Last Goalie Post. Promise.

Braden Holtby, Semyon Varlamov, & Michal Neuvirth

It’s important when using statistics to make arguments or try to sway public opinion that you are not above learning something new. For me, this includes the effect of zone starts on scoring, how Corsi correlates to winning, and why scoring chances are an important metric to track. Each one helps put a player’s performance into context to try and determine the true skill level of a hockey player– both past and present. This is why I don’t change my mind, but I do make new decisions based on new information.

Take Washington’s goalies for instance. At first, I felt Michal Neuvirth deserved the right to be the “number one,” and I wasn’t wrong. He had a wonderful November and performed admirably. Then Semyon Varlamov’s groin started to heal and showed that he was ready to reclaim the number one spot– especially in the context of how he performed in different game situations.

Well, here I am again; only this time I have added a new point of data to my scoring chance database to help me see the level of competition each line combo or d pair is playing against– and that includes goalies.

I have taken the average relative Corsi for each skater on the ice at the time of the scoring chance against.

Relative corsi represents the difference in Corsi rating (the balance of Total Shots For & Against) when a player is on the ice, compared to when he’s on the bench.

So how do the goalies stack up?

Goalie EV Sv% Corsi QoC
Varly 0.869 0.926
Neuvy 0.848 0.603
Holtby 0.722 0.063

Still Varly. He has a higher scoring chance save percentage (which is an indicator of a goalie’s true skill) than Neuvy, and he faces tougher competition. Braden Holtby is, well, not ready for the NHL quite yet. A line with Sidney Crosby has a Corsi QoC value of about 1.20– that should put into context the quality of scoring chances Varly has kept out of the net.

Unfortunately, we don’t know what “average” should be, since only five or six teams are currently being tracked in scoring chances. I also agree trying to figure out quality of competition this way has its flaws (and I am open to suggestions for improvement), but for a ‘quick and dirty’ way to see the opposition skaters go up against I think I am on the right track.

  • Brad

    I think Bruce needs to loosen up on how long one goalie stays in between the pipes. Just because a goalie loses a game doesnt mean you replace him the next. GIve him time to get a groove going in the crease.

  • Rockin’ The Red in VA

    Here’s a thought, and this is mainly just a conversation piece. Isn’t Evgeni Nabokov still available? If so do you (Peter and Ian) and us (the readers) think he might be a good acquisition? Here are his career stats:

    563 553 32490:53 293 178 37 1294 2.39 14757 13463 .912

    Just a nugget to chew on. . .

  • Fedor Fedin

    He is and his name usually not far away from the Caps in the rumors.

  • Agree that maybe the Caps should rethink the “win and you’re in” strategy for the goalies…perhaps if Varly had just started the other day instead of Neuvy (who they knew might be nursing a sore groin), Neuvy wouldn’t have added more stress to the injury and Varly wouldn’t have gone in cold in the second.

  • Peter

    @Rockin’ The Red in VA

    I defer to Neil in matters of trades. I always have the “dance with the girl that brung you” attitude.


  • Ian

    @Rockin’ The Red in VA

    You have to remember that Nabokov was 8-8-5 with a 3.02 goals-against average and .888 save percentage with SKA. If your George McPhee you look at a player like this and immediately pass. He hasn’t been going up against NHL competition for about a half a year, and you already have two young netminders who are keeping you in every game. I’m personally satisfied with the Caps’ goaltending, but at this point, if we can pick up a veteran ‘tender at the deadline, let’s do it. Regardless of what happens, I hope Bruce eventually anoints one of these guys as the #1 towards the end of the season so we have some stability at that position come playoff time.

  • Just another thought…maybe with scoring chance sv% QualComp is better to use (measuring scoring) whereas when measuring scoring chance rations Corsi Rel QoC is better to use.

    If Varlamov’s performance this season is for real…he’ll be winning something shiny at the end, in some tourney.

  • Neil – RMNB

    @RedArmyLine I thought of that, and actually have QCOMP in my database as well. The challenge with using that is Sh% is not sustainable, nor correlated to winning on the level that Corsi is.

    My decision to use Corsi as the barometer was to see which goalies see scoring chances against from skaters who put more “pucks in play” against, thus giving credence to the Sv%.

    Also, I worry about the “eyeball test.” The QCOMP numbers for a line with Crosby & Stamkos were negative – which would cause some people to doubt the numbers.

    I am experimenting with translating both QoC numbers to a scale of 1 to 10 using avg +- Stdev. Any thoughts on that – or the metric as is from anyone – is appreciated.

  • Neil – RMNB

    @Rockin’ The Red in VA Dan Ellis (.907) has almost the same career save percentage as Nabokov (.912). At this point you are talking about saving maybe 5 more goals over the next 30 games or so – really not enough to make a difference. Plus his age is worrisome.

    IMO a decision needs to be made: Either both goalies’ health is not an issue, in which case you roll the dice OR you feel one is going to be hurt most of the time (like this season) and you need a clear, healthy #1 option via trade.

  • Neil – RMNB

    @RedArmyLine let me re-phrase: “The challenge with using that is Sh% is not sustainable, nor correlated to winning on the level that Corsi is.”

    Obviously Sh% is correlated to winning since more goals = more wins. I meant in a “repeatable, controllable” sense.

  • …and Nabokov was just signed by Detroit, but it’s not like he would be any better than Varly or Neuvy anyway.

  • abrlcklnthewall

    Clearly we need to hold onto both for as long as possible.

  • Sean

    On the Nabokov issue, with both Neuvy and Varly hurting, I’d take a gamble. Detroit signed him, but he has to pass through waivers first since he played in Europe this season. Caps are currently 23rd in waiver priority.

    Nabokov’s stats in the KHL are pretty worthless. Bigger rinks, worse defense in front of him, family issues, etc. It’s not a direct corellation to the NHL.

    More on point, it’s essentially a no risk move. Both Varly and Neuvy are waiver exempt, and won’t have to pass through waivers if either are sent to Hershey. Additionally, at only $250k salary pro-rated, Nabby wouldn’t cost Leonsis much nor impact our salary cap situation enough to prevent the addition of a significant player towards the TDL.

    If Nabokov struggles, make him the backup. We’re right back where we are currently with Varly or Neuvy starting most games. If he shines, we’ve upgraded. It’s a no risk move with obvious upside. Especially given that I don’t try Varly’s groin and I don’t want Holtby starting more games that necessary.