Saturday, April 11, 2009

The Difference, When It Matters

I made a claim to Boneman partway through the 2008/09 NHL season that I've wanted to follow up on for a while. As long as I was in the midst of an embargo, though, it wasn't going to happen. Now, on the other hand...

So my claim was this: I think that teams on the bubble of making or missing the playoffs step up their game in the late stages of the season. Well, duh, you might say. But where common sense might say that some teams would get hot while others ran cold, I think that the overall trend around the playoff cutoff point is strictly upward.

To be more specific: I hypothesized that the # of games over 0.500 for the teams in the 8th and 9th slot wouldn't prove to be linear. Logically, you might think that, for example, if 2 or more games over 0.500 would put you in the top 8 teams within your conference 1/4 of the way into the season, then 8 or more games over 0.500 should do it by the time things wrapped up in April (a linear increase). But my gut feel, in recent years, has been that that relationship doesn't hold true.

To see if I was right or not, I mined some data. Here's a graph showing what the results look like in the Eastern Conference this year, focusing on the teams that were in the 8th and 9th slot at the end of each month (for April's results, I used current stats):

As you can see, both the 8th and 9th place slots experienced a sharp jump in March (from +5 to +12 and from +6 to +10, respectively). Those each represent huge increases (on the order of 65% - 140%), well above what you'd expect from the changes shown in preceding months.

March therefore seems to be the month when teams around the middle of the conference start to pour it on after figuratively screwing the pooch for the first 3/4 of the season. Now, the teams in the 8th and 9th positions changed over time, but if we consider the difference between the 2 slots as being the difference between making a run at the Cup and starting your golfing season early, then it's pretty significant that "the bar" gets raised so disproportionately in the final 1/4 of the season. Or, at least, that's my theory. For what's it worth.

No comments: