Tuesday, April 8, 2014

Game of the Day (4/7/84)

Tigers 4, White Sox 0. Noted non-Hall of Fame inductee Jack Morris facing off with former strikeout titlist Floyd Bannister.

Both starters had a fairly easy time of it in the first; Lou Whitaker's leadoff single made him the only runner to reach in the inning, and Alan Trammell hit into a double play to eliminate him from the bases. The second went slightly better for the Tigers, as Lance Parrish drew a leadoff walk and Chet Lemon hit a go-ahead 2-run homer two batters later.

From there, the bases remained clear until the bottom of the fourth. Rudy Law drew a leadoff walk and stole second, and Carlton Fisk and Harold Baines also drew free passes to load the bases with nobody out. But Greg Luzinski hit into a 1-2-3 double play, and Ron Kittle struck out, allowing Morris to preserve the lead. And the lead grew larger in the next half-inning, as Lemon led off with a double, Kirk Gibson doubled him home, Tom Brookens bunted Gibson to third, and Whitaker hit into a fielder's poor choice - the Sox tried and failed to throw Gibson out at home, and his run made it a 4-0 margin.

The game proceeded quietly from there. Vance Law walked in the bottom of the fifth, as did Gibson in the top of the seventh against reliever Tom Brennan. Luzinski drew a free pass in the bottom of the seventh, and again with two outs in the bottom of the ninth, but Morris fanned Kittle to end the game.

And that's it. In terms of determining who would win, this game held very little drama. So what makes it the best of the day?

You may have noticed that the recap did not contain any mention of hits by the White Sox. That's because there weren't any. Morris allowed 6 walks, but threw a no-hitter.

WPL is a very strong system for measuring the excitement generated by changes of fortune in terms of victory and defeat over the course of the game. But there are other things on a baseball field that also create drama, and no-hitters (and perfect games) are among the best. They are also responsible for the only subjective adjustment I make to WPL scores when picking a day's best game - a no-hitter gets a 3-point bonus; a perfect game gets a 4-point bonus (that is, the no-hitter bonus plus an extra point).

There are two questions that need to be addressed here. First, where do the bonus numbers come from? Frankly, they are almost entirely arbitrary and subject to potential future adjustment. The reasoning behind them is as follows: The best 9-inning game in my large and ever-expanding database scores just over 6 points on the WPL scale. A no-hitter is about as exciting as a 9-inning game gets - except that the excitement is only occurring when the no-hitter is in jeopardy; when the other team is batting, you're just waiting for their half-innings to end so the good stuff can start back up (especially if the game isn't otherwise in doubt). So the no-hitter bonus is equal to half of the best nine-inning game I've measured. The perfect game bonus is equal to half of a really exceptional extra-inning game, although not the best extra-inning game I've measured or especially close to it; I don't have a particularly good reason for this.

Second, why do no-hitters get a bonus when other exceptional individual performances (like high-strikeout games for pitchers or 4-homer games for hitters) don't? The short version of the answer is that no-hitters are different. If a pitcher has 12 strikeouts through 5 innings and retires the first batter of the sixth on a popup, he still has 12 strikeouts; he's no longer as likely to break the record, but he hasn't actually lost anything. If, however, a pitcher has a no-hitter through 7 and the first batter of the eighth hits a single, the no-hitter is irretrievably gone. The ever-present risk of immediate and total loss adds a continuous drama that's lacking in other cases. (Again, this is arbitrary, but defensible, I think.)

As a secondary consideration, a no-hitter or a perfect game can also be readily identified through the numbers I enter into the spreadsheet anyway (runs/hits/errors for each team picks out the no-hitter; a perfect game will have exactly 0 positive WPA contributions from the victimized lineup). Strikeouts or home runs would require extra searching that would add time to the not-inconsiderable process of data entry. Moreover, looking for near no-hitters would require even more time to examine the play-by-play accounts of each game, and that's time that I can better spend on other things.

All of this arbitrariness certainly adds seams to the WPL method - but then, since I'm trying to assign a value to the excitement of a game, the method was hardly seamless before. At least now it can (and does) pick out a no-hitter as the best game of the year to date.

No comments:

Post a Comment