Todd Nelson’s Edmonton Oilers

Todd Nelson7

The Edmonton Oilers are now 9-9-2 under Todd Nelson, and they’ve carved that record out on merit. They’ve done it without David Perron (18 games), Taylor Hall (eight games) and Benoit Pouliot (seven games) for long stretches.

How is this happening?

All Situations

Dallas Eakins 19

  • All-situations, score-adjusted Corsi under Dallas Eakins: 50.2%
  • All-situations, score-adjusted Corsi under Todd Nelson: 49.8%

(Note: I’ve nixed the five games the Oilers played under Nelson/general manager Craig MacTavish because I don’t for a moment believe that the interim coach of record was calling the shots while his direct superior, a career coach, listened meekly on the bench beside him. Mileage may vary, but to me it does not seem reasonable to ascribe those five games to Nelson.)

Those numbers are open to all sorts of different interpretations, but I see them as an endorsement of the Oilers’ interim head coach. Eakins doesn’t get a lot of credit for the fairly substantial improvement in Edmonton’s underlying numbers this season because the team didn’t win under his watch, but that improvement was vital; no team has sustainable success if its pinned in its own end of the rink for most of the night. Whether the Oilers’ brutal shooting/save percentage was Eakins fault or not is debatable and continues to be debated. All I’ll say is that I’ve seen some really good coaches (Mike Babcock in Detroit, Darryl Sutter in L.A.) go long stretches with great Corsi numbers and garbage percentages so it can happen to even the best. That’s not the purpose of this piece, though. We’ll get a definitive answer on Eakins at his next coaching stop.

For now those numbers are included as a baseline for the purpose of examining what Nelson has done. With inferior personnel, his team is basically a rounding error away in terms of Corsi (in all situations, and adjusting for score), and of course the team’s record is a lot better.

There’s been a lot of talk of intangibles, and from spending 2013-14 in Oklahoma City I can understand why; I know the atmosphere that Nelson endeavours to create and I know the amount of respect his players have for him. When we see an improved record and couple it with those things it’s easy to say that the changes Nelson has wrought are intangible and can’t be tracked by the numbers. Maybe that’s true, but I haven’t really seen anyone try, so let’s try now.

Even Strength


  • Even-strength goal differential under Dallas Eakins: +47/-75 (-0.90 per game)
  • Even-strength goal differential under Todd Nelson: +32/-45 (-0.65 per game)

Under Eakins’ watch, the Oilers had a 50.9% Corsi, a 6.5 shooting percentage and a 0.902 save percentage at five-on-five (goal differential above is both five-on-five and four-on-four). Under Nelson’s watch, the Corsi is 46.9%, the shooting percentage 7.4 percent and the save percentage 0.909. We should note that Nelson has had inferior personnel, and the Corsi numbers get quite a bit closer when we adjust for the fact that Eakins’ Oilers were trailing more than Nelson’s Oilers.

We’ve seen a reasonably big shooting percentage bump at five-on-five. There’s a tendency to say that Nelson’s systems are so much better that it’s seen the Oilers score on a higher percentage of their chances. Before we endorse that point of view, let’s try one thing:

  • Oilers five-on-five shooting percentage, minus Eberle, under Eakins: 42 goals, 648 shots, 6.5 percent
  • Oilers five-on-five shooting percentage, minus Eberle, under Nelson: 23 goals, 381 shots, 6.0 percent


Jordan Eberle has scored eight times on 38 even-strength shots under Nelson; he scored four times on 57 shots under Eakins. Remove him from the equation, and suddenly the Oilers under Nelson are the same wretched shooting percentage team they were under Eakins (actually, a little worse).

When the gap is so small that we can reduce it to one player, I’m not inclined to say that there’s a massive team change. There’s a massive Eberle change, for sure, but if we take his goals out of the equation the Oilers actually had a slightly better even-strength goal differential under Eakins than under Nelson. Add to that the wretched save percentage under both coaches and I think it’s fair to say that coaching probably isn’t driving the shooting/save percentage results at evens.

I’ve noted previously that Nelson has inferior personnel and has played with the lead more often; that’s true. On balance I’m inclined to say that at even-strength on the whole the majority of the team has taken a small step back, perhaps owing to coaching and perhaps owing to personnel. It’s been compensated for because Eberle has been three times as likely to score on any given shot under Nelson than he was under Eakins. Maybe that’s coaching, maybe that’s injury, maybe it’s personal circumstance; we really don’t know.

