|
do you think there's some readily available (and preferably already analyzed) evidence out there that could refute or support the notion of truly great pitchers pitching to the score, as heyman contends?
i could envision looking at comparing runs allowed versus run support for hall of fame pitchers in individual games, and seeing if there's any notable correlation. i would expect that in blowout wins, the elite pitchers would have no problem allowing 4 or 5 runs, but in shutout losses, the elite pitchers would only ever allow one (unearned) run. rarely would an elite pitcher allow a bunch of runs in a blowout loss, and rarely would an elite pitcher bother shutting out his opponent in a blowout win.
i mean, that's gotta be the way these things work, right?
in fact, if i do a quick look at sandy koufax, in '64 and '65, i find something very interesting. using bbref's game logs, i plotted sandy koufax's runs allowed in each game, and, based on the score of the game at the time he exited the game, tabulated the run support he received. i did this for only the games he started, throwing out the three relief appearances as not necessarily representative. then i plotted a linear trend line.
mind you, the r2 values on these trendlines render them essentially meaningless, but the direction the trendline points in nonetheless interesting. you see, in 1964, the trendline slopes slightly downward, suggesting that for each run his team scored, he allowed 0.04 runs fewer. for all intents, this is probably flat. in 1965, the trendline slopes upward, but gently so. for each run that his team scored, he allowed an additional 0.13 runs. this is bolstered tremendously by a single game in which he allowed 5 runs with 12 runs of support. combining the two seasons together shows that for each run he received in support, he allowed 0.10 additional runs.
a tenth of a run. with negligible at best correlation in an admittedly miniscule sample size.
is that what we mean by "pitching to the score"?
| 
