Revisiting Flames game scores, 30 games in

Which Calgary Flames players have contributed the most to the team’s success this season? If you want to be completely unscientific and subjective about it, the answer is “Chad Johnson and Johnny Gaudreau,” of course. But if you want to dig into the nuances of each player’s contributions, you need a tool to do so.

That’s where game score comes in.

WHAT IS GAME SCORE?

Our pal Dom Luszczyszyn (of Leafs Nation and Hockey Graphs fame) formulated game score earlier this year, adapting a concept from baseball:

Game Score is originally a baseball stat created by Bill James, but the basketball version was created by John Hollinger. It was meant as an extension and simpler alternative of player efficiency rating (PER)
and aims to “give a rough measure of a player’s productivity for a
single game” with a similar scale to points scored, something the
average basketball fan would immediately be familiar with. A Game Score
of 40 would be outstanding (James’ game six performance netted a 44.1)
intuitively because we know that a 40 point game in basketball is a damn
good game. It includes every box score stat and assigns an appropriate
value to it to provide a rough overall picture of how a player played
during that game.

Dom formulated a version for hockey that includes various things that contribute to wins, including goals, (primary and secondary) assists, shots, blocks, faceoff differential, penalty differential, Corsi differential and even strength goal differential. (Dom limits Corsi and goal differential to 5-on-5, I calculated using even strength. The difference is negligible.) For goalies, it’s goals and saves that contribute. The weighing for each factor was constructed to give goalies and skaters equal opportunity to garner the same score.

THE RANGES

In Dom’s game score updates from Leafs Nation, he’s used a scale to assess performances. Since it’s a simple scale, it’ll work well for our purposes of assessing who’s helping the Flames win (or who isn’t).

  • “Awful”: Below -0.15.
  • “Bad”: Between -0.15 and 0.15.
  • “Fine”: Between 0.15 and 0.45.
  • “Good”: Between 0.45 and 0.95.
  • “Great”: Above 0.95.

A BIG TABLE

Player Avg. Awful Bad Fine Good Great
Johnson 1.263 3 2 0 2 10
Gaudreau 0.754 6 1 1 0 12
Frolik 0.641 5 5 2 8 10
Hamilton (D) 0.625 4 6 5 5 10
Tkachuk 0.613 4 3 5 3 11
Giordano 0.535 5 3 6 8 8
Backlund 0.525 6 6 3 5 10
Versteeg 0.509 2 6 2 4 5
Stajan 0.392 6 7 5 7 5
Monahan 0.385 7 9 1 5 8
Bennett 0.380 6 6 9 4 5
Chiasson 0.376 7 8 3 7 5
Wideman 0.371 5 1 10 5 3
Ferland 0.355 4 6 8 10 2
Brouwer 0.233 10 3 7 6 4
Kulak 0.260 2 5 5 1 2
Engelland 0.259 7 9 6 4 4
Jokipakka 0.174 4 6 3 4 1
Brodie 0.146 9 10 5 2 4
Hathaway 0.104 4 5 2 0 2
Shinkaruk 0.075 3 3 0 0 1
Hamilton (F) 0.073 5 6 3 2 0
Elliott 0.058 8 0 0 0 5
Bouma 0.030 5 3 4 2 0
Vey 0.004 1 1 2 0 0
Jankowski -0.030 0 1 0 0 0
Grossmann -0.225 2 0 1 0 0

SO WHAT?

The best goalie? Johnson, by far.

The best defenders? Hamilton and Gordano by a fair margin. Note that in terms of game score, Brodie is the worst regular defender on the team (and slightly behind Jokipakka). He leads the entire team in awful/bad games (19 combined).

The best forwards? I hope you don’t have a heart condition, because prepare to be shocked: Gaudreau leads the team’s forwards. He’s followed rather closely by Frolik, Tkachuk and Backlund (in that order). Frolik leads the entire team in good/great games (18 combined). Gaudreau leads the team in great games (12), but doesn’t have any good games yet. Bouma and Hamilton are among the worst regular forwards by this metric.

A general trend: Based on game score, the Flames are split into “good players” (Johnson, Gaudreau, Frolik, D.Hamilton, Tkachuk, Giordano, Backlund and Versteeg, all averaging over .450 per game) and everyone else, who are much less consistent.

SUM IT UP

Do the various game scores hold weight? Do they match up with what’s been happening on the ice? Personally I’m a fan of the concept since it’s simple and comparable, and I think it passes the smell test given that Johnson and Gaudreau do better than everyone else. 

But do the player performances as broken down by game score match up with your impressions? Sound off in the comments!

  • MattyFranchise

    I’m interested in pursuing this line of thinking. The system as you have presented seems to pass the eye test as well.

