Welcome to the glossary! Here you will find just about any statistical term that I’ll use at View from the Rooftop. If I do happen to use something in an article that you can’t find here or even if you stumble upon a term somewhere else that you don’t understand, please let me know so that I can add it here!
WAR is essentially an approximation (meaning a player with 5 WAR is not necessarily better than a player with 4.8 WAR, but he is almost certainly better than a player with 2 WAR) of a player’s total value to his team based on his offense, defense, baserunning, and position relative to a “replacement level” player, who is someone who might be called up from the minor leagues or might be a borderline Major League player available on the bench. One thing that’s important to note here is that a replacement level player is the same for every player on every team. It is not at all dependent on who a certain player’s team has on their bench or in the minor leagues. The replacement level for the Rockies’ second baseman is the same as it is for the Yankees’ catcher. There are two different formulas used to calculate WAR, one by FanGraphs (fWAR for short) and one by Baseball-Reference (rWAR). Both attempt to calculate the same thing and both use the same scale, the numbers are just slightly different. For quick reference, here is a chart that tells you what WAR numbers mean for position players and starting pitchers:
|2+ WAR||Solid starter|
|4+ WAR||All-Star caliber|
|6+ WAR||MVP candidate|
*For relief pitchers, anything above 1 WAR is considered superb.
There really is no end all of baseball statistics and you should always consider more than one when looking at a player, but if you ask me, WAR is the best place to start.
WPA – Win Probability Added
WPA attempts to figure out how much a player did or did not help his team win games. Unlike WAR and most other stats here, WPA does take into account the relative importance to the game when things happen. The basic difference is that WAR gives equal value to a solo home run in an 8-0 game in the 3rd inning and in a 2-2 game in the 9th inning, while WPA views the home run in the late, tied game as far more beneficial to the team. WPA is much less predictive than WAR, much more dependent on what situations a player happens to be put in, and does not factor in defense or base running, but it’s definitely fun to look at. The way it’s calculated is pretty simple. Say Nolan Arenado comes to bat in a situation where the Rockies have a 50% (0.5) chance to win the game and hits a home run, leaving his team with an 80% (0.8) chance to win after his at-bat, meaning he increased the Rockies’ chances of winning by 30% (0.3). This number is his WPA. A player can also have negative WPA for situations where he makes his team less likely to win. Say in Arenado’s next plate appearance, he decreases the Rockies’ chances of winning by 10% (0.1). That makes his total WPA for the game 0.2 (0.3-0.1).
LI is a number that tells you how important a certain situation in a game is. It can be a pretty complex calculation to find the LI of a certain moment in a game, but it basically revolves around the maximum amount a team’s probability of winning could change; say they’re at 50% now with the best possible outcome increasing their chances to 75% and the worst possible outcome decreasing their chances to 25%, as well as the average amount it has changed historically. The important thing to know about LI is that 1 is average, with higher being more important and lower being less important.
WPA/LI – Context Neutral Wins
One of the flaws of WPA is that it’s a cumulative stat, meaning players with more playing time or players who happen to come up at more critical times have a natural advantage. WPA/LI tries to balance that out by finding how much a player helped his team win regardless of how much opportunity he had and ends up giving you a number that is more in tune with the talent level of the player, though it still is not as predictive as WAR. WPA/LI, aside from being the shorthand terminology for Context Neutral Wins, is also the formula used to produce each player’s score. For WPA/LI, 0 is average. To give you some more context, Jose Bautista led the league in WPA/LI for position players in 2014 with a score of 5.55, while Matt Dominguez finished last at -3.45. For starting pitchers, Clayton Kershaw led the league at 5.37 with Travis Wood finishing last at -2.78. For relief pitchers, Wade Davis was the league’s best at 2.60 while Jim Johnson brought up the rear with a score of -1.57.
The total number of times a player has come to the plate.
AB – At-Bats
The total number of times a player has come to the plate and had an outcome that affected his batting average. Essentially, this is PA – walks, hit by pitches, catcher’s interference, sacrifice bunts, and sacrifice flies.
AVG – Batting Average
A player’s hits divided by his number of at-bats. Major League hitters as a whole had a batting average of .251 during the 2014 season.
