The Taylor Hall trade has been controversial since the minute it was announced, and former Oilers teammate Oscar Klefbom has only added to it.
On Sunday, Klefbom made a comment to Ola Winther in an interview for HockeySverige.se that Hall’s best games were never against the best teams, but he was great against the weak ones. Naturally, the hockey community exploded after seeing the comments, forcing Klefbom to immediately back-pedal, saying that he meant everyone struggled against better teams, not just Hall.
The comments have blown over since then, but the question remains: did Hall actually play his best games against weaker teams?
Before answering that we need a way to measure single game performance. Points are fine, but they don’t capture everything and the same goes for Corsi or plus-minus which are heavily team dependent.
I came up with my own method earlier this summer called Game Score. It’s a concept stolen from basketball where you combine every basic box score stat, weighing each component by its relative importance. For hockey I used goals, assists, shots, blocks, faceoffs, and penalties as well as Corsi and goal differential at 5-on-5. The scale is mirrored to points-per-game so that a “good” Game Score is immediately familiar. I’ll spare you the math, but you can read more about it here.
As for finding a player’s “best” games, I used any game with a Game Score higher than two. There were roughly 1,500 of those last season out of over 47,000 individual games. That’s an elite performance, and Hall had 10 of them. Only 21 players had more this season, with Sidney Crosby leading with 21 games.
Looking at those 10 games it turns out that Klefbom was absolutely right: Hall’s best games came mostly against weaker opponents, at least according to Game Score.
Those elite games came against three playoff teams: Florida, Philadelphia and the Rangers – all of whom lost in Round 1. The rest come against Vancouver, Buffalo, Colorado, Columbus, New Jersey and twice against Calgary. The average point total of those teams is 85.2.
So yes, his best games were against bad teams, but it’s also apparent that the relationship between performance and opponent is extremely weak. There is a slight downward trend in Hall’s Game Score when opponent strength goes up, but it’s simply too noisy to be all that meaningful. Hall still had good games against good teams, while also having poor games against bad teams. He also rarely had a negative performance which is astounding on a team as bad as Edmonton.
But does this even matter? Klefbom is right that Hall played his best against the league’s worst, but so does almost every player in the league.
Looking at the best 100-250-500-1500 games last season shows that the best games by pretty much any player will almost always be against the worst teams. The effect only increases as the games become more impressive. Of the 100 best games last season, 38 of them came against the bottom five teams in the league. Just six were against the top five.
That shouldn’t be all that surprising either. They’re bad teams for a reason and it’s because they’re easier to play against. The good teams are good for the opposite reason. The teams that allowed the fewest “good” games last year were Anaheim, Washington and Los Angeles. On the opposite end of the spectrum was Vancouver, Colorado and Columbus.
The other side of this is related to Klefbom’s stating it was the whole team that didn’t play their best against top teams. Again, he was right. Of the 55 Oiler games above a Game Score of two, the average opponent’s point total was just 86.3, fourth lowest in the league.
What’s interesting though is that the team with the lowest total was Chicago with an average opponent point total of 84.5. If it works for Chicago, it’s probably not a very good excuse for why the team struggled. Last year’s NHL scoring leader on that very team had 20 “good” games and they were mostly against weak opponents too, with an average of 86.3 points. In fact, most of the teams that feasted on weak opponents happened to be playoff teams (eight of the top 10).
And if you wanted one more reason why “only plays best games versus weak teams” is a silly narrative to peddle you can look at the player who has the lowest opponent points average for players with five or more “good” games. It’s Connor McDavid at 76.6 over seven games against Toronto, Calgary, Vancouver, Buffalo, Montreal, and Columbus twice.
Every game in an 82-game season is important whether it’s against a good or bad team. Winning is what matters and playing your best helps that cause. Taylor Hall did that more than anyone else on the Oilers last season and it really doesn’t matter who those games were against.