    The one surprise on the “good” list is Versteeg, by eye, but obviously his underlying numbers are very good.

    I think this is a good metric so far.

    • Flames Fan in Edmonchuck

      Offside doesnt help but i think its mainly Wideman…. Brodie covering for him instead of playing his game…. getting a dman to play with Brodie should be priority 1/1a (along with a high end RW).

  • Lucky 13

    Hmmm, interesting for sure. Game score seems to reflect the on ice realities and it’s simplicity makes for easier digestion.

    To me (not being a Corsi, Fenwick frequent follower) it appears fairly accurate in terms of assumptions.

    Johnson has been nothing short of great in the last dozen games.
    Same for Gaudreaus performance before his injury and since returning.

    The 3M line is as it appears,steady, dependable and great at times.

    Most surprising to me would be Brody and (Kulak vs Jokipakka)

  • jupiter

    Looks to me like the Analytics crowd trying to dumb down there stats. Reminds me of a grade 3 report card.

    When the Goalie gets mixed in with players, in regards to performance, you got to question the logic. There job is so different.

  • BobB

    This simplification can be dangerous, and skewed by sample size (see:Kulak) but with game score I’ve divided it directly from percent of cap hit to get what I call BobBValue (TM). I find it interesting maybe you folks might too. In a sense this is like p/60 vs raw counting stats but /$ if you will. Johnson has been great, and Johhny very good. But Johnson makes a small portion of what Johnny makes so his BobBValue is much higher. (Sorry about formatting) :

    Kulak 73,3

    Johnson 54,2

    Versteeg 39,1

    Chiasson 34,3

    Ferland 31,4

    Tkachuk 25,2

    Jokipakka 14,1

    Hathaway 11,8

    Frolik 10,9

    Backlund 10,7

    Stajan 9,5

    HamiltonF 8,7

    Bennett 8,7

    Gaudreau 8,2

    HamiltonD 7,9

    Engelland 6,5

    Shinkaruk 5,9

    Giordano 5,8

    Vey 5,7

    Wideman 5,2

    Monahan 4,4

    Brouwer 3,8

    Brodie 2,3

    Elliott 1,7

    Bouma 1,0

    I haven’t looked really closely at this, or even gone through a proofing/validation process but my gut is that 8-10 would be the magic range. You’re scoring between 8 &10 and you’re playing at full contract value. Over ten? Likely in for a raise on the next contract (some small raises…Hathaway to large raises… Johnson, sample size acknowleged). Less than 8ish down to 1? Bad value contracts…. see the sidebar poll.

  • BobB

    Further to the above. Going forward, as this is just 30 games into the season…. maybe I’ll look at last year for more sample… but IMHO you leave Brouwer exposed if we can and playoff status considered, I’d look long and hard at trading Gio, and get some good from him while the getting is good.

    If they continue at poor value for the duration of the season, I don’t see much upside in years to come.

  • KiLLKiND

    I’ve never been a fan of the Elliot trade as we gave up far too much for a stop gap goalie. If Hiller was too old it should have been obvious Elliot almost is as well. We coul’ve and should’ve drafted Debrincat last year instead. Debrincat now has 60 points through 28 games and is a right winger. That sounds exactly like what Calgary is missing. On a side note resigning Elliot is also not worth it. I advocated for a Reimer and Johnson, or Enroth, or Ramo combo, this would have been a fine combo as Enroth will likely bounce back and Ramo is around Enroth’s ability although he is injured still. Reimer also is doing bad, but wouldn’t have cost Debrincat.

  • dontcryWOLF88

    Would have sworn Grossman was top 3? No?…!?

    haha..okay. Joking aside. Yeah, that list goes along with other data metrics I study more. I have Bennett, Ferland, and Kulak little bits higher. Monahan I have lower, though havnt dug into his numbers for a bit, and he has been racking up points recently. Brodie I also have lower.

    Any journalists here have any sort of info on Brodie’s situation? Smells like personal issues. When the rest of the team was struggling it was difficult to hassle him for his play. Now that the bulk of the rest of the team has stepped it up, the fact Brodie hasnt come along suggests there is an issue here. Most of the time he plays like his reliable, strong skating, puck moving self. Then youll see him break down and just leave you shaking your head. Its so hard to comment accurately on these things when you dont know the player personally though… Anybody here have some insight on this? He is way below his career averages in every category.

    • Lucky 13

      I’ve read that TJ’s fiancé has health issues. MS is believe?
      Not sure if it’s accurate, it was a fan who replied about it in another article.

      If that’s the case, I’d be off my game too..
      Gotta feel for these players who are trying to keep their lives private, yet they have bigger issues that they have to deal with at home.