OBP – On-Base Percentage
A simple calculation of what percentage of his total plate appearances a player gets on base by a hit, a walk, or being hit by pitch. Reaching base on a defensive error does not help your on-base percentage. This is a more useful number than batting average because it takes into account walks and hit by pitches, giving you the total amount a player is able to reach base. OBP is written out as a decimal rather than a percentage, so a player who gets on base 40% of the time will have an OBP of .400. Here are your general rules for evaluating a player’s OBP:
|Less than .300||Poor|
SLG – Slugging Percentage
SLG is calculated the same way as batting average, except a double counts as two hits, a triple counts as three hits, and a home run counts as four hits. The Major League average SLG during the 2014 season was .386.
A hitter’s slash line is a quick and basic way to give an overview of how a hitter has done. It’s formatted as AVG/OBP/SLG. For example, Nolan Arenado in 2014 had an AVG of .287, an OBP of .328, and a SLG of .500, so his slash line was .287/.328/.500.
OPS and OPS+ – On-base plus Slugging Percentage
Fairly self-explanatory here, this is a player’s on-base percentage plus their slugging percentage. On-base percentage has proven to be more important than slugging percentage, but this is still a useful statistic to give you a general idea of how good (or bad) a hitter is. Major League hitters in 2014 had an OPS of .700.
OPS+ looks at the same thing but is calculated a bit differently. It adjusts a player’s OPS for the ballparks they’ve played in as well as the league average for that particular year. League average OPS+ is always 100, with each number higher or lower than 100 representing one percent better or worse than league average. An OPS+ of 105 means the hitter was 5% better than a league average hitter.
Weighted runs created takes all of a player’s offensive contributions (singles, doubles, triples, home runs, walks, strikeouts, and base running) and translates it into a number for how many runs that player generated for his team.
wRC+ is the same statistic, but adjusted for league average and ballpark like OPS+ is. It is on the same scale, with league average wRC+ always coming out to 100 with each number higher or lower than 100 representing one percent better or worse than league average. A wRC+ of 95 means the hitter was 5% worse than a league average hitter. wRC+ is considered a more accurate stat than OPS+, so when in doubt, use this one.
ISO is an attempt to measure how much raw power a player has. It is calculated as SLG – AVG. This, of course, can vary from year to year, but here are your general rules when looking at ISO:
|Less than .100||Terrible|
BABIP measures the percentage of balls put in play that ended up as hits. Essentially, it is a player’s batting average if you remove his strikeouts and home runs. One component that contributes to BABIP is luck, but players have control of it to some degree as well. For instance, hitting more line drives or being a faster runner will generally lead to a higher BABIP while hitting more fly balls or being a slower runner will generally lead to a lower BABIP. In 2014, the league had a cumulative .299 BABIP, but this is a statistic where it is typically more useful to compare to a player’s previous BABIP than it is to compare it to league average. As long as that player has a large enough Major League sample size, of course. For instance, if a player has a .340 BABIP, you might compare that to league average and conclude that he has gotten lucky, thus expecting a drop in performance. However, if you look at his previous BABIP numbers and see that they are also around .340, it means his luck has probably been right about average for him and you can expect his level of production to continue.
HR/FB tells what percentage of a batter’s fly balls ended up as home runs. This can help to explain a few things; an increase/decrease in power or an increase/decrease in luck. It isn’t always clear which, but HR/FB can be something useful to look at if your favorite player has more or less home runs than you were expecting.
Ultimate zone rating is an attempt to attach a run value to a player’s defense based on the number of runs he did (or didn’t) save based on four different defensive abilities: outfield arm, the ability to turn a double play, range (how many balls they’re able to get to), and errors.
UZR/150 is the same thing, but calculated to find a player’s UZR for every 150 games played at the position in an effort to put everyone on the same scale. One thing that is important to note about all defensive metrics is that they require a very large amount of data to be reliable. If someone has only played 10 games at third base, his UZR score tells you next to nothing about how good he is there defensively. In fact, it’s estimated that you need at least three full seasons of defensive data before you will be able to get an accurate number.
Defensive runs saved uses a different method of attempting to calculate the number of runs a player did or did not save with his defense. Joe Posnanski of Sports Illustrated gives the most basic explanation of how DRS works here:
“…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.”
