Players should be assigned defensive ratings only for the positions they normally play. All positions have a range rating that reflects the player's ability to reach batted balls and turn them into outs and an error rating that reflects his tendency to make errors. Throwing ratings are assigned to catchers and outfielders. Catchers also have passed ball ratings.
The defensive ratings are grouped in a box in the upper-right corner of the player ratings (Ed Note: You see this in the spreadsheet under positions)
This rating indicates a player's ability to reach balls hit in his direction and turn those batted balls into outs. Most baseball announcers use the "range" to mean the ability to cover ground, and that's certainly an important part of what goes into our range ratings. But it's not the whole story.
Our range rating (which takes values from Excellent to Poor) measures each fielder's overall playmaking ability (minus his tendency to commit errors, as we have a separate rating for that). Playmaking ability is not just about range, it's also about positioning, handling the ball cleanly, throwing quickly and accurately, and making good decisions about where and when to throw the ball.
For modern seasons, we carry out very extensive studies of play-by-play data when assigning our range ratings. We look at each player's individual performance on the balls hit his way, overall team defense, the effects of neighboring fielders (3B often take balls that the SS could have handled anyway, so we don't punish the SS for failing to make those plays), and ballpark effects.
Unfortunately, good fielding data can be hard to get for past seasons. The best sources we've found are the team section of the Macmillan Baseball Encyclopedia and the STATS All-time Major League Handbook.
By comparing putouts (for outfielders) and assists (for infielders), and adjusting for playing time, you can get an idea how a player compares with his peers. These types of measures (commonly known as range factors) can sometimes be very misleading, however, as they don't take into account the groundball/flyball nature or left/right mix of the pitching staff. The more of these factors you can take into account when assigning range ratings, the more accurate your ratings will be.
This rating is a percentage indicating how this player's error rate compares to the average fielder at his position in the era in which he played. A rating of 100 means the player is average -- that is, he makes 100% of the errors expected of someone at that position. A player who makes only 50% as many errors as his peers is rated 50. Someone who makes twice as many errors as his peers is rated 200.
The strength and accuracy of an outfielder's throwing arm are indicated in this rating, which is used whenever a runner tries to take an extra base on a single, double or fly ball. These ratings take values from Excellent to Poor.
When we assign throwing ratings for modern outfielders, we use detailed information about the number of extra bases opposing runners took on singles, doubles and fly balls hit to that outfielder. We also look at the number of runners thrown out, but outfielder assists can be misleading. Some outfielders pick up meaningless assists on plays where one or two runners score on a weak throw to the plate, and the batter is retired when the throw is cut off.
This rating indicates the strength and accuracy of the catcher's throwing arm and is used whenever a runner tries to steal second or third. It has values from Excellent to Poor.
When we assign catcher throwing ratings for modern seasons, we use detailed studies of play-by-play data to see how often opposing runners challenged each catchers arm and what percentage of those runners were thrown out. Our studies take into account any SB that were credited to trailing runners on double steals and how often a runner was caught stealing as a result of a pickoff throw by a pitcher. Most importantly, we look at the performance of each pitcher-catcher pair, an approach that helps us determine whether it's the pitcher or catcher who deserves the credit or blame for the results.
Passed ball rating
This number indicates how many times a catcher will allow a passed ball in 1,000 plate appearances with runners on base. The formula is similar to that for wild pitch ratings for pitchers:
rating = (passed balls * 1000) / (batters caught * .43)
Official statistics don't include batters caught, so you'll need to estimate it. For example, if a team's pitchers faced 6300 batters and this catcher was behind the plate 72% of time, he caught about 6300 * .72 = 4536 batters.
The .43 factor indicates that about 43% of all plate appearances occur with runners on base. This number rises and falls with the level of offense in the league.