The Case Against Kris Russell Part 1

Alright. I’ve been overly damned cheery for weeks on this site. I think it’s a pretty much non-stop stream of happiness since the season began, but it’s becoming increasingly clear that the Oilers are on the verge of doing something incredibly unadvisable: Extending Kris Russell. It makes so little sense for so many reasons, and yet here we are. Let’s have a conversation about this.

This is part 1 of 2 articles that will talk about the prospects of extending Kris Russell. Today the focus is on the player himself and where he fits on the club. Tomorrow the second article will run discussing the various reason why extending Russell is a truly awful idea. 

Let’s start with the cold hard truth. Kris Russell is Edmonton’s worst defender.

Advertisement - Continue Reading Below

I’ve already lost a segment of readers and I accept that. For some people, saying the obvious is offensive and there are others that just can’t acknowledge the truth. It’s understandable that the statement may seem shocking, although it’s dangerous to come to any other conclusion than the one stated above.

I know that Kris Russell skates really well. I know that he blocks lots of shots and is always willing to sacrifice his body. There are times when he uses those skills and good things happen on the ice as a direct result. It’s fantastic that he can skate with the puck like he did during a 4v4 recently and create an opportunity for the team.

The problem with Russell, and it’s what makes him so thoroughly dangerous to the Oilers, is that the things he does well are really noticeable but what he does poorly often goes by invisibly. For example, he will block two shots in a shift and retrieve the puck in the corner, take some strides, then move the puck out of the zone. The eyes catch all of that really well. What they are not generally trained to see is that Russell’s side was targeted by the attacking team, he gave them unobstructed entry to the zone by giving a 10-foot gap, and after he threw the puck up the ice it was immediately picked up by the attacking team and brought back in again.

This imaginary sequence of events I’ve used as an example happens very frequently with Russell and it’s bearing out in all of the non-traditional stats we can get out hands on. Once you start looking for entries allowed and successful transitions back up to offense, there is one Oiler defender who starts standing out in the least positive way possible: Kris Russell. It’s then impossible to unsee it.

Advertisement - Continue Reading Below

Every player has a Risk/Reward benefit. Every single one. It’s all about what you create minus what you give up. For a player like Dustin Byfuglien he’s constantly derided for weaknesses in his own zone, but he more than makes up for them with his impact in the offensive zone. Someone like Adam Larsson we recognize has limits in the offensive zone but he’s stronger defensively.

Kris Russell, however, is neither strong offensively nor stellar defensively. This is a bad combination, but it’s actually his impact on offense that should be most alarming. Of course, it’s the WAY in which he chooses to defend that negatively impacts his ability to transition back up to offense. No matter which way you look at it, Kris Russell is a ticking time-bomb for the team.

Edmonton has had seven defensemen play at least 200 minutes 5v5 this season. Here is where Russell ranks of those seven in some key statistical categories.

Corsi For per 60: 7th (50.03 CF/60)

Corsi Against per 60: 7th (56.36 CA/60)

Advertisement - Continue Reading Below

Corsi For Percentage: 7th (47.0%)

Fenwick For per 60: 7th (38.34 FF/60)

Fenwick Against per 60: 6th (39.87 FA/60) *This should be one of the stats he performs best in because he’s a shot blocker. Fenwick counts only unblocked shot attempts.*

Fenwick For Percentage: 7th (49.0%)

Shots For per 60: 7th (28.18 SF/60)

Shots Against per 60: 3rd (28.37 SA/60) *Good Job!*

Advertisement - Continue Reading Below

Shots For Percentage: 7th (49.8%)

Scoring Chances For per 60: 7th (18.96 SCF/60)

Scoring Chances Against per 60: 3rd (19.98 SCA/60)

Scoring Chance For Percentage: 5th (48.69%)

Offensive Zone Faceoff Percentage: 4th (33.7%)

Defensive Zone Faceoff percentage: 3rd (32.7%)

On-Ice Shooting Percentage: 2nd (10.2%)

Save Percentage: 2nd (94.59%)

PDO: 2nd (104.8)


So that’s a pretty big wall of information. Here are the key takeaways for me: Russell is last or next to last in every single shot metric we have available. In Scoring Chances against he’s doing an OK job of staying in the middle of the pack, but the team also generates the lowest number of chances for when he’s on the ice. He’s also getting a pretty even distribution of zone starts so I can’t blame his low numbers on bizarre usage.

