I started playing MLB.com's "Beat the Streak" recently as a statistical project. The way I figure it, there are a number of factors that can influence whether or not a player will get a hit in the coming game. Without going into too much detail, I'm trying to find any factor that can positively influence the number of at bats that my hitter-of-choice will have, decrease the potential for strikeouts against my hitter-of-choice, increase the probability of solid contact against either the starting pitcher or the opposing bullpen, and (depending on the hitter) either increasing the probability of a home run or a bloop single.
I figured a potential factor for this would be whether or not the game was played at home or on the road. After all, the home team doesn't bat in the 9th if they've got the lead, so that can potentially decrease the amount of PA's a batter has. However, what I found does not seem to agree with that hypothesis.
In 2008, Ichiro Suzuki started in 80 road games, averaging 4.775 PA per game, with a standard deviation of 0.7287 PAs. He also started all 81 home games, averaging 4.667 PA per game, with a standard deviation of 0.7583 PAs. I plug these into the old TI-83+ and I find that the difference in average is not statistically significant (p value of .179).
So I figured that maybe, because Ichiro is a leadoff hitter, it wouldn't be as big a difference as it might be for someone who hits deeper in the lineup. To that end, I went and looked up Alex Rodriguez.
Rodriguez had 4.4203 PA/G over 69 road games last year, with a standard deviation of .8118. At home, he averaged 4.3088 over 68 games with a standard deviation of 0.7966. Interestingly that's even less significant (p value of .209). Rodriguez batted fourth a great majority of those games (a couple games in the 3-hole).
So it looks like I'm overblowing the home/road situation in terms of plate appearances. After all, it would only come up if the home team is ahead. The Mariners and Yankees lost 46 and 33 home games respectively, suggesting they were probably behind in the 9th in virtually all those games.
It's interesting to think about, anyway.
Edited on 8/5, 1:10pm... I just went around to do some more research, comparing Ichiro to Ian Kinsler of Texas. The Rangers last year scored 901 runs last year, the most in the league. The M's on the other hand, scored 671 runs. There were only a handful of teams that scored fewer runs than that (SF, OAK, WAS, SD).
I would expect that a team that scores more runs would bring more men to the plate and ultimately inflate each player's PA/G. Kinsler batted first for Texas in 120 games last year, and had 4.8167 PA/G with a standard deviation of 0.7445. Ichiro's totals were 4.7205 PA/G over 161 games, with a standard deviation of 0.7434. The p-value for this one is the most significant of the three so far in this test, but still doesn't pass most significance tests (0.1423). Still odd.