Alberta Baseball Confederacy: Week 22
With two weeks left in the regular season, the Alberta Baseball Confederacy's playoff-bound teams are almost decided. Three of the six playoff teams have clinched, and only the race for the third position in the Greater Western Trapper Alliance Division is really close.
It's also the time of year where those teams on the outside wonder what might have been. It's cold comfort to list off the injuries or managing decisions that cost the team the few more wins it needed. A more effective coddling, perhaps, would be provided by evidence that the team performed well enough to win -- indeed, deserved to win -- but fell upon some hard luck.
Fortunately, we have access to an excellent tool: Pythagorean standings. (To be more precise, the following calculations are based on the Pythagenpat method.) Now widely used, the Pythagorean standings are a Jamesian invention that derived a team's 'expected' win-loss record from the total runs scored for and against in its matches. Those with a greater number of actual wins than expected can take the result as a cautionary sign and fret about a fall; those who are underperforming can hope that their win total will soon rebound.
What do the results show for our league? In each division, the team on the outside of the playoff race would be going to the playoffs if the Expected Win totals were used. The injustice is greatest in the NEU Division, where the Nuke LaLoosh Fan Club underperformed their expected wins by 7, and the playoff-bound On Top outperformed their expected wins by 5. The difference of 12 wins far exceeds the gap of 6 wins that separates the teams. In the GWTA division there is still an opportunity for reality to conform to expectation, as the Zingari Italiani (underperforming by 2 wins) seek to close their 3 1/2 game gap with the Tyler Roughnecks (outperforming by 4 wins).
Other differences are more subtle, or not significant in terms of the standings. The league-leading Doctrine appear to be outperforming, but would still be tied for first overall if their 6 excess wins were stripped away like so many steroid-tainted medals. The cellar-dwelling Helmet would be third-worst if justice prevailed. The worst news must be the verdict levelled at the Supernauts: their 33-50 season is, by Pythagorean rules, entirely deserved. Sometimes there's no place to hide.
Greater Western Trapper Alliance Division
| W | L | T | PCT | ExWins | Delta | |
|---|---|---|---|---|---|---|
| The Pete Munro Doctrine* | 57 | 27 | 0 | 0.679 | 51 | 6 |
| Vancouver Drizzle | 48 | 36 | 0 | 0.571 | 48 | 0 |
| Tyler Roughnecks | 46 | 38 | 0 | 0.548 | 42 | 4 |
| Zingari Italiani | 42 | 41 | 1 | 0.506 | 44 | -2 |
| The Saint Patricians | 34 | 49 | 1 | 0.411 | 38 | -4 |
| The Willie McGee Helmet | 31 | 53 | 0 | 0.369 | 36 | -5 |
Northern Eskimo Union Division
| W | L | T | PCT | ExWins | Delta | |
|---|---|---|---|---|---|---|
| Northern Miners* | 52 | 31 | 1 | 0.625 | 51 | 1 |
| The Royal Rooters of Boston* | 48 | 36 | 0 | 0.571 | 48 | 0 |
| On Top | 41 | 42 | 1 | 0.494 | 36 | 5 |
| Nuke LaLoosh Fan CLub | 35 | 48 | 1 | 0.423 | 42 | -7 |
| Man-Ram Sula-Ram | 34 | 50 | 0 | 0.405 | 35 | -1 |
| Spruce Avenue Supernauts | 33 | 50 | 1 | 0.399 | 33 | 0 |
*Indicates team has clinched playoff spot. Top three teams in each division qualify, with each division leader earning a bye in week-one.
posted by Avi Schaumberg @ 12:07 AM
11 long time listeners, first time callers
11 Comments:
THAT was funny.
Wait a minute. I'm supposed to feel bad because I got good value for the runs my players created? The Pythagorean method with jibbly customized exponents is an empirical finding from, y'know, actual baseball. The interpretation of deviation from the Pythag estimate as "luck" applies to the game itself: you have zero warrant for applying it to models of the game.
And plus, (a) the season's not over, and (b) fuck you.
Cosh is absolutely right: the season's not over yet. And King Felix just threw what will turn out to be a 33-point shutout win.
As for our model, I think the formula still holds. What it measures is the efficiency of the distribution of runs created (and counted against).
Cosh's run total (for and against) has been distributed in a way that's produced the number of wins you'd expect it to. The poor NLFC, however, have concentrated their scoring very inefficiently: a huge number of points in a few weeks (and quite likely during weeks when their opponents also scored well). If they'd been 'lucky' enough to have those points spread out a little better, their actual record would conform better to the expected one.
If you're going to argue that Pythagoras measures run-usage efficiency as a matter of first principle, you should at least revert to the purely theoretical version with the clean square exponents--especially since your observed residues have a suspiciously nonrandom distribution. As I understand it, James & Adams generated the Pythagorean theorem (actually the log5, from which it's derived) on paper and only then confirmed its effectiveness for most-leagues-most-times. The proper exponent should orbit around 2: any other choice, like Pythagenport/Pythagenpat/Pythagenpenis, will be a product of a particular league/park/run environment and can't be defended philosophically as an a priori measure of efficiency for all models. It probably doesn't make much difference, but I'm a strong believer in shooting the messenger.
Since the piss is being taken out of my managerial performance, I'd like to point out that starting Felix was the tough call of this week for me, since he's likely to get an extra day's rest and miss his second start. No doubt it's too soon to laugh at the (10 or so) owners who ripped him when I made the trade (which will not stop me), but Matt Cain would have been the popular choice after Felix's bad outing last week.
Scary. Why don't you check up on Mussina and see how he's doing for the other side of the trade this week?
"The proper exponent should orbit around 2..." is nonsense. The exponent should be based on the run environment.
Compared to MLB, our league has a high run environment -- teams average 225 points per game. Using an exponent of 2 for that run environment compresses the Expected Wins dramatically, making nearly all the clubs .500 teams (the Doctrine, for example, would become a .519 team; the Supernauts a .467 team).
As for the Moose+Smoltz for Felix deal, I still think it was good for both sides. I landed a 50+ point week last week from big John, and Cosh landed a keeper.
We'll see if mudcrutch puts his money where his mouth is in the offseason: i.e. will he keep Verlander. Given his batting lineup, I could see all his pitchers being subject to catch-and-release. But my money says Felix, Verlander and Kazmir are kept, despite the injury risks.
Give that man an Emmy for disingenuousness! Literally everybody, up to and including Pujols, is available in trade if the price is right; that's got nothing to do with the issue.
I do have to admit one embarrassing thing, though: if I'd known that On Top was going to shit the bed so badly in the last two months of the season, I'd have held onto Moose and Smoltzie. OT sure looks like a lame wildebeest straggling behind the pack right now. I could maybe put on a burst of speed if my goddamn second basemen wouldn't keep getting hurt.
What the hell is with the Red Sox? They're actually physically falling apart right now. Lester is being tested for cancer? WTF?
I guess this gives Abboud the injury trump card in the offseason. "Oh, sure, but how many of your guys got cancer?"
Auuuggghhh! Four starts on Sunday and my guys go a collective 0-3??? What the HELL, PITCHERS?
Waaaaaaahhhh! Ryan Howard hit HOW many home runs??? Splaaaaack!!!
Post a Comment
<< Home