Study: If the Leafs had a good game, how did they play in the next game?

Gus has done some good work looking at 10-game segments for the Toronto Maple Leafs this season, and I wanted to quickly look at momentum, specifically, whether the Leafs’ performance over one 10-game segment makes much of a difference compared to the next segment.

It can also be applied to games as well. What is the Leafs record, for example, coming off of a win as opposed to coming off of a loss? Well, in the most recent season, the Leafs were 14-16-5 off a win, and 20-21-5 after a loss. Not enough evidence there to make you think that the team gets into a groove or starts playing with confidence or momentum.

Look at it this way: Flip a coin 20 times. If the coin lands on “heads” four times in a row, does “heads” have all the momentum? Regardless of the last flip, there’s always a 50-50 chance the coin will hit heads or tails. In hockey, the coin is weighted, and it’s been weighted against the Leafs for quite some time now.

From a further perspective, we can tally up what happened the game after every game the Toronto Maple Leafs played since the lockout, and see whether the team’s goal differential in any game impacted the next result:

It doesn’t appear to follow any sort of pattern. Rather, over a huge span of games, it looks to be completely random. But let’s tally it all up here and see how the team’s record does after wins and losses, in every game since the 2005 lockout:

  W L OTL GF GA Points/82 GD/82
After Wins 104 113 32 716 844 79.0 -42.2
After Losses 113 93 34 742 775 88.8 -11.3
After OTL 33 36 9 184 236 78.8 -54.7

Funny that the reverse seems to be true: The Leafs play better after losses than wins. That can be some sort of regression, or just due to general game-to-game adjustments. The team is below .500 coming off of wins, but coming off of regulation losses, the team is closer to playing like a playoff team.

If you do it by 10-game segments, you get a similar, but less drastic effect. There are seven segments to every season, with the seventh segment having 12 games (they count games 81 and 82). I split up all 49 segments over the 7 years and sorted them from best to worst based on goal differential. Then I checked the Leafs’ performance in each of those segments stacked up against the next one:

  Pts/82 GD/82 Next Pts/82 Next GD/82
Best 12 105.9 34.9 78.6 -41.7
Next 12 84.7 -13.0 87.3 -31.1
Next 12 81.3 -49.2 88.5 -3.9
Last 13 54.2 -95.2 78.3 -46.6

The lesson here is that the better the team plays, the further they’ll come back down to earth. The reverse doesn’t appear to be true, however. When the team gets into a funk, they tend to stay there. The worst the Leafs did in either points per 82 or goal differential seemed to come the 10-game segment after their very worst ones. The Leafs’ last two segments of this past season fit into the “Last 13” category. When they fell apart, they really fell apart. It didn’t seem to affect individual game performance however.

(With much love to Gus and the indispensible Hockey Reference)