One of the most widely used pitching statistics, ERA is the average number of earned runs allowed by a pitcher per 9 innings pitched. It gives you a decent baseline for looking at past performance, but is not a good number to use when trying to predict the future, as runs allowed can largely be due to good/bad luck and other things outside of the pitcher’s control. As a general rule, relief pitchers will have a slightly lower ERA than starting pitchers. In 2014, relief pitchers combined for a 3.58 ERA and starting pitchers combined for a 3.82 ERA and there is a gap of a similar size every season. ERA- is a park and league adjusted stat like OPS+ or wRC+ and it uses the same scale with 100 representing league average (though again, relief pitchers tend to be slightly better). The only difference is that with ERA-, lower is better. A pitcher with a 95 ERA- was 5% better than league average.
FIP attempts to remove defense from the equation entirely and focus completely on only the things that the pitcher had control over: strikeouts, walks, and home runs allowed. FIP can tell you whether a pitcher has done better, worse, or about the same as expected based on only his statistics. If a pitcher has a 3.00 ERA and a 4.00 FIP, that pitcher has probably had some luck go his way and FIP has been a far better indicator of future performance than ERA, so going forward it would be wise to expect that pitcher to have an ERA closer to 4.00. FIP- works the same way as ERA-, adjusting for league and park, with league average at 100 and lower numbers being better.
xFIP and xFIP- – Expected Fielding Independent Pitching
xFIP is identical to FIP apart from one small difference. xFIP assumes that every pitcher allowed a league average number of HR/FB in an attempt to weed out some of the luck that can come from home runs allowed. FIP has been slightly better at predicting future performance than xFIP, but there really isn’t one number that tells you everything you need to know with pitching, so it’s wise to look at multiple sources. xFIP- uses the same scale as ERA- and FIP-.
SIERA – Skill-Interactive ERA
SIERA is another method of predicting future ERA, but it differs from FIP and xFIP in that it doesn’t completely ignore balls in play. Instead, SIERA has studied what things pitchers do that causes them to have better or worse results on balls in play. For example, pitchers that strike out more hitters also tend to allow fewer hits on balls in play because batters aren’t able to make solid contact. SIERA is another very good indication of future performance that should be looked at in conjunction with FIP and xFIP.
WHIP is exactly what it sounds like and it tells you how well a pitcher has done at keeping people off the bases. If a pitcher allowed 10 total walks and hits in 10 innings pitched, his WHIP would be 1.00. League average WHIP in 2014 was 1.28. The lower a pitcher’s WHIP, the better.
LOB% is also exactly what it sounds like, the percentage of runners who reach base and are left on base without coming around to score. This can tell you whether a pitcher has had good or bad luck in the sequencing of the hits and walks they have allowed, which can result in either more or fewer runs scored. League average LOB% is about 72% and most pitchers will eventually regress to something close to that number. However, there are some exceptions. Elite pitchers who allow fewer baserunners also allow fewer of those baserunners to come around to score, leaving them with a higher LOB% than most. Clayton Kershaw, for example, has a LOB% in his career of 78.3%. This doesn’t mean he’s gotten lucky, it means he’s a great pitcher who doesn’t allow very many baserunners.
The more strikeouts a pitcher gets, the better. K/9 tells us how many batters a pitcher strikes out per 9 innings pitched. League average K/9 for starting pitchers in 2014 was 7.36, for relief pitchers it was 8.46. Higher numbers are good, lower numbers are bad.
The fewer walks a pitcher allows, the better. BB/9 tells us how many batters a pitcher walks per 9 innings pitched. League average BB/9 for starting pitchers in 2014 was 2.69, for relief pitchers it was 3.29. Lower numbers are good, higher numbers are bad.
This is the number of batters a pitcher strikes out for every batter he walks. You can calculate it simply as strikeouts divided by walks. If a pitcher had four strikeouts and one walk, his K/BB would be 4. League average K/BB for starting pitchers in 2014 was 2.73, for relievers it was 2.57. When talking about K/BB, the higher the number is, the better.
*Special thanks to the guys over at FanGraphs.com, which I used as a reference point for some of these definitions.