Monday, September 19, 2011

Explaining "adjusted" goals and assists

Apples-to-oranges debates occur all the time in sports, particularly when trying to determine where two players of different eras rank historically. There's a reason that Greg Vaughn's 50 home-run season in 1998 doesn't register as much as, say, a comparable Mickey Mantle season: when Vaughn was hitting his homers, it was in the midst of the "steroid era", when home runs were being hit at a record pace. So it is with hockey. It's no coincidence that the majority of the NHL's all-time career leaders in points enjoyed their prime seasons from the early-1980s to the mid-1990s: skaters had taken that evolutionary leap to a new level of speed and playmaking ability (no doubt aided by the influences of Orr and Gretzky), while goaltenders were still relying on primitive technique and laughably small (by today's standards) equipment.

This leads to a few false assumptions among hockey fans. On one hand, there's the temptation to shrug aside the truly remarkable seasons that Gretzky and Lemieux put up during that era. On the other hand, one can overrate the historical impact of players like Mike Gartner and Dino Ciccarelli, simply because they amassed a boatload of career goals.

For a solution, I like to apply the "adjusted scoring" formula (used by hockey-reference.com and many other sites, no doubt). The history of the league has goal-scoring at approximately 6.15 goals per game played. The concept of the formula goes like this: if you scored, say, 100 points during a time when league-wide goal scoring was at 8 goals per game, then we should look at what that total would have been if you were playing in an average season (6.15 GPG). In that example, it would be about 77 points. There's a big difference in how fans view a 100-point season and a 77-point season, so the system provides a much better view of what the impact of that season was like at the time it was happening.

Taking it a step further, I like to adjust for players who didn't play when the NHL was in its 82-game schedule. Gordie Howe in his prime played 70 games a season out of 70 possible games. It's certainly possible that those extra twelve games would have worn him down, but there's no evidence to suggest that. So players who played in non-82 game seasons are prorated for a full 82 game season (this cuts both ways for injuries...if you played 35 out of 70 games, that would adjust to 41 out of 82 games).

Is the system perfect? Of course not. Who's to say that certain players wouldn't have thrived even more in higher-scoring eras, or that others would have still continued to produce even in a heavily-defensive era. And as mentioned in the paragraph above, who's to say that the players of past eras could have handled a full 82 games? But at least the system helps to level the playing field a lot, so that a conversation can ensue. And it's remarkably accurate for demonstrating which players mattered. Gordie Howe benefits immensely, because we can see how dominant a scorer he truly was. Interestingly, a player like Mark Messier suffers...many of his gaudy numbers from the 1980s weren't even enough to place him in the top ten in scoring.

The debates about the factors of an era still exist when judging players, but at least the "adjusted scoring" system allows for a common-ground discussion about what types of numbers players were actually putting up.

No comments:

Post a Comment