What we can say is that, based on goal differential, the Oilers at evens were a bad team both under Eakins and under Nelson because their shooting percentage and save percentage are terrible.

Special Teams


  • Special teams goal differential under Dallas Eakins: +14/-23 (-0.29 per game)
  • Special teams goal differential under Todd Nelson: +12/-10 (+0.10 per game)

What we have seen under Nelson is massive improvements on both special teams, both in terms of goal differential and underlying numbers.

On the power play, the Oilers averaged 97.1 shot attempts/hour under Eakins; that ranked them 18th in the NHL over that span. Under Nelson, the Oilers have averaged an incredible 111.6 shot attempts/hour, the fourth best number in the NHL over that span. Nelson’s power play is +10/-0 in terms of goal differential over 20 games; Eakins was +12/-5 over 31. The improvement is real, and it’s spectacular.

Ditto for the penalty kill. Under Eakins, the Oilers penalty kill allowed 98.0 shot attempts/hour, ranking them 15th in the NHL over that span. Under Nelson, the penalty kill has allowed 86.5 shot attempts/hour, the third-best number in the entire league.

We don’t need to parse these numbers. Almost overnight, Edmonton has improved on both special teams to a massive degree. Personally, I credit that to coaching, but because it’s confined to special teams I’m tempted to say that if we dig into it we’ll find tactical changes that mattered. That’s something we’ll do in the coming weeks; this piece is long enough as it is.

In Total

Craig MacTavish15

When we actually start digging, we don’t have to just shrug our shoulders and say that Nelson has the team buying in, that he hasn’t alienated the young players, that Eakins lost the room and he’s won it back.

When we actually start digging, we see that the Oilers (minus Eberle) are actually a somewhat lesser team under the new coach at even-strength, but that between Eberle’s improvement and massive changes on special teams they’re a better team overall.

If I were general manager Craig MacTavish and I was forced to make decisions today, I’d do the following:

  • Take the interim tag off Todd Nelson. He’s made big upgrades at special teams and at even-strength he’s had incredible challenges which go a long way toward explaining the step back in the team’s underlying numbers.
  • Upgrade personnel. Goaltending has been a problem, the blue line has been a problem and shooting percentage has been a problem. Those problems have not gone away under Nelson at evens; under two coaches this team has underperformed percentage-wise and *just maybe* it’s because of the obvious massive holes on the blue line and at centre. Fix the roster. Don’t blame the coach.

All advanced statistics taken from the wonderful, with my profound thanks.


  • Zarny

    If the stats don’t confirm my pre-conceived notions, then the stats are useless.

    If the stats confirm what I already believed, then they are a good tool that I can use to justify my pre-conceived notions.

  • Jayz

    It’s amusing to see so many people (including JW) trying to sound smart by using advanced stats for their articles/arguments.

    Just stop.

    You are not Dellow. Yes we all miss his articles. I hate the oilers for signing him because I think we all could use his insight into what is happening not only with our team but with that team down the road.

    You can’t remove Eberle from the equation. Eakins was bad mojo. No other way to say it. Also there is a direct correlation to the age of teams as well as how long the “core” group has been together and possession stats.

    This isn’t rocket science. More mature (25-29 age group) teams that haven’t had much turn over ( Detroit, La ect ) have very strong possession numbers. Teams that are still quite young especially in the top 6 F group ( oil, flames, av’s) and have plenty of new faces every year have poor corsi.

    The most important thing I’ve noticed is the positive vibe that just seems to beem out of our most embattled players.
    That is why I trust in Todd.

  • Rusty Patenaude

    This article clearly demonstrates the problem with advanced stats and that is…nothing. There is nothing at all wrong in using them as ONE tool for understanding what is contributing to the results on the ice. Corsis are not the best, let alone only method for evaluating what happens on the ice, but they have some validity.

    It is clear that the problem with Corsi stats is that some people overestimate their value. It seems that these “stats guys” can’t actually watch a team play well and believe their eyes, they have to crunch the numbers first. That is foolish. This is not a computer simulation of hockey, this is hockey played by real people on ice we are talking about. There are so many more variables than these advanced stat systems can analyze that all any of these stats can be is ONE more way to look at the game. But it is ignorant to use Corsi numbers to trump an experienced and unbiased eye.

    The Oilers are playing better under Nelson. If you have watched the games you will know that without checking the Corsi numbers. One important area of improvement is that the team is more mentally resilient now. They can give up a lead in the third period and shake it off and compete confidently in the overtime. They can be beaten like a rented mule in Toronto without triggering an 11 game losing skid. They can come back against a clearly superior team on the road in the third period. That resiliency is not the result of myopic fixation on advanced stats. That is the result of a group of players who have bought in to Nelson’s vision of the game.

