Off-Day Nats Stats: Baseball Stats 101, Hitting

As part of our continuing effort to make Nats fans the best fanbase in baseball (and in a shameless attempt to encourage more comments to stat-related posts), I've decided to start an occasional series about basic baseball stats.  Don't worry, there's not going to be a bunch of math, and I'll try to minimize the weird acronyms.  My goal here is to explain what the basic stats are and what they can tell you about a player.  These will also be the same numbers that I'll throw around with little or no explanation in upcoming stats posts as the season progresses.  Today's lesson is basic hitting stats.  Class starts after the jump.  (Don't worry, there won't be a quiz later.)  Fun hitting-stat tidbits at the end as your reward for staying with it!

The slash line

The most common way to summarize a hitter's ability is the so-called triple-slash line.  This is a short list of three different batting stats, separated by slashes.  The stats, in order, are batting average, on-base percentage, and slugging percentage.  They're all decimal numbers out to three decimal places. For instance, Ryan Zimmerman's slash line in 2010 was .307/.388/.510.  Let's talk about what that actually means.

Batting average (BA)

Batting average is the fraction of at-bats in which a player got a hit.  Batting average is a little screwy because "at bats" don't include walks, hit-by-pitch, or sacrifices.  The idea is that BBs, HBP and sacs are all good things, so a player's batting average shouldn't be penalized by treating them as instances of "didn't get a hit."  League average is about .280.  A BA around .300 or higher is good, while .250 and below is bad.  BA has been around for well over 100 years, but newer stats are more useful for showing how good a player is at hitting.  For instance, BA doesn't include walks, and it also doesn't differentiate between slap-hitters and sluggers--which brings us to the next two stats...

On-base percentage (OBP)

On-base percentage is the fraction of plate appearances in which the player gets on base.  It includes walks, HBP, and sacrifices.  Another way to think of OBP is the percentage of time that a player doesn't make an out.  This makes OBP better than AVG for evaluating hitter performance, since avoiding outs is ultimately more useful for the offense than getting hits (although getting hits is part of avoiding outs).  League average OBP is about .340; less than .320 is bad, while greater than .360 is good.  Comparing BA and OBP can tell you something about a hitter's approach.  Remember Cristian Guzman's swing-at-everything philosophy?  In 2009, his BA was a decent .284, but his OBP was an out-machine-like .306.  Compare that to 2009's Discerning-Eyed Nick Johnson, who had a similar .295 BA, but a stellar .408 OBP; by drawing more walks, NJ got on base a third more often than the Guz.

Slugging percentage (SLG)

Slugging percentage is like batting average, except extra-base hits count extra: a double counts as two hits, a triple is three hits, and a home run is four hits.  This means that SLG could theoretically be as high as 4.000 if a batter hit a home run in every at-bat (OBP and BA max out at 1.000 if you always get on base or always get a hit).  Comparing BA and SLG tells you something about what kind of hits a batter gets.  If every hit is a single, then BA and SLG will be the same.  The more extra-base hits a hitter gets, the higher his SLG will be compared to his BA.   For instance, Adam Dunn had a .260 BA and a .536 SLG in 2010.  His SLG was a bit over twice his BA, meaning that on average, every hit he got was a bit more than a double--a rare accomplishment.  League-average SLG is around .420; more than .500 is good, less than .350 is bad.

Beyond the slash line

The slash line is the basic summary of hitting prowess that you'll see on every scoreboard and on-screen ticker in MLB.  However, it takes three numbers to sum up how well a player is hitting.  Isn't there some way to sum it all up in one number?

On-base-plus-slugging (OPS)

OPS is an attempt to combine a player's ability to avoid making an out (OBP) with his ability to hit for power (SLG) into one number.  Like the name says, you figure out OPS by adding OBP to SLG. For example, Zimmy had an OPS of .898 in 2010.  Average OPS is around .750 or so; more than .800 is good, and less than .700 is bad.  OPS is appealing because it's so easy to figure out from the slash line, but it has problems, too.  OBP and SLG aren't equally important for figuring out how good player is at offense, and OPS treats them as equal.  Also, if you're trying to put a player's entire offensive contributions into one number, why not roll in baserunning, too?

