Pitchers W-L records & run support

[BiZmaRK]

In 1972, Steve Carlton went 27-10 in 41 starts. In the games in which he received a decision, the Phillies averaged 3.5 runs per game and the games in which he didn't receive a decision, the Phillies averaged 3.1 runs per game. This is definitely a trend. Given such a large sampling size, how can this .4 run discrepancy be explained?

This year with the Giants, Tim Lincecum is getting run support of about 4.9 per game and Matt Cain about 4.3. Meanwhile, Barry Zito is getting run support of about 3.1 per game. If you were to take one run of support away from each of Lincecum & Cain and give it to Zito, that would put Lincecum at 3.9, Cain at 3.3 and Zito at 5.1. That would amount to about 3 to 4 extra wins for the Giants up to this date.

After a fairly large sampling size is in, why do these numbers not even out? What does it say about the motivation of the hitters with respect to who is on the mound?

[STLCards2]

After a fairly large sampling size is in, why do these numbers not even out? What does it say about the motivation of the hitters with respect to who is on the mound?

The larger the sample size, the more the run support will even out - 25 starts is still not a large sample. Seems to me like the largest factors that could keep run support different after huge sample sizes would be 1. if the pitcher was a very good or very bad hitter, or 2. if a pitcher uses a special-catcher (ala Maddux and Perez instead of Lopez). It is not supsing, then, that Maddux had the worst run support of the three HOFers.

As far as hitters being more confident or relaxed with a certain guy on the mound casuing them to hit better - that has been speculated many times, but I haven't seen any evidence. Maybe somebody has seen a study that suggests otherwise.