At the same time as all of that, Kris Russell has a stellar PDO. Our eyes should be in love with Kris Russell and our brains should associate him with success because he is riding a wave of percentages that are covering up his deficiencies. In the four years prior to joining the Oilers, Kris Russell’s on-ice save percentage was 91.74%. I think it’s highly unlikely that at 29 years of age Kris Russel has figured out how to make his goalie jump to 94.59%. The inventor of PDO himself (Brian King) says that the stat was created because the Oilers were extending players with high numbers and dealing away for pennies the ones with low numbers. Teams mistake great PDO for great talent then make awful decisions. 

There are things that no metric can tell you, like how a player is in the dressing room, what kind of character they have, etc. All of those things are inputs. They are tools that the player has available to help him/her get the job done. The metrics help inform us as to how well the job is getting done. In Russell’s case, we have a lot of information that says the team struggles to move the puck in the right direction when he’s on the ice. 

Where does he fit on the blue line? If you had to keep one player moving forward, Russell or _____, how many Oiler defenders do you go through before you keep Russell? He makes it through the entire list for me.

  • Prometheus


    Thank you for being a pillar of light in a sea of chaos. Obviously it’s your choice, but I’d suggest you not waste your breath. Sometimes individuals refuse to have an open mind.

    The beauty of science (statistics included) is that if new evidence appears, you can adjust your beliefs. Numbers never lie – but what they say is not always what is heard.

    • ¯\_(ツ)_/¯

      Numbers never lie – but what they say is not always what is heard.

      You might have missed that recent election in the US where the “numbers” in the form of statistical analysis from polling data were way off in polls across a wide group of battleground states. Numbers don’t lie but they don’t tell the truth either.

      We interpret them, we don’t use them to promote self-evident truths. And sometimes we don’t have enough to extrapolate from a data set to a meaningful or representative conclusion. I’d say that’s what’s happening here.

      In the Russell example, it’s worth asking: what’s the meaning or value (in games won or any value that can meaningfully be applied) of the statistical difference between the 7th ranked D (Russell) and the 3rd or 4th ranked D?

      In other words, are these merely distinctions without a difference? If they are distinctions with a difference, then what is that difference? What’s the measurable difference between #7 and #4, for example?

      If he’s so bad and there’s really a case to be made about taking out Russell, then why wait? How many more games do the Oilers win if we take Russell out of the line up today?

      According to the numbers that never lie, Russell should never suit up for the Oilers again. Why isn’t that the argument being made here? Forget the contract, get Fayne on the phone and put him in the line up. Where’s that argument?

  • S cottV

    Holy reaction.

    Haven’t had this much fun since Schultz debates.

    When you go off the charts – Matt, like with Schultz, Hall and Yak – I’ve not been on side.

    However – in this case I agree. Russell is at best a stop gap measure. Maybe a year more at best, unless something better materializes on the right side.

    When the going gets real tough – like running for playoff contention or dare I say – in the playoffs, Russell gets thrown around like a rag doll.

    His deal only works when the going isn’t ratcheted up.

    We haven’t seen these conditions yet.

  • Armchair genius

    Forgive me if this has been said, I read the article and immediately rolled my eyes. These advanced stats can also be skewed with false positives. How else would Sekera be so positive, because by my eye he is terrible without Russell and makes more poor plays per shift than Oilersnation bloggers goat kris Russell. If you want to truly write something of merit, start looking at Sekera’s bloated contract, and his pros and cons. By my eye, he looks absolutely terrible on the ice at times, and I think some of this may reflect poorly on Russell. I’d like the fancy stats on him analyzed. But what do I know, I just watch the game, instead of analyzing stats. A majority of NHL gm’s don’t rely a whole bunch solely on these advanced stats because of this very reason.
    I don’t know if Russell should be extended, probably not right now, but to say he’s the teams worst blue liner by these stats alone is ignorant in my opinion.
    Arizona and now Florida will tell the tale I suppose.

    • Gravis82

      You make a valid point that @matthenderson should consider.

