• Goalies in B2B Games

    by Tyler Dellow • November 29, 2013 • Hockey • 7 Comments

    Ilya Bryzgalov started for the Oilers last night in Nashville and the Oilers won 3-0. Ilya Bryzgalov is starting again tonight for the Oilers. Eric Tulsky provided some pretty convincing evidence last year that starting goalies in back to back games is insane. Eric looked at instances of teams playing a goalie in back to back games and found that, from 2011-13, goalies who were playing in the second half of a B2B who had played in the previous game posted a .901 save percentage while those who hadn’t play in the previous game posted a .912 save percentage.

    I went and took a quick look at this year’s data and the same trend is evident. Goalies playing the second half of a B2B who were rested the night before have a .922 save percentage so far this year; those who didn’t play the previous night have a save percentage of .916.

    It sure seems like it’s a no brainer to start both of your goalies in back to back games. A six to ten point difference in save percentage is actually a pretty large thing, particularly when you take into account that teams playing back to back games are more likely to be fatigued and give up more shots. Imagine a team that takes and allows 28 shots per game. Imagine further that this team scores on 8.5% of its shots and allows goals on 8.5% of its shots, a .915 save percentage. They score 2.38 goals per game and allow 2.38 goals per game. This is a pretty average team that I’m describing, one that will put up 92 points or so.

    If you change things so that this team has a .905 save percentage – basically apply a second half of a B2B penalty – the team will go from allowing 2.38 goals per game to allowing 2.66 goals per game. They go from being an even goal difference team to one that goes about -23 over the course of 82 games. That will turn a 92 point team into an 84 point team or so. It’s a mammoth change.

    Teams play so few B2B games that botching this decision may not show up in the course of a year. That’s not really an excuse though. Part of coaching is taking the players that you’ve been dealt and trying to maximize every little edge that you can find. The Oilers have a perfectly cromulent second goalie in Devan Dubnyk and he’s played well lately. Playing Bryzgalov against Columbus might not cost the Oilers but, unless Dubnyk has an undisclosed injury of some sort, Dallas Eakins is hitting on 18 here.

    Email Tyler Dellow at mc79hockey@gmail.com

    About Tyler Dellow

    7 Responses to Goalies in B2B Games

    1. November 29, 2013 at

      I’d be curious to know if any goalies have posted two shutouts in back-to-back games.

    2. November 29, 2013 at

      Bryzgalov becomes the third different Oilers goalie to start on consecutive nights this year.

      LaBarbera Oct 14/15: .800 / .893

      Dubnyk Nov 9/10: .882 / .861
      Dubnyk Nov 15/16: .919 / .943

      Bryz Nov 28/29: 1.000 / ???

      Also Bachman got consecutive starts with a day off in between, but where he absolutely got pummelled the first game with 48 shots + 3 more in a shootout in a long night vs. the Kings. He went .979 / .846.

      It’s odd in that if you’d asked me two months ago, I’d have bet on Dubnyk being the only Oilers goalie all season to get back-to-back starts, & preferably not too many of them at that.

      • Bruce McCurdy
        November 29, 2013 at

        Bryz Nov 28/29: 1.000 / .826

    3. skatinginsand
      November 29, 2013 at

      Dallas Eakins has proven himself to be much smarter than any analytics… oh wait…. 4-0 Columbus, end of second. Never mind.

    4. Triumph
      December 1, 2013 at

      I mentioned this to someone on Twitter who cited Eric’s study, but I was sort of amazed that Martin Brodeur doesn’t demonstrate this tendency. I theorized that Brodeur’s lower career SV% might just be because he started so many back to backs, but his SV% in those games is the same as it always is.

    5. mattyc
      December 1, 2013 at

      Even if your starter in B2B is handicapped, wouldn’t still make sense to play them if you backup is even worse? Ex. if Bryzgalov SV% – handicap > Dubnyk SV% (not meaning to say Dubnyk sucks…)

      • Pierce Cunneen
        December 1, 2013 at

        yes, but you need to give your starter rest at some point, so why not do it when he would otherwise not be at his best. Thus, you can play him the next game and he will be at his best and rested.

    Leave a Reply

    Your email address will not be published. Required fields are marked *