    If you know anything about hockey you can see how much confidence Justin Schultz and Oscar Klefbom are playing with now. You can see how Yakupov trusts himself, how Eberle has regained his swagger, how Lander is no longer scared to make a mistake. Those very obvious, very observable changes on the ice have just as much merit as stats.

    Hockey does not lend itself to advanced stats the same way baseball does. Hockey is a fluid, spontaneous game where the two teams interact on the same surface with the same team goals. Baseball can be broken down more effectively because in many ways it is 9 individuals competing in 9 different disciplines. Hockey advanced stats are imperfect at best and should never be used as the only way to judge performance.

  • Zarny

    The Eakins picture literally made me cringe at first glance. I didn’t hate Eakins as the coach until right before he was fired, but man that picture takes me back to a bad place.

  • Nelson has the team playing a better game that allows better quality chances, and one that gives us multiple chances to retrieve the puck when we turn it over.

    The team is faster, sends more traffic to the net with more players stopping in front of the net rather than skating through the slot, and employ a more simpler brand of goal creation. We saw something similar under Krueger and this was lost under Eakins.

    Part of this is varied zone entries. The Oilers no longer try to carry the puck in at all times (which certain players were great at), they now purposefully lose possession (which some stats people disliked) by dumping it in, winning a battle and then generating some offensive zone time. The lack of this made nights easy for opponent Dmen who rarely got hit or fatigued from retrieving the puck. It’s something we are doing now and it leads to poor decision making and poor execution on the opponent D’s part. If we lose the battle deep in the zone, worst case scenario is that it results in the other team moving the puck 200ft to our net. It gives us a chance to forecheck, it gives us a chance to cause a turnover in the neutral zone, and it gives our D a chance to close the gap when the opponents are coming in with less speed. Sure, the Oilers still carry it in on occassion, but when it makes sense. How many times have we seen Hall or Yak come into the zone with 4 opponents waiting for them? How many times have we seen them hold onto the puck despite that? There was too much stress on retaining possession coming into the zone.

    So I don’t see things as “intangible” I see it as playing a great simple hockey style. Which is exactly how the Flames are achieving their success, they employ an offensive game that could have 12 monkeys for forwards and still have a good amount of success. They knew they didn’t have the offense to create goals (as admitted in numerous articles by management), so they simplified the game to capitalize on rebounds (which creates a sky high shooting percentage), capitalize on sending the puck to the net without the intention of scoring.

    When it comes to advance stats measuring the difference between Nelson and Eakins, I’d argue that we don’t have enough advance stats yet to make a comparison. There are realities to the game that advance stats do not yet cover. Is there a stat that covers how many players are on either side of the puck when a turnover occurs? Is there a stat that covers how many feet the puck is from your net when a turnover occurs? Is there a stat that covers rebounds generated? We know already that 50% of rebound shots are goals. Why not track what causes this? Again, many of the Flames players are not shooting to score, they are shooting to create a rebound. Call it “garbage goal”, sure, but in the playoffs you’d take goals any way they come.

    TDLR: Nelson has improved the team in many facets, areas that advanced stats cannot measure as of yet as advanced stats are still in their infancy. 😛

  • Anton CP

    Considering about how many times that Oilers have lost a close game before Nelson took over (18 out of 29 losses with 2 goals or less, 5 out of 7 wins are with 2 goals or less) that special team makes a huge difference. Some of those losses can be wins if their special teams can execute better. The most horrendous stats from last year was shorthand goal against, Oilers were league worst. Eakins need to be fired a lot sooner.

  • toprightcorner

    I thought Ramsey was in charge of the special teams under Eakins? Did Nelson take those duties over himself since the change? If this is the case is it lack of Eakins taking charge to change special teams or that Ramsey having been removed from those responsibilities was a genius move by Nelson.

    In most cases, the head coach is not in charge of the special teams, it is usually an associate or assistant in charge though the head coach input probably varies from team to team.

    • Anton CP

      The lockout year that Oilers have the same assistant coaches before Eakins took over, under Krueger that Oilers special teams were decent. Maybe it is the other way around that Eakins was also getting involved with special teams.

  • toprightcorner

    If the Oilers are playing the same or worse statistically then there doesn’t seem to be any substantial reason that Nelson has earned the head coaching title and if a more experienced coach is available that should be one of the first moves MacT makes in the off season.