      The biggest bias in this study is actually publication bias.

      This is what is causing alot of the backlash. How do we know that Russel is the worst if you never write about the others? Could it be because you have a hidden agenda and dont want to talk about others because it would invalidate your argument?

      This is the heart of the issue.

      Perceived lack of transparency

  • tileguy

    Its all how you present the numbers eh Matt

    Canada’s jobless rate falls to 6.8%, 11000 jobs created in November hours ago
    Canada’s jobless rate dipped in November to 6.8 per cent, as the economy created 10,700 jobs and fewer people looked for work, Statistics …

    Guess its all how you look at the numbers

    Part-time work is fuelling Canada’s job growth this year, a discouraging trend for a country still desperately trying to recover from the oil slump.

    For the second consecutive month, the creation of part-time work outpaced full-time hours, leading to an unexpected net gain of 11,000 new jobs in November.

    Some people can make lemonade fro lemons others choose to make vinegar.

  • tileguy

    I am going to post this again since I had so much fun cutting and pasting

    Its all how you present the numbers eh Matt

    Canada’s jobless rate falls to 6.8%, 11000 jobs created in November hours ago Canada’s jobless rate dipped in November to 6.8 per cent, as the economy created 10,700 jobs and fewer people looked for work, Statistics …

    Guess its all how you look at the numbers

    Part-time work is fuelling Canada’s job growth this year, a discouraging trend for a country still desperately trying to recover from the oil slump.

    For the second consecutive month, the creation of part-time work outpaced full-time hours, leading to an unexpected net gain of 11,000 new jobs in November.

    Some people can make lemonade fro lemons others choose to make vinegar.

    • Gravis82

      OK, that is an example.

      Now apply that line of thinking to this situation. Prove to me, with numbers as you have done in your job example, that Henderson is falling into the same trap. Show how you know it is valid to make the comparison between your example and Matt’s analysis.

      I actually seem to like russle when I watch the game, but these numbers cause me to pause.

      They theory presented as to why he may be worse that he looks, also seems logical and could be true. But maybe not? I’m not going to say all of this is wrong though just because it doesn’t align with what I thought initially. To do so would be the worst definition of bias of any that have been proposed in this comment section

      • tileguy

        I’ll try but no guarentee that my words on paper reflects thoughts in my mind.

        First off know I absolutly hate hockey advance stats. This was something invented only a few years ago by people who enjoyed working with spread sheets and thought they could mimic baseball. In baseball when you have the ball in your hand it is up to you, throw it, catch it, hit it, your teamates can not make you a better player, statistics have real meaning. Hockey advance stats can not take into account the emotions that play out in hockey, see Canada versus Russin circa 1970. In hocket your teamates can make you a better player, not so in baseball.

        The author says in his article

        Corsi For per 60: 7th (50.03 CF/60)

        Corsi Against per 60: 7th (56.36 CA/60)

        Corsi For Percentage: 7th (47.0%)

        Fenwick For per 60: 7th (38.34 FF/60

        What he is saying is that Gryba, Bennng are better players than Russel. That is just not true, he is using these numbers to advance his position, Its all hogwash I say.

          • Gravis82

            I would say that i SEE that benning might be better than Russell.

            But I also see that Russell might be better than Nurse…sometimes.

            This data suggest something other than what I would have thought. But that doesnt mean I automatically think the data is wrong. I’m just as likely to be wrong too.

            So now, I’m going to watch Russell more closely. Not when he has the puck or is battling, but Im going to watch how he approaches gaps and defending zone entries. And then I am going to watch how other players do it.

        • Rock11

          This is just not true is a statement backed by absolutely nothing. I get that we now live in a world where we get to choose our own facts. But if you are going to make declarative statement like that at least try and have fact to base it on. Even if it is made up in your own little mind.

          • Rock11

            You declared that any statement made that Gryba and Benning were better than Russell “is just not true”. Like it is a fact as unopposed as the sun rising in the east but offered exactly zero in the way of fact or explanation to back up the assertion. So far the only actual facts on the table suggest otherwise and I’ve yet to hear a cogent counterargument.

            As for the word great, it’s obviously a throwaway word meant to convey people think he has real value. Which you are of course aware, but if you would like to be pedantic I retract it and substitute the words “so good” in its place.

