | 35 | 0.87 | 43.63 | 11.80 |
| Leon.Draisaitl | CL | 31 | 0.88 | 52.76 | 11.05 | 2 | 5.73 | 56.76 | 10.47 |
| Tyler.Pitlick | RC | 7 | 0.68 | 51.54 | 12.57 | 2 | 0.00 | 30.43 | 10.70 |
| Iiro.Pakarinen | RL | 5 | 1.65 | 50.00 | 7.25 | 12 | 0.00 | 47.17 | 10.79 |
| David.Perron | RL | 31 | 1.73 | 52.01 | 13.43 | 2 | 2.49 | 57.14 | 12.07 |
| Teddy.Purcell | RL | 31 | 1.23 | 54.70 | 12.56 | 37 | 0.85 | 46.23 | 13.37 |
| Justin.Schultz | D | 30 | 0.92 | 50.99 | 17.32 | 37 | 0.44 | 50.58 | 18.59 |
| Jeff.Petry | D | 30 | 0.33 | 53.00 | 17.97 | 24 | 0.74 | 43.09 | 16.93 |
| Keith.Aulie | D | 12 | 0.00 | 51.60 | 12.59 | 10 | 0.45 | 36.86 | 13.28 |
| Mark.Fayne | D | 31 | 0.53 | 49.49 | 14.52 | 37 | 0.22 | 44.19 | 15.05 |
| Andrew.Ference | D | 28 | 0.24 | 48.45 | 18.04 | 37 | 0.94 | 41.98 | 15.46 |
| Oscar.Klefbom | D | 10 | 0.33 | 53.52 | 18.34 | 37 | 0.97 | 50.91 | 18.40 |
| Martin.Marincin | D | 12 | 0.34 | 51.29 | 14.56 | 20 | 0.00 | 49.00 | 16.07 |
| Nikita.Nikitin | D | 22 | 0.53 | 50.57 | 15.46 | 15 | 0.27 | 45.83 | 14.83 |
Looking at the point production (points per 60), the two players that saw an increase of their 5-on-5 production under Nelson are Eberle and Yakupov. Hall’s numbers increase, but that may have been because he was banged up early in the season. What’s surprising is the decrease in productivity for players like RNH, Pouliot, Gordon, Purcell and even Schultz. What’s troubling is the decrease in the possession numbers (Corsi For %) across the board. We are seeing that the team does struggle with possession in all score situations (whether they’re trailing, leading or the game is tied) under Nelson, while Eakins had something figured out when it comes to 5-on-5 play.
And here are the players who were coached by one and not the other. Included are guys like Lander, Roy and Klinkhammer who have all done relatively well with Nelson behind the bench, but still struggle possession wise.
| Eakins | Nelson | ||||||||
| Name |
pos |
Gm | P60 | CF% | TOI/Gm | Gm | P60 | CF% |
TOI/Gm |
| Anton.Lander | C | – | – | – | – | 28 | 1.19 | 48.29 | 10.82 |
| Derek.Roy | C | – | – | – | – | 37 | 1.67 | 45.93 | 13.63 |
| Ryan.Hamilton | L | – | – | – | – | 16 | 0.32 | 40.99 | 11.54 |
| Rob.Klinkhammer | L | – | – | – | – | 32 | 0.48 | 46.22 | 11.67 |
| Matt.Fraser | LR | – | – | – | – | 28 | 1.21 | 41.71 | 10.62 |
| Drew.Miller | RC | – | – | – | – | 3 | 0.00 | 53.25 | 11.19 |
| Jordan.Oesterle | D | – | – | – | – | 6 | 0.75 | 49.64 | 13.42 |
| Will.Acton | C | 3 | 0.00 | 44.68 | 9.22 | – | – | – | – |
| Bogdan.Yakimov | C | 1 | 0.00 | 61.54 | 10.05 | – | – | – | – |
| Mark.Arcobello | CR | 31 | 0.89 | 49.86 | 13.10 | – | – | – | – |
| Steven.Pinizzotto | R | 13 | 1.20 | 43.51 | 7.68 | – | – | – | – |
| Jesse.Joensuu | RL | 20 | 0.63 | 45.93 | 9.57 | – | – | – | – |
| Brandon.Davidson | D | 3 | 0.00 | 42.86 | 10.52 | – | – | – | – |
| Darnell.Nurse | D | 2 | 0.00 | 56.36 | 15.21 | – | – | – | – |
| Brad.Hunt | D | 6 | 0.00 | 50.00 | 15.82 | – | – | – | – |
What’s become apparent is that individuals are producing more points, but it’s due in large part to the successful powerplay. Stripping the powerplay away, however, gives us a better assessment on how the team is doing for the majority of the game. In this case, the production has increased for some and decreased for others. When it comes to possession, which is a key indicator of team success, the entire team is struggling mightily.
Both Eakins and Nelson are qualified NHL coaches, having found success at the AHL level, and will likely be employed in some capacity next season in the NHL. Nelson should definitely be considered for the OIlers head coaching position next season along with other experienced coaches available this summer. The problem is that the Oilers are struggling to assemble an NHL caliber roster, and until they do, it really won’t matter who the coach is next season.
The SuperFan 
Hi
Well you have forgotten something in all your stats, eakins had 2 pre-seasons, countless practices and a 113 games to do his best with what he had and most of the players went sideways under his watch. On top of that the special teams became the worst in the league. Nelson has had 37 games to change the team direction and moral. The best thing to change the quickest would be the special teams and come 2015 pre-season we will see what he can do if he is allowed to coach next season.
eakins missed injured players 33 gmes to nelson’s 82 missed games to injured players.
eakins had petry and perron 62 games where as nelson only had perron and petry for 28 games
eakins for roy 0 games and nelson 37 game. i would take either petry or perron over roy anytime.
If you watched the practices they have become nhl caliber and if you watch the oilers in the first period they have be alot better and ready.
so stats can easily be bent to whatever direction the person using them decides.
possesion numbers under eakins are a mirage.
shots or should i say scoring chances are also a mirage under eakins. Look where the shots where taken from and you will also see a huge difference.
thx vern
Those are all fair points Vern. Stats won’t tell you the whole story, but it can shed some light on the key issues impacting this team. No question, the team is built poorly, and has been for years. But having two sets of coaches gives you two different game plans and tactics, which are typically reflected in the stats and metrics derived from analytics.
This post is an extension of my last post, which questioned some of the positive support Nelson has received. To me, the team is still bad, so I just wanted to see if there were any patterns when it came to player production. Based on the comments and feedback I’ve received, I’m hoping to continue exploring this case,
vern i couldnt agree with u more!! even looking back to ralph era pp was improving then plummeted to basement under eakins. ive said all along the way theres no reason for this team to have a bad pp specially with the roster make up. its built as a strong pp team. way to much structure under eakins and not nearly enough creativity. for three yrs theres absolutely no net presence in either end of ice and no fault to any coach but this team is still a perimeter hockey club. its getting better though and some roster changes would help. perron helped with this but its to bad he and taylor hall dont get along. should have been hall to go not perron!! credit nelson though for using anton lander infront of net on pp. go figure a guy that couldnt stay in line up before is not only succeeding now but has an important role with club!
If you’re interested, Neal Livingston of Tend the Farm, who did amazing work covering the Barons, has an insightful post on Nelson’s powerplay system in OKC: http://www.tendthefarm.com/things-you-get-from-a-todd-nelson-coached-team/
cool thanks i will check it out!!
Pingback: Hockey News: Chris Tanev Contract; Devan Dubnyk for the Hart?
I think if your “underlying numbers” can’t tell you if a team was playing like garbage, they’re not worth much at all. Eakins knew how to jig the advanced stats by having the team play a certain way. A non-effective, terrible to watch, brand of hockey. Eberle said as much recently. He was coaching them to get better numbers, but not honoring the actual hockey playing that it takes for those numbers to mean anything real.
Nelson is coaching this team to play hockey. And the stats you see “look worse”, but they are what they are. They were arrived at honestly. At least you know what needs to be fixed. His team is playing better. They’re competitive most nights. They are nowhere near where they need to be, however. There is no way they could be with this mix of players. That’s on management. And I bet MacT doesn’t like that Nelson’s performance shines the light on what is really wrong with this team: The Old Boys Club.
We shall see how that plays out soon enough. But the suggestion that Eakins had this team playing better, in any meaningful way, suggests that you’re taking the data at face value, and ignoring the actual hockey being played. This team was out of the playoff race by Halloween for the second straight year. Instead of saying “the Oilers were playing better 5v5”, why not say “How come the stats can’t capture how bad this team is?”.
I’m not sure if Eakins was rigging any stats. If you’re interested Matt Henderson has an excellent write up on that theory here: http://oilersnation.com/2015/3/16/gaming-corsi-on-the-power-play
As for publishing stats, I’m very mindful of the fact that they could be misleading in some ways. What I’m trying to do is start with a question, find some data that can confirm or deny it, and then use that data to continue exploring additional questions or ideas. For example, possession was better under Eakins, but I’m curious as to how he got those results. I’m in no way defending Eakins and his results…he’s obviously made mistakes. Nelson, to me, is a good coach, but has some deficiencies, so I’m trying to find where those weaknesses are. Using data alone won’t give you all of the answers, but it can at least point us in the right direction.
David Staples pointed out recently that analyses of the Oilers have missed out in accounting for quality of competition. I tracked the Oilers possession (& all teams) on a weekly basis and found that their possession numbers dropped dramatically in December. (This was during the long-ass losing streak.) Trusting Staples on this, that’s when the Oilers played more Western than Eastern teams. I computed three metrics comparing the 2 divisions in the West vs. Eastern conference. Both Western divisions have stronger Corsi & Weighted-Shot differentials than the East.
I think what I’ll do is compare the Oilers possession numbers vs. Eastern conference, Central division, & Pacific division. It could be that quality of competition–let’s call it Staples’ hypothesis–accounts for more of the difference in 5v5 than the coaches. Now it may not be cut & dry for it could be a bit of both. But I think it would valuable to do the analysis. Will get back to you when I’m done.
In the meantime, here’s my calculations of the Oilers quality of competition last season and this season.
http://bit.ly/1y7F0Gr
And here I document in detail how much Yakupov has improved with Roy at evens & on the PP under Nelson. For example, with Roy at 5v5, Yakupov’s shot generation & point production are comparable to any 2L winger. On the PP, Yakupov’s shot attempt rate is top 30 in the league.
http://bit.ly/18VijPl
That would be awesome. It’s almost too easy to do a Eakins vs Nelson comparison over and over, so I’ve been mulling over how else we can break out player performance. I know the perception is that players are better now, but having the kind of data you’re proposing could test that.
Your breakdown of Yakupov’s performance is excellent stuff.