I’m sure you’ve been anxiously awaiting this, but it’s time once again (groan) for the annual bit-past-the-one-quarter-mark* event: The AHL, forced run through the Pythagorean method.
The math isn’t that scary: Basically, by taking a team’s totals of goals for and goals against, we can get a good projection of what that team’s winning percentage should be. The shootout adds noise in hockey: Basically, I chop out the shootout, convert records and goals-for-and-against back to regulation-plus-overtime totals, and plug that data into the formula.
So you get three numbers. First is “real winning percentage,” old-school wins and ties. The second is “Pythagorean winning percentage,” based on (GF^2/(GF^2+GA^2)). The third is the difference between them, sort of an “overachieve” (if the real percentage is higher) or “underachieve” (if the real percentage is lower), if you want to use those charged words.
I don’t know if it’s actually mathemagically sound, the way I do it, but heck, it’s for fun and giggles here, anyway.
For a refresher, here’s the 2006-07 version, and here’s last year’s. Again, all caveats apply; again, especially the Simpsons reference**.
Through Tuesday’s games:
Team | “Real” pct. | Pyth. Pct. | “Overachieve” |
IOW | 0.646 | 0.574 | 0.071 |
BIN | 0.460 | 0.393 | 0.067 |
MIL | 0.595 | 0.547 | 0.048 |
PEO | 0.542 | 0.494 | 0.048 |
TOR | 0.500 | 0.455 | 0.045 |
PRO | 0.583 | 0.542 | 0.041 |
POR | 0.625 | 0.594 | 0.031 |
WOR | 0.521 | 0.493 | 0.028 |
SYR | 0.542 | 0.524 | 0.018 |
PHI | 0.538 | 0.526 | 0.013 |
BPT | 0.615 | 0.603 | 0.013 |
SPR | 0.438 | 0.427 | 0.011 |
HER | 0.750 | 0.748 | 0.002 |
MCR | 0.457 | 0.458 | -0.002 |
HFD | 0.438 | 0.443 | -0.006 |
SA | 0.241 | 0.247 | -0.006 |
MTB | 0.696 | 0.706 | -0.011 |
NOR | 0.370 | 0.383 | -0.013 |
LOW | 0.479 | 0.493 | -0.014 |
GR | 0.542 | 0.555 | -0.014 |
HOU | 0.500 | 0.515 | -0.015 |
ROC | 0.200 | 0.217 | -0.017 |
LE | 0.381 | 0.410 | -0.029 |
WBS | 0.520 | 0.552 | -0.032 |
RCK | 0.522 | 0.555 | -0.033 |
ALB | 0.313 | 0.356 | -0.044 |
HAM | 0.580 | 0.624 | -0.044 |
CHI | 0.540 | 0.598 | -0.058 |
QC | 0.400 | 0.465 | -0.065 |
(Some fractional differences are from rounding three numbers.)
For fun, I converted the Pythagorean percentage to a points total, then put the bonus points back in: added a point for an overtime loss and a point for a shootout win. Remarkably, the order barely changed; only three teams changed, and by no more than two places.
In theory, a team with a big difference between its real and Pythagorean percentages is due for a correction. So if you’re surprised that Iowa is ahead of Chicago and that Peoria is right with them, well, there you go; Iowa is about three and a half points ahead of pace, Peoria is about two and a half ahead, and Chicago is three points behind.
To me, by far the most interesting thing on that whole chart is that Hershey, mighty Hershey, unbeatable Hershey, they-can’t-keep-this-up-can-they Hershey… is right where it belongs.
So then I got to wondering. You know, doing this three years: It’s all well and good, but what does it mean in the long term? So I crunched out last year’s final numbers.
Team | 20-game “overachieve” | Final “overachieve” |
TOR | 0.124 | 0.049 |
HER | 0.114 | 0.034 |
HAM | -0.095 | 0.028 |
CHI | 0.04 | 0.028 |
SPR | 0.046 | 0.026 |
BPT | -0.04 | 0.017 |
RCK | 0.011 | 0.016 |
IOW | 0.044 | 0.011 |
LOW | 0.01 | 0.009 |
PHI | 0.020 | 0.009 |
POR | -0.029 | 0.007 |
PRO | 0.056 | 0.004 |
SA | -0.02 | 0.004 |
HFD | 0.033 | -0.001 |
MIL | -0.009 | -0.002 |
WOR | 0.054 | -0.002 |
ALB | -0.061 | -0.004 |
MTB | 0.001 | -0.012 |
SYR | -0.113 | -0.012 |
QC | -0.026 | -0.012 |
HOU | -0.022 | -0.013 |
WBS | -0.032 | -0.013 |
ROC | 0.055 | -0.015 |
NOR | -0.043 | -0.024 |
GR | -0.0006 | -0.026 |
LE | 0.010 | -0.028 |
BIN | -0.033 | -0.033 |
PEO | -0.059 | -0.035 |
MCR | -0.062 | -0.039 |
I’ll let you tell me if you see anything interesting in there. Or in here, anywhere.
*-Not “quarter pole.” That means you’ve got a quarter of a mile to go, not elapsed. (Pet peeve.)
**-“I wish, I wish, I hadn’t killed that fish.” (2F03)