  • Osmosis_jones

    We should’ve traded Nuge for Seth Jones when we had the chance. Then we wouldn’t even have to argue about Kris Russell and we could worry about who’s going to be our 3rd line centre.

  • dabears318

    kris russell is a perfect example of just how immature hockey analytics is.

    by eye he is killing it. stats should back up an opinion, not form them.

    this article is poorly researched.

    • Counterpoint: Kris Russell is exactly why hockey Analytics are more important than ever. They are identifying severe inefficiencies in a player that otherwise go unnoticed or our standard narratives fail to properly describe.

      As for the poorly researched part of your complaint, I’m open to the possibility I provided incorrect information. Can you point out where I did so I can fix it?

      • Jay (not J)

        Best thing about analytics is that they kept Russell cheap and available until the end of the summer when by rights he should have been snatched up July first.

        • Gravis82

          Russell was not cheap. The simple fact he was available for that long means that 3mill was too much.

          Although Chia had to take the risk, and there were worse options (which include doing nothing).

          But you do not extend him.

          Based on his stats, as you say, he should be there next summer in Sept as well, which should give the oilers leverage to get him locked up for another 1 year deal, this time at 1 million if they so choose.

      • Shameless Plugger

        The one stat analytical experts never talk about is wins and losses. Even though it is THE most important stat in hockey it is somehow dismissed. My gut feeling is because it doesn’t always support their analysis.

        Don’t discount there are smart creative human beings beyond equations and black and white. Embrace the gray area!

        • Picking out the players who most frequently contribute to Wins and Losses is kind of the whole point, but picking players who have been on winning teams isn’t analysis at all. The Penguins record was better without Crosby than it was with him for the past couple seasons. Does that mean he’s bad? If that’s the kind of analysis you think is good then we will always be at odds.

      • Harry2

        UNBELIEVABLE! Your arrogance is shocking Matt. Your saying that Chiarelli cant identify quality dmen and he should be using analytics to identify the players he should sign?

        How is it that you havent been scooped up ala Tyler Delow?

        • Gravis82

          Saying that Russell is not a good pickup, does not mean that therefore Chiarelli is bad at assessing all defensemen.

          Not sure where you are getting that extrapolation from.

          Also, what do you have against Dellow?

          • Harry2

            No but saying Chiarelli is wrong to resign Russell after having a year to evaluate him because of fancy stats is very arrogant.

            Dellow was hired by MacT/Eakins. Enough said

          • Gravis82

            what does Dellow have to do with anything?

            Dellow has not said a word since he shut his blog. How on earth do you have enough information to make any judgement on the quality of his work?

            1) Dellow was awesome, but Eakins didn’t listen.
            2) Dellow awful and Eakins listened

            If you think Dellow was awful then you are therefore saying he was bad at his advanced stats job. That implies that others could be good at their advanced stats job.

            So right now, do you think Henderson is wrong in his interpretation, and is just as bad at this as Dellow was? Or do you think that Henderson did most steps correctly (as you imply would be possible), but its the numbers themselves that are biased and wrong?

            If you think its the numbers that are biased, then tell me how you know that. If your answer is, “Because Dellow….ect again”,or bescause “they dont agree with what I see and I am always right!!”…then perhaps consider the circularity of the logic here.

          • Harry2

            Wow ykur love for Dellow is kinda creepy. Ill trust a cup winning GM over a guy writting about a player from behind his computer. Not to mention Edm’s record with and without the player in the lineup

          • Gravis82

            1) you never answered my questions

            2) I have no opinion on Dellow, either way, and I have not seen his work nor will I spend the time evaluating it because it is completely irrelevant to this discussion.

            3) what is relevant is that your dislike of someone that is completely unrelated to the current issue is somehow influencing your current interpretation of this situation.

            4) If Russell absence was related to losing 5 games as you say, and then a complete reversal of fortunes after his re-introduction, then he is probably worth 11 million. If that affect was all him, then folks, we have the best D man that has ever played on our roster.

          • passelin

            Love the huge response engagement on the math – who would’ve thought your high school math teacher was right. This is how they should teach it in school.

