Stats Glossary

Thoughts On Statistics

As you've probably noticed by now, I use statistics quite a bit around here, and a lot of them are of the sort that don't get shown on TV broadcasts or discussed by our illustrious broadcast crew.  You won't see mich discussion of pitcher wins or RBIs here.  I use stats that usually get filed under "advanced metrics" or "sabermetrics" and here's why (and I promise I'll make every attempt to avoid a long winded diatribe).  It's not to set myself up on one side of the "stats vs. scouting" or "old school vs. new school" debate or another.  I think both sides have plenty to offer to our understanding or baseball, but I tend to lean more towards a "new school" approach to the game while trying to avoid too much scorn for the old school approach.  

However, regardless of how we feel about the (false) dichotomies above, we all use stats and always have.  RBIs and ERA are just as much stats as as wRC+ and FIP.  It's just that traditionally we have used imperfect stats that give us a flawed or limited understanding of what a particular player contributes.  Often it is because, as with pitcher wins and RBIs, they measure things that aren't within the control of the player (e.g. runs scored by their team or men on base when they're hitting) or because they include other statistical noise that obscures what we're really trying to measure.  I don't feel there's any debate over whether or not stats have value, since we all use them.  I just figure if we're going to use stats, we might as well use the best ones available, the ones that give us the clearest understanding of the player or game.  I believe those are usually the "advanced" stats. 

They can seem intimidating at first, but anyone can figure them out.  I did, and I'm no economist or statistician, just a guy that took the time to do some reading and thinking. 

To that end, here's a glossary of some of the more common advanced stats that I refer to.  This is not comprehensive as some metrics are pretty self explanatory (K% or HR/FB for example), or I just don't refer to them very much, so I won't put them here.  Just enough so you can play along at home.  I can't take credit for pretty much any of this.  The explanations are either directly copied from, or heavily paraphrased from FanGraphs as that's where I started learning. I don't intend to take credit for their work, just to give it to you in a nutshell. If you want to read more (and I strongly encourage you to do so!) just follow each link to FanGraphs' more thorough explanations or check out their complete glossary here ---> [FanGraphs's Stats Glossary]

Offensive Stats:
  • OBP - On-Base Percentage.  Measures the percentage of times a player reaches base more accurately than Batting Average because it includes all of the scenarios in which a batter can reach base: Hits, walks and hit by pitch.

  • OPS and OPS+ - On-Base Percentage Plus Slugging.  Exactly what it sounds like, a player's OBP added to their Slugging Percentage.  It's not perfect, but it's the most mainstream catch-all stat for a player's overall offensive contributions.  OPS+ measure the same things, but is scaled so that 100 is league average OPS, and every point above or below 100 is a point above or below league average.

  • wOBA - Weighted On-Base Average.  This is one of the best catch-all offenseive metrics going, as it ascribes different values to different types of batting outcomes in terms of their likelihood of creating runs, and does so more accurately than Slugging Percentage.  Basically it's HRs > 3Bs > 2Bs > 1Bs > HBP > (unintentional) BBs in terms of value.  It is scaled to OBP, so .320 is about league average, .400 is excellent and .290 is pretty awful. 

  • wRC and wRC+ - Weighted Runs Created.  Also a great catch-all offensive metric.  It quantifies the same things as wOBA, but measures it in terms of how many runs a player creates for their team.  Like OPS+, wRC+ uses is scaled to 100 as league average and each point above or below 100 is a percentage point above or below league average.  So a player who has a wRC+ of 125 creates 25% more runs than league average.

