Thursday, September 06, 2007

Another year, another title

The Cougars rallied to win the B Division playoffs in the Sudbury Men’s Modified Fast-Pitch Softball League. We lost the first game in the series 18-14, then in the second game hung on for a 9-8 victory as the potential game-winning run took a called third strike, setting the stage for our heroics. Copying the template that won us the C league championship two years ago, we went down big (8-1), began rallying with a Parinella home run, and scored several in the last inning to win it.

[Quick rules: 10 fielders, fast-pitch but with no windmilling, no leading or stealing, “courtesy” runners allowed for gimps, everyone bats, otherwise regular rules except for the occasional 6 feet high/4 feet wide strike zone. We averaged about 10 runs per 7 inning game this year.]

The final inning was as boring as a comeback could be. Down three, we got a bunt single, two outs, a bloop, an infield single, three outs, and another single up the middle, followed by complete dejection from the bitter opposing captain, who normally pitched for the team but got relegated to catcher due to wildness in his previous outings (roughly two walks per inning in 5 games against us this year).

Our team got outhomered 35-7 this year, and even outscored 218-208 despite putting up a 12-8 record (4-1 in the playoffs). We had three blowout losses (16-1, 11-1, 16-4) and no blowout wins (biggest was 15-9).

I put up a discouraring offensive line of .412/.424/.588 (BA/OBP/SLG), not very far off from the team line of .397/.468/.518. In a high run environment like this (about 1.5 runs per inning), slugging is less important than in baseball, since each runner is more likely to score. Previous years:
2006: .535/.549/1.070
2005: .439/.455/.756
2004: (just 2 games) .667/.571/1.333
2003: .536/.567/1.000
career: .508/.523/.958, an extra base hit every 4.2 AB

This year I just could not hit the ball squarely, with just two HR, a gift 2B and 3B, and a smattering of hard-struck singles and line outs in 51 AB. I figured out the end problem (bat is hitting bottom of ball, popping it up) and the preceding cause (right shoulder too low) but could not get at fixing the root cause (too upright of a swing plane). I can't say that I tried to fix it with practice, as almost all my swings all season were in games (took a few swings in BP a couple times, made it to a cage about a week before the end of the year). It was my most frustrating season in ball since age 11 when my coach thought he had all these hot shot 12 year olds.

On the plus side, this was easily my best fielding season, aided in large part (I think) by the purchase of a new glove. I had been using a Bobby (yes, Bobby) Bonds baseball outfielder glove purchased in the '70s, and the larger softball never seemed to bounce too easily into it, leading to a fair number of imperfectly fielded grounders. But I got a new softball infielder's glove for Father's Day and immediately noticed a difference, and fielded practically every grounder cleanly the rest of the season. That combined with my strong arm and above-average range led to a high Range Factor/Zone Rating/Plus-Minus/pick your advanced fielding metric.

Just like completion percentage in ultimate, fielding percentage in baseball or softball is only telling when other things are equal, and they usually aren't. Even in a high-quality league like the majors, the variation in errors at a position is significantly smaller than the variation in number of plays made, even after correcting for opportunities. Advanced fielding metrics today calculate on a play-by-play basis, comparing the fielder in question to a league average on balls hit in that area at that speed.

The parallel to ultimate stats is that an individual's contribution to scoring goals and avoiding turnovers is only partly captured by the number and percentage of passes he completes. A guy who clogs or cuts at bad angles is every bit the turnover machine as a teammate without a forehand, yet might show up high in the throwing stats. An effective cutter could actually appear worse relative to his teammates because he always provides them an easy target while never getting to throw to himself. And at a higher level, throwing percentage is to some extent simply a choice, inversely related to the yards per throw.

So, I actually got a little belligerent at the bar afterwards when discussing fielding, as one of my teammates was listing errors as a proxy for how well the team fielded, so I countered with "plays not made" (which I actually listed as "errors"), including what would have been just a single had he cut it off before it hit the gap, and a hard-hit ball that the third baseman couldn't field, and a short fly that fell in, etc. There are many more of those in a game than errors, and even if each is only a 50/50 proposition while errors happen on balls that should be outs 95% of the time, the net number of extra outs is greater due to plays not made.

And I would be remiss if I didn't trumpet the triumph of our demographic in this game. we have two guys younger than 35, and our pitcher is 62 (but still throws hard and locates the ball well), while our opponents were probably a median age of 22, with many of the players members of the state champion high school baseball team a few years ago.


itchy said...

Absolutely. As Bill James always said, there is a much larger deviation between players in "plays made vs. plays not made," as opposed to "plays attempted and successfully executed."

We'd need a much better stat-taking method in order to properly analyze these in ultimate.

parinella said...

Bill James came up with Range Factor 20-30 years ago, counting the plays a player made in the game, before eventually coming around to the view that this too was fatally flawed.

The first flaw was that no matter how bad a defense is, it will get 3 outs per inning, and somebody has to make the plays eventually.

The second flaw was that opportunities at a position were not equal across teams. Some pitching staffs strike out a lot and/or give up lots of flyballs, reducing the groundball opportunities for infielders.

Any stat for which there is a large team effect has these problems for measuring individual stats. To take another example, yards per carry for a running back depends a lot on the offensive line and types of run (line plunge on 3rd and short vs draws on 2nd and long). You can get a good sense of the team effectiveness with this stat (although that too will suffer some), but the variation at the individual level will be due more to context than to skill.

So, for ultimate, I would summarize thus:
1. Individual aggregated stats provide only a crude estimate of skill or value.
2. Team aggregated stats can provide a better estimate of team effectivenss (and possibly the effectiveness of those who are in on those points).
3. To get good individual stats, play-by-play data is necessary.

Edward Lee said...

I put up a discouraging offensive line of .412/.424/.588 (BA/OBP/SLG), not very far off from the team line of .397/.468/.518.

Why so few BBs?

parinella said...

Several reasons. One, my pitch recognition hasn't been great. Mostly this translates into swinging at borderline pitches when I don't have to. Two, I don't hit a lot of foul balls, so if I swing, there is a 75% chance the ball is in play. Three, I'm there to hit, so I'm choosing to play non-optimally (in terms of creating runs) because I enjoy it more that way.

Anonymous said...

Who knows where to download XRumer 5.0 Palladium?
Help, please. All recommend this program to effectively advertise on the Internet, this is the best program!