Weighted on-base average (wOBA)

The wOBA stat takes almost everything a player can do to score runs--hits by type, walks, stolen bases, and HBP--and rolls them into one number.  With the huge amount of statistical baseball history, it's possible to calculate how many runs any offensive event is worth (just like you can calculate the odds of winning a game from any base/out situation for finding WPA).  For instance, a dinger is worth 1.7 runs on average, and a walk is worth 0.62 runs.  wOBA is the average run value per plate appearance of everything a hitter does, weighted to put it on the same scale as OBP.  Like with OBP, a wOBA of .340 is league average; less than .320 is bad and more than .360 is good.  Ryan Zimmerman had a wOBA of .389 in 2010, which is nice, indeed.

The "plus" stats (OPS+, wRC+)

The "plus" stats take one more step to make a single number that expresses a player's offensive value in a way that is easy to compare among players.  The "plus" stats (OPS+ and wRC+) take the underlying stat and do a couple of things to them to enable comparisons on as much of an apples-to-apples basis as possible.  First, they adjust the data for park effects.  If a hitter spends most of his time in a pitchers' park like Petco (or old RFK), he gets more credit for hitting HRs than a hitter who gets a lot of AB in a hitters' park like Citizens Bank Park or Great American Ballpark.  More sophisticated analysis will actually weight individual hit types differently (for instance, it's easier to get a hit to drop for a double in the spacious outfield of CitiField compared to the average ballpark, but it's harder to hit a HR over the absurd outfield wall).  The second part is to "normalize" the number by scaling it so that a league-average hitter has a "plus" stat of exactly 100; more than 100 is better than average, and less than 100 is worse than average.  This makes comparing players a snap.  You don't have to think in your head that so-and-so's SLG is inflated by playing in a hitters' park, or try to sort out which parts of his slash line are above and below average.  A higher plus stat is better, period.

The two main "plus" stats I like are OPS+ and wRC+.  OPS+ is the park-adjusted, normalized version of OPS.  It's the most common hitting "plus" stat, since OPS is generally about as far into hitting stat math as the casual fan is willing to go.  However, it shares the limitations of OPS by weighting OBP and SLG equally, and by not including baserunning.  If you're using OPS+, you should really be using wRC+ instead.  You can think of it as the park-adjusted, normalized version of wOBA (technically it's based on a stat called wRC that also includes SB and CS--the details aren't important). wRC+ wraps up everything a player can do to help his team score runs in one number: a player with a wRC+ of 100 helped his team create exactly a league-average number of runs; a player with a wRC+ of 120 created 20-percent more runs than average, while a player with a wRC+ of 75 created 25-percent less runs than average.  Who cares if the math behind it is complicated when the end result is so easy to compare and interpret?

Nats with(out?) Bats

Given what we've just learned, let's see how the Nats are doing after the weekend sweep!

• As a team, the Nats are hitting .226/.316/.351 with an 80 wRC+.  They're creating 20-percent fewer runs than average, and they are below average in every part of the slash line. That puts them in the bottom four in the NL in team offense
• Danny Espinosa leads the team among players with at least 50 PA, sporting a 119 wRC+ and a .256/.364/.512 slash line.
• Rick Ankiel lags the 50+ PA club, with only a 60 wRC+ and a .211/.286/.298 line.
• Expanding our net to smaller sample sizes, Ryan Zimmerman had a 172 wRC+ in his 37 PA before hitting the DL (.357/.486/.536), while Wilson Ramos leads among active players with a 150 wRC+ in 34 PA (.414/.500/.483).
• Putting aside Jesus Flores' 457 wRC+ (the result of getting a hit in his single PA so far this season--talk about SSS!), every other Nat I haven't mentioned so far has a wRC+ of less than 100.
• Jason Marquis' 52 wRC+ is higher than Mike Morse (46), Jerry Hairston (48), Matt Stairs (37), and Ivan Rodriguez (27).
• Adam "league average" LaRoche is almost there, with a .245/.339/.367 line and a 96 wRC+.
• Despite his painful-to-watch struggles at the plate, Ian Desmond's .217/.266/.383 line is only about 15-percent worse than average, for an 86 wRC+.
• For all his wolfish fearsomeness, Jayson Werth's .200/.302/.382 so far is only good for a below-average 89 wRC+.

In conclusion

The best stat to quickly compare overall offense is probably wRC+: 100 is league average, it includes ballpark effects, and it scales directly with the number of runs above or below average that a player creates (more is better).  You can find it on the invaluable fangraphs website. Right now, wRC+ tells us that the Nats aren't providing much offense outside of Spinner and Chubs.  Who says fancy stats have to disagree with what you can see with your own eyes?