          • Gravis82

            Not everyone who uses analytics knows what they are doing. Does Eakins have a graduate degree with at the very least a statistical component? Um no.

            I would bet that most who do this work for teams, do.

            And its not necessarily about calculating the numbers. Its about interpreting the numbers and making sound recommendations based on known limitations. That is not a skill that develops without much guidance from people much more experienced and smarter than yourself, and years of practice.

      • Oiler Al

        There might be a place for advanced analytics in hockey,[no so much for the average eye fan],which might be both curious and suspicious of the numbers.

        I just think the numbers geeks have taken this too far. They have created paradigms for every capricious
        notion happening on the ice. eg; 5X5, then you have
        5X5 tied, leading, trailing, etc etc. It gets insane, next you will have 5X5.. ice is bad, crowd is small…
        Some guy will come up with numbers that will predict the weather on how Russell played last night!

      • Oiler Al

        There might be a place for advanced analytics in hockey,[no so much for the average eye fan],which might be both curious and suspicious of the numbers.

        I just think the numbers geeks have taken this too far. They have created paradigms for every capricious
        notion happening on the ice. eg; 5X5, then you have
        5X5 tied, leading, trailing, etc etc. It gets insane, next you will have 5X5.. ice is bad, crowd is small…
        Some guy will come up with numbers that will predict the weather on how Russell played last night!

  • Shameless Plugger

    It’s amazing how many Stanley cups were won without the use of analytics. I guess all those eyeball testers are just lucky.

    I don’t have a problem with analytics in sports. What I have a problem with is the arrogance of the people who’s stand by them as the only means to evaluate something.

    There are many factors which contribute to wins and losses. To completely dismiss heart, funny bounces, crappy ice, terrible officiating, travel fatigue, problems on the Homefront…etc…etc is quite frankly ignorant!!!

    • Gravis82

      Problem is, that no one was using analytics then. So it evened out.

      Now some are. Those that use it the best will have a competitive advantage. Its in the Oilers best interest to get on board and learn how to apply them productively.

      Is it really worth the risk of not getting on the analytics train, or at least looking into it, and then finding out too late that its profoundly powerful as the leafs win a cup before us based heavily on analystic driven acquisions?

      I cannot think of a worse fate

  • Reg Dunlop

    Well done Henderson. Now you can go relax on Lowetide’s site and mock the Luddites and Neanderthals of Oilersnation during your nightly stat-circlejerk. Don’t forget to mention your BA! They don’t give those babies out in cereal boxes. At least I don’t think they do anymore.

  • hockey1099

    I stopped reading at kris Russell is our worst defender and went right to the comments.

    My first thought was that it is better to remain silent and be thought a fool then open ones mouth and remove all doubt. If this article isn’t a satirical joke apparently it takes a two part blog to thoroughly remove all doubt.

  • Explicit

    Sure gets out of hand quickly. I like Russell, couldn’t care less what the advance stats say. He’s not the perfect #4 dman, but he’s the best one we’ve had in a long time. I wouldn’t be against having a better player to fill his spot but it sure the heck isn’t Fayne, or anyone else we have in the system.

    • Gravis82

      I like Fayne better than Russell. Fayne is low event. Russell looks like hes doing alot because he gets himself in bad spots, and then gets to try and play hero to get himself out of it.

  • Rock11

    I simply can’t believe how many people who claim they watch the games and don’t care about what the stats say think Kris Russell is great. We definitely are not watching the same game. You can criticize the stats all you want, which I think is shortsighted so we’ll have to agree to disagree, but it seems when I watch the games and Kris Russell is on the ice I usually quickly get to find out how Talbot is playing that night. For all those who think KR is great just take a moment next game and do a quick mental calculation of the amount of time the Oil spends in their own zone vs the oppo zone when he is on the ice. No fancy stats required. You wont like what you see.

  • CDNinATL

    The problem this isn’t Moneyball (the movie for those who don’t know reference…watch it). Baseball is far easier to quantify. Hockey is far more fluid.

    And this question which has been asked before, who do we replace Russell with? Plus he’s playing on his off side. I didn’t see any adjustments there.

    • Gravis82

      So because hockey is more complex than baseball, it cant be done so why bother trying?

