Friday, September 5, 2008

My Favorite Stats: Part 2 of 2

I covered my favorite pitching statistics here in Part 1. Today I will look at batting statistics that I frequently reference.

The Best Stats

“On-base percentage." This measures how good a player is at not making outs, which is the main goal of a batter. It has traditionally not been as valued as batting average, although appreciation of OBP has increased in recent years. It is listed during Mets broadcasts on SNY alongside batting average, HR, and RBI. League average OBP is .334 this year. Barry Bonds OBP’d an outrageous .609 in 2004.

“Slugging percentage.” A statistic that many fans cannot easily understand, slugging percentage is a good way to measure a hitters power. League average SLG is .421 this year. The formula is: 1B + (2B * 2) + (3B *3) + (HR * 4) / AB. Think of it this way: a double counts as 2 singles, a triple 3 singles, and a homerun 4 singles. Tally up the total singles and divide by AB. Albert Pujols leads the NL in SLG at .640.

(side note: whenever you see something like “his numbers are .299/.345/.487”, this is Batting Average/OBP/SLG. It’s a strong way to get a quick look at a player’s production, and takes like 5 seconds to look at.)

“Value of replacement player.” This is where it gets a little nerdy and people stop listening. I can see why most do not want to hear about VORP – the main problem is the name I guess. However, it’s a great stat, and measures how much a player contributes to his team compared to a theoretical “replacement player” at the same position. This is a counting stat (like RBI or runs scored) as opposed to a percentage stat (like OBP and SLG). A player with a VORP of 22.5 has created 22.5 more runs than a “replacement player” at the same position would. VORP tells us that Hanley Ramirez and Jose Reyes are extremely valuable players, because they play shortstop, a position that lacks many great offensive players. Same goes for Joe Mauer and Geovany Soto at the catcher position. Click here to see VORP leaders. VORP exists for pitchers as well.

“Equivalent Average.” EqA attempts to measure a player’s production independent of park and league factors. EqA is notable for including stolen bases and stolen base attempts into its simple yet scary looking formula, which is linked here. Every year, average EqA will be .260. Similar to batting average, a strong performer will have an EqA at .300 or higher, while those around .220 are poor performers. David Wright leads all qualifying Mets in EqA at .311. Damion Easley’s is .241.

“OPS Plus.” See below for a description of OPS. Similar to ERA+, OPS+ measure a player’s OPS against the league average, adjusting for park factors. It is a great way to measure player performance across eras. Babe Ruth’s career OPS+ was 207, meaning he was 107% better than the average player during his career. Rey Ordonez’s career OPS+ was 59. Ouch.

Good Stats

“On base percentage plus slugging.” Add up OBP and SLG, and this is what you get. The main problem with OPS is that it implies OBP and SLG are equal in value, when in reality OBP is more important. Nonetheless, it’s a useful barometer of a player’s performance.

“Homerun.” The best possible outcome for a batter in any plate appearance. Contrary to popular belief, home runs do not kill rallies.

“Batting average on balls in play.” This is very similar to the pitcher’s version of BABIP, except league average will not normalize to .290. Instead, a player’s career BABIP should be used as the baseline to determine if he’s lucky or unlucky during a period of time. For instance, Carlos Beltran’s career BABIP is .299, so if one year it’s .240, he’s probably getting unlucky. There is a strong correlation between BABIP and LD% (line drive %). A hitter's BABIP will be about 120 points higher than his LD%. For instance, Wright's career LD% is 22.9%, and his career BABIP is .338 - 109 points higher than his LD%.

Poor Stats

“Batting average.” This is the stat traditionally used to measure a batter's performance. Do NOT be fooled – it does not tell the whole story. Obviously, it does not factor in walks to its equation. It is borderline worthless, and will have some thinking Derek Jeter and his .294 BA is having a better year than Adam Dunn and his .240 BA.

“Runs batted in.” Most still think RBI is a way to tell who is the best at driving runs in. It is a counting stat, and does not tell us how many opportunities a batter had to arrive at his RBI total. For instance, if player A has 100 RBI and player B has 70 RBI, one could assume that player A is a better run producer right? Wrong. Looking deeper, player A had a RISP in 400 of his PA’s, while player B only in 100 PA’s. With this extra info, player B is the superior run producer. Check out this FJM destruction of an answer to a mailbag question I asked Marty Noble about this topic. Also, read my post on David Wright’s RBI total here for more.

“Runs scored.” Similar to RBI, this is a flawed counting statistic. Also like RBI, it largely hinges on the production of other players on the team. Avoid it.

There are some other stats out there that I haven’t covered, such as WPA, WARP and runs created. However, basic knowledge of the “great” and “good” stats listed above should allow one to understand most of what I write on here.

Here are a couple other acronyms I use:
- RISP (“runners in scoring position”)
- PA (“plate appearances”)
- LOB ("left on base", as in runners left on base)
- XBH ("extra base hits")
- LD% ("line drive %")
- BAA ("batting average against")

Special Addendum regarding defensive metrics
I’m cautious when it comes to placing a lot of value in defense metrics, mostly because they are somewhat harder to quantify than offense metrics. Stats like range factor and defensive efficiency have their merits. However, the defensive metrics I’ll really use here are FRAA (“fielding runs above average”), UZR, and Dewan's +/- system. FRAA is a counting stat, like VORP, with an FRAA of 0 being league average. A negative FRAA implies below average fielding, a positive implies above average fielding. Endy Chavez’s FRAA this year is an amazing +12, meaning his defense has been good enough to save 12 runs. By comparison, Torii Hunter’s FRAA is +14. Dewan's +/- system tells us how many plays a player has made above or below league average. A player with a +10 has made 10 more plays than a league average defender would be expected to make.

No comments: