August 8, 2003
Back To Basics
So, as a service to the people who may be exploring serious baseball analysis for the first time, or who may be new to Baseball Prospectus, here's a brief rundown of some basics of performance assessment. It's spotty, but it's a start. For you longtime readers, please consider this a cheat sheet you can use when discussing baseball in bars, or with Bob Feller.
1. How come you don't use RBI to evaluate hitters?
There are a few problems with RBI, in terms of using it to evaluate the performance of an individual hitter. RBI are dependent on other players. A hitter, by himself, can't drive in more than one run in a plate appearance without a contribution from his teammates. Hitting a home run with the bases empty counts for one RBI. Hitting a home run with the bases loaded counts for four RBI. The individual contribution of the hitter doesn't change, but in one case, the hitter gets four RBI, and in the other, the hitter gets only one, and the difference is entirely dependent on the performance of the hitter's teammates, not the hitter himself. So using RBI to evaluate individual players is problematic.
2. What are park adjustments, and why should I care about them?
The performance of the players on the field is dramatically affected by the environment in which they play. In recent years, people have paid more attention to park effects among the general baseball populace, largely because of the introduction of the most extreme hitters' park (Coors Field) in the history of baseball. But there were park effects before Coors Field, and there will always be park effects, no matter where the game is played.
In order to compare player performances accurately, it's important to understand the extent to which a player's home park (and road parks, and even league) affect that player's performance. We have methods for adjusting statistics based on where a player plays, to allow better comparisons of players who play in different parks. The methods aren't absolutely perfect, but they are very good. We also have methods for interpreting stats from the minor leagues, foreign leagues, and the independent leagues. Check out our statistics pages for more information.
3. I've always used Batting Average, Home Runs, and RBI for hitters, and Wins, Losses, and Saves for pitchers. I don't want to go overboard, but I do want to understand players a little better. What should I use instead?
That's a matter of taste. No pile of numbers is ever going to tell you everything about a player. It's always a tradeoff between how easy it is to acquire and understand the numbers, and how much information you really want or need. Personally, as a shorthand, I use the three-number set of BA/OBA/SLG for hitters, and IP, H, BB, K, and ERA for pitchers.
BA/OBA/SLG (Batting Average/On-Base Average/Slugging Percentage) gives you an idea of the "shape" of a player's performance in a concise way. A guy who hits .250/.380/.550 is a patient slugger; his average isn't great, but he walks a lot (high OBA for that batting average), and a lot of his hits are for extra bases. (high SLG for that batting average) A guy who hits .305/.340/.380 is a slap-hitting guy; he doesn't walk much, and he doesn't accumulate many extra-base hits.
For pitchers, wins, losses, and saves are highly dependent on team performance, much like RBI for individual hitters. It's a little more difficult to encapsulate a pitcher's performance than a hitter's. By using the IP/H/BB/K/ERA set, you get an idea of a pitcher's usage pattern, his control, his stuff, and his effectiveness, at least to a pretty reasonable extent. If a pitcher's giving up few hits and striking out a ton of guys with a low ERA, he's probably pitching fairly well, no matter what his Win-Loss Record.
Over time, you'll get adjusted to these numbers, or whatever set you find yourself using, and you'll have the same comfort level you now enjoy with the more traditional stats. The entire point of using a particular set of numbers, though, is because they reflect the performance of the players on the field. Baseball stats are really just records in an arcane accounting system, and it's important that the system you use be as empirically close to reality as possible. For more information on this and other statistics, check out our Glossary.
4. How come I get scoffed at when I use fielding percentage in an argument about a player's defensive ability?
Defense is very difficult to measure effectively. For a very long time, people have been using errors and fielding percentage, largely because: (1) errors are highly visible and memorable, (2) we really didn't have anything a whole hell of a lot better, and (3) the game was played very differently 100 years ago than it is now. But fielding percentage really doesn't tell you much, and can, in fact, confuse the issue with bad information. Consider two hypothetical outfielders, Albert Hands and Bob Range, both of whom played 162 games behind identical pitching staffs in identical parks:
Player Putouts Assists Errors Fielding Percentage Albert Hands 324 3 1 .997 Bob Range 486 24 10 .981
Bob Range made 10 times the errors Albert Hands did, but made an additional out per game, and was an immensely more valuable fielder. But using either errors or fielding percentage, Albert Hands appears to be the more proficient defender. Fielding percentage overrates the rangeless defender with good hands, while penalizing players with superior range who may make (and attempt) more difficult plays.
We'll try to mix in a few of these from time to time, even though this is probably redundant for most of the readers. For those of you for whom this is totally redundant, I offer this gratuitous link to a hot babe. Or not.
Join us for a Baseball Prospectus Pizza Feed in Menlo Park on August 21! Click here for details!