I have to give Brian Giles partial credit for this idea. Last week at our 1997 Cleveland Indians survivors meeting; Giles, Richie Sexson, Sean Casey and I were talking about park factors.
OK, that's not true, but Giles did kind of inspire my post for today.
Giles is the marquee hitter on the market this off-season, but one specific factor will keep him from getting major dollars: his down trending power numbers. My idea (one that is shared by many others, I'm sure) is that while Petco is a bad park for power hitters, especially lefties, it was far harsher on Giles. I won't go too far into this because I still might use it as a separate post.
We have universal park factors to tell us how much a stadium elevated run scoring. There are even park factors (in the Bill James handbook, but I'm not sure where else) broken down for left-handers and right-handers. That number gives us a nice overview, but it doesn't tell us the entire story.
Suppose we have two hitters both playing for team C. Hitter A has a road slugging percentage of .450 while hitter B has a road slugging percentage of .400. Logic suggests that hitter A should hit for a higher slugging percentage at home. But there could be something about team C's home park that makes it tougher for Hitter A. Let's say hitter A is a line drive gap hitter mostly to left field. Maybe park C has a big wall in left field that turns a lot of player A's hard hit line drives into singles. Let's say that hitter B, left-handed fly ball hitter, gets a lot of cheap home runs at home.
I'm pretty sure you all get where I am going with this. When I actually sat down and looked at the numbers it was so simple that I was upset for not realizing it sooner. At my blog I've been writing a lot about finding value. If your park is more tailored to certain types of hitters, you can extract more value based on what you pay them. If you can sign a guy with a .750 OPS, maybe he puts up an OPS of .800 because of hitting in your ball park. This doesn't make this player better or any different than he was before, but because he can exploit a certain part of the ball park into more hits or more power than the park factor suggests he should be able to, he does have extra value.
I hate to always bring up the Giants, but that's the team I'm interested in so I use them to illustrate my point. I compared the road/home Isolated Power (Iso) of every player who batted at least 100 times in Pac Bell while playing for the Giants. I used Iso rather than slugging percentage because I wanted to see how the park influenced pure power. I ran a regression for the numbers (which correlated pretty well at .635) and came up with this equation for expected home Iso. The equation is .94*Road Iso-.0005. Mind you this equation only works for Giants players from 2000-2005.
Here are the top 10 overachievers.
Here are the bottom 10 underachievers.
My next idea was to decide what types of hitters would hit for more power at Pac Bell. I decided I had a choice between separating the players into groups based on what types of hitters they are (LH, RH, good strike zone judgement, etc.) or trying to correlate something with Home Iso/xhome Iso number.
The first idea would've taken too long plus there can be a lot of human error in a survey study like that. I would essentially have to rate players on a scale to fit them into categories.
Well, the second option didn't work out either. I compared players to line drive percentage, fly ball percentage, line drives broken down by what part of the field they were hit to, ditto for fly balls. I separated the hitters into groups based on their handedness, but nothing correlated at a significant level.
Why couldn't I find anything significant? Maybe because the sample sizes needed to be bigger.
Maybe I should've broken it down by season instead of looking at the numbers collectively. All are valid ideas, but I don't have the time to try to research them. So what type of hitters should hit for better power?
19 players out performed their expected home Iso, only 4 of them were lefties.
23 players underperformed their expected home Iso, 10 of them were lefties. So given the choice between a lefty and a righty with similar numbers, the right-hander might be your better bet.
Also of note; more flyballs to left field and more line drives to right field both resulted in a decrease in power. The reasons for this probably have to do with shortcomings in the data.
Obviously this exercise could be done with any team. Off the top of my head Fenway, Petco and US Cellular field peak my interest. It's useful, especially if your team is looking for hidden value.