  • BABIP - Batting Average on Balls In Play.  This measures the percentage of balls that a batter puts into play (so excluding HRs, because they don't count as "in play") that go for hits.  It is affected by three factors, talent, defense, and good old fashioned luck.  The idea is, that once a player puts a ball into play they have very little control over what happens to it.  Maybe it's a hard liner right at a fielder, maybe it's a lazy blooper just over the infield for a hit, maybe they smoke a ground ball that gets gobbled up by a great defensive play.  BABIP recognizes that over the course of a seaon, about 30% of balls put in play will end up going for hits, so league average is about .290 - .310.  While some players may have a consistently higher or lower than average BABIP, most will fall within this league average range, so a BABIP dramatically higher or lower may be an indicator that a player is due for positive or negative regression back to average because they've been getting unusually lucky or unlucky, or have often faced above or below average defense.  (Note: Regression does not mean to get worse!  It just means their numbers will get closer to league or career averages as time goes on, whether that is better or worse that current performance.)  Also, this:

  • ISO - Isolated Power.  This is a measure of a hitter’s raw power. Or, to look at it another way, it measures how good a player is at hitting for extra bases. The simplest way to calculate ISO is to subtract a player’s Batting Average from their Slugging Percentage, which leaves us with a measure of just a player’s extra bases per at bat.  Basically Slugging Percentage without the Batting Average noise.  .140 is about league average, .250 is excellent and .080 is awful.

There are a number of these and they are not as well developed or reliable as offensive or pitching metrics.  They take huge sample sizes to tell us much of value, and even then their accuracy is a topic for debate.  When I refer to them, it will likely be to one of two metrics:
  • DRS - Defensive Runs Saved.  This measures the number of runs above or below average that a player saves for their team.  0 is average, +15 is a Gold Glover, and -15 is awful.  How it's calculated is both complicated and probably boring to you, but a short explanation is:
“…as I understand it, the numbers determines (using film study and computer comparisons) how many more or fewer successful plays a defensive player will make than league average. For instance, if a shortstop makes a play that only 24% of shortstops make, he will get .76 of a point (1 full point minus .24). If a shortstop BLOWS a play that 82% of shortstops make, then you subtract .82 of a point. And at the end, you add it all up and get a plus/minus.” (Joe Posnanski, Sports Illustrated)

  • UZR and UZR/150 - I generally use this less than DRS, and I'm not even about to try to condense the explanation for them.  Instead I will once more rely on the way smarter people at FanGraphs.  For more on UZR go here.  UZR is just UZR/150 defensive games played.
  • ERA- -  I assume if you follow baseball enough to be on this blog, you're familiar with ERA.  ERA- is similar to OPS+ and wRC+ in that 100 is league average and every point above or below is a percentage point above or below average.  However, since a lower ERA is better than a higher ERA, a lower ERA- is also better than a higher ERA-.

  • FIP and FIP- -  Fielding Independent Pitching.  The problem with ERA, is that it is subject to BABIP.  Luck and defense have a significant impact on it.  FIP is an attempt to show what a pitcher's ERA should have been assuming outcomes on balls in play had been league average.  It is calculated using only outcomes that are within the pitcher's control, which is to say strikeouts, walks, hit by pitches and home runs.  It has been shown to predict a pitchers ERA from year to year better than the previous year's ERA does.  It is scaled to ERA, so league average for FIP and ERA is about equal at 4.00.  2.90 is excellent and 5.00 is awful.  FIP is helpful for identifying pitchers whose ERA results may not be the best representation of how well they actually pitched.  FIP- is like ERA-, but with FIP.

  • xFIP and xFIP- - Expected Fielding Independent Pitching.  A pitcher's HR/FB rate will generally fluctuate significantly from year to year, with the thinking being that while FB rates are somewhat within a pitcher's control, how many of those fly balls leave the yard and how many stay in the park for outs or hits is largely due to luck and the parks they play in.  xFIP is just like FIP, but with an assumed league average HR/FB rate.  xFIP is helpful for identifying pitcher's whose FIP may have been affected by abnormally high or low HR/FB rates. xFIP- is like ERA-, but with xFIP.
WAR: What is it good for?! Never heard that one before right?! Just kidding... Coming Soon!

No comments:

Post a Comment