    Better results with the same or worse statistical play could simply be an anomaly of a honey-moon period for the team.

    Sure, without an under performing Hall or Pouliot that missed many games under Eakins hurts the team but an actual NHL 2C balances that out in my opinion.

    Player respect is the main thing that Nelson had over Eakins as a coach and that itself goes a long way but I don’t see Nelson as the coaching answer for this team. Give me a head coach with lots of NHL experience please and then something can be done with the player additions over the off season to achieve the best results

  • MacT's Neglected Helmet

    A good, thought provoking column. Thank you JW. Some thoughts. The special teams improvement, especially the power play improvement is really a return, with significant personnel changes, to the successful power play numbers when Krueger was associate coach and then head coach. I have no idea how you apportion the blame for the PP failures between Eakins and Ramsey.

    The Eberle stats are interesting, but I presume there are some players with weaker numbers under Nelson, although not as negative as Eberle’s are positive, whose subtraction would favour the job Nelson is doing. I note that Eberle is attributing his vastly improved play to not obsessing about mistakes as much as he was. It certainly appear that the more up tempo game and Nelson’s encouragement of Eberle’s change in attitude account for his recent success, something that is difficult to measure.

    I have a question and I should add that I was in Asia until Dec. 7th so missed most of this season under Eakins. I know that he is a deep believer in analytics, as am I, and that he values putting more shots on goal, as I believe he has said. The question is was he pushing too hard for more shots and mainly getting more low percentage shots, which would account for the lower shooting percentage under him. One would have to grade the shots qualitatively, an onerous task, to determine if that is true. Even counting scoring chances would only partially answer that question.

    My fear is that recent comments by MacT, and to a lesser extent Lowe, reflect a reluctance to give Nelson credit for the Oilers far better record since he took over. MacT insists he has given Nelson a better roster, but it is hard to argue that Roy, Fraser and Klinkhammer are more than a slight improvement over Perron and Arcobello, if an improvement at all. I am concerned that they have another head coach in mind or at least want to take a shot at a bigger name. Given the near certain departure of Petry and maybe others in an attempt to gather more draft picks, it is going to be very difficult for Nelson to maintain the team’s point-a-game pace.

  • JW does this corsi stat take into account sustained offensive zone pressure, quality of shots, and the pressure applied by the forwards in the o-zone and the defenceman in the d-zone. The Oilers are more aggressive now, that can’t be questioned and it’s no longer one shot and done and they are crashing the net more consistently. I like analytics for the fact that it gives tangible info but basing your theories completely on analytics makes for an incomplete thesis/statement.

  • Pouzar99 wrote:

    A good, thought provoking column. Thank you JW. Some thoughts. The special teams improvement, especially the power play improvement is really a return, with significant personnel changes, to the successful power play numbers when Krueger was associate coach and then head coach. I have no idea how you apportion the blame for the PP failures between Eakins and Ramsey.

    Not actually.

    Ralph Krueger’s power play was built not on generating a lot of shots but rather on a crazy-high percentage of those shots going in. Todd Nelson’s power play is built on generating crazy numbers of shots.

    I know which of the two I’d bet on sustaining their success.

  • Dwayne Roloson 35

    Yeah Nelson has a worse team on paper. Advanced stats(which are overrated IMO) are very similar but at the end of the day, nelson is getting more out of our top guys/deploying them in a way that gives them success.

    Ebs has looked great and klefbom has quietley put up 2G-6A-8pts-+3 in his last 10 games. the bottom 6 is producing at a decent level too.

    we are better and we look competitive most nights even without hall and poo.

  • MacT's Neglected Helmet

    I think everyone agrees that the Oilers look way better under Nelson, the question Willis is trying to answer is “is this improvement for real? or are we getting *some* of the puck luck or PDO regression we expected to see after a horrible start to the season.

    score adjusted Corsi is absolutely the best way to predict future success but we’re not close to a point with Nelson where we have a big enough sample size to use corsi properly

  • nugeformayor

    Love this guy! Eakins on the other hand would have got us baby mcjesus for sure,too bad but I cant stand to even see his pic makes me cringe. We need to donate to Todd to get him some new suits and get someone that knows how to pic ties.

  • BubbaZanetti

    Quick question here, Whether Nelson is coach or someone new at the start of next season…if the team still sucks, do they get rid of that coach too? Whomever the coach is, should that coach get the full season. And the players should understand this. Sooner or later, these players are going to have to be held accountable, and I’m sure as hell not talking about the bottom six.
    Just curious as it seems it will be more or less the same line-up next year.