      I would be interested in see what types of analytics world class football (i.e. soccer) teams are employing. Its very similar to hockey in its fluidity.

  • toprightcorner

    I don’t care if the stats you use show he is the worst dman on the team. All I know is that the Oilers are a better team with him in the lineup versus out of the lineup. The record proves it.

    Do I want the Oilers to extend Russel next year, heck no, I want a natural right shot replacement with some offensive skill.

    What people forget is that it doesn’t matter how good or bad a player is at any given time, the question is do the Oilers have someone how is a definite step better that can replace him. Don’t you dare tell me that Gryba could replace Russel and improve the dcore as any reputation you had would be gone in an instant.

    Russel is a legit NHL player and is part of the reason the Oilers defense has improved this season, that is an undeniable fact.

    He is still not a long term fit on this team moving forward. If the Oilers are in a playoff spot by the traded deadline, fantastic, if not, Trade Russel for a 2nd, fantastic.

  • papler

    Analytics are a support tool and Henderson does it exactly right. The reaction of some of the commenters is clearly because he started with a statement that is not compatible with their view of the player. A fine example of confirmation bias.

    But every team should (and most probably all of them do) use analytics as a measure to check their view of a certain player and discover areas to improve in an unbiased way. it
    shows you red flags where you have to dig deeper to find more information you otherwise wouldn’t be looking for.

    kinda like when you have a car that starts to use more fuel or motor oil. some don’t mind because hey, the truck drives great as always, and drive on until the darn thing goes up in smoke. others have their engine check to discover a part needa to be replaced.

    the NHL has plenty of players where the eye told a different story than stats. McDonald and Clarkson are good examples but there many more and they aren’t always that dramatic
    Pat Maroon is another example: he was traded by his team for a nickel because he wasn’t “producing” when in fact the stats said he actually was. Good for us!

    also to consider: Kris Russell did NOT have a good run with Dallas. Ever thought why he should be so much better now in Edmonton? sudden change of abilities during the offseason?

    stats are here to check on an opinion. if it matches, good, if not you check why not by digging into the videos and find a reason for the numbers.

  • I am Batman

    Russell rules. The Oilers are better with him than with Jeff “useless” Petry.

    The end.

    How is my article so much better than Henderson’s already? Guess I didn’t set the bar too high….

  • Dwayne Roloson 35

    I’m not saying Russell is the best dman in the world but when he lays down, he often gives Talbot a few extra seconds to square up to the shot or read the play.

  • Anton CP

    The old saying, it is better be lucky than good.

    Even if Russell is the worst d-man on the team so what? They are not paying him top dollars nor long turn contract. If with him that team wins more often than losses then I won’t even care if Russell is just a cardboard cutout, wins is the only number that matters. Instead of trying to make him the worst, how about making an actual effort to explain why the team wins more than losses when he plays?

    I’ve not bother with hockey advance stats at all simply because it is incomplete. The true analyzer is someone will figure out a number that can actually match the performance, not the other way around.

    • GCW

      Kris Russell wants a 5 × 5 deal. The oilers are one of the teams stupid enough in the league to mistake PDO/luck for talent. That’s a dangerous combination.

      No one cares that much this year. I hope his luck holds through the season. But it is foolish to assume it will hold in perpetuity.

      Signing him to an extension when what they need is a right handed defender who can run a powerplay just compounds the foolishness.

  • camdog

    The advanced stats say that the Oilers are a better team without Hall then they were with Hall. Of course I’m not going to leap to the conclusion that Hall was part of the problem, that said the numbers suggest he may have been. Do we simply ignore the numbers that do not support our own beliefs? Seems like some people do…

    • Gravis82

      If you presented that argument in a boardroom, to the company CEO using an analogous business example, you would be fired. Your with and without analysis as presented has too many uncontrolled and unaccounted for variables, rendering it invalid. They would have listened to you and likely lost millions of dollars

    • Given how many biases exist in the world, I’m sure there are several that apply, heck Kahneman and Gladwell are making real money raising the cognitive awareness of their nefarious existence.

      remember when “theres an app for that” solved such problems. 😉

  • CMG30

    Reading through the comments here it’s plainly obvious that far too many commentators are angry that someone dared to challenge their personal dogma rather than any legit beef with the article.