Sabermetrics

Declaring WAR on Clayton Kershaw’s 2011

One of baseball’s most distinguishing traits is the enormous number of statistics available from eras new and old.  If you are so inclined, you can find information and stats on just about any season, team, and player since professional baseball became a thing in the 1800s.  As a matter of fact, there is an entire organization called the Society for American Baseball Research (SABR for short) dedicated to recording and preserving teh history of the game.  (It is from this group’s acronym that the term “sabermetrics” was formed).  One of the best places to do this kind of research online is Baseball-Reference – you can purchase a subscription for some of the data but there are a lot of free features.

With the abundance of stats, it is inevitable that the urge will arise to compare players from all eras of baseball with each other.  However, this argument typically proves futile as one quickly realizes that many era-specific factors – such as different opponents, league size and composition, advances in equipment – affect a particular player’s stats and performance.  Does hitting 50 home runs in one season in the 1960s mean the same thing as hitting 50 home runs in a season in the 1990s?  How well do ERAs compare over the course of the 20th century?  Both are good questions, and until very recently, we were left to resort to qualitative assessments of players based on assumptions and opinions to answer them.

Fortunately, some very smart baseball minds developed a way to quantify a player’s value in a way that allows for comparisons across eras.  This quantity is known as Wins Above Replacement (or, simply WAR).  While it’s a very complicated formula involving some advanced metrics, the end result is a number that describes the value, in wins, that a play is worth to his team when compared to a replacement level player.  Put another way, it answers the question,”If Player A is unable to perform and is replaced by a minor leaguer or bench player, how many fewer (or mare in some cases) wins will his team have?”

One of the beauties of WAR is that it is context, league, and ballpark neutral.  This means that the effects of game situations, which league a player played in (AL vs. NL), and the run environments of different ballparks are accounted for in the calculations.  Rather than give you all the dirty details, I will point you to the Fangraphs library for further reading on the nuts and bolts of WAR.  As noted at Fangraphs, using one metric for player evaluation is not ideal, but WAR provides a great starting point since it accounts for all aspects of player performance – batting, baserunning, defense for hitters and runs saved (or prevented) for pitchers.

With that in mind, let’s take a look at Clayton Kershaw’s 2011 performance from a WAR standpoint and see how this campaign fares historically.  If you look at Kershaw’s player page over at Baseball-Reference, you can see that his rWAR* value for the 2011 season was 7.0.  2011 was Kershaw’s age 23 season (his age at midnight of June 30, 2011 was 23) and since the advent of the CY Young Award in 1956, there have only been twenty seasons that saw a starting pitcher with a rWAR of 7.0 or greater in their age 23 season or below.

Of those twenty, only Kershaw, Dwight Gooden in 1985 (age 20) and Roger Clemens in 1986 (age 23) came away with the Cy Young Award.  This may seem arbitrary at first glance since the Cy Young Award is determined by a vote, but the bigger picture is that Kershaw is one of three pitchers to have performed at such a high level at a young age and be perceived as the most outstanding pitcher in his respective league.

Another point worth noting is that, going back to 1880, there have only been 25 seasons in which a left-handed starter posted a rWAR of 7.0 or greater at age 23 or younger.  (Awesome side note: a guy named Noodles Hahn had two such seasons in 1899 and 1902).  Before Kershaw’s 2011, Jim Abbott’s 1991 season and Frank Tanana’s 1975-77 seasons with the California Angels were the most recent occurrences.

To reiterate, it is typically not safe to pass too much judgment based on any one statistic – Kershaw actually accumulated a lower rWAR than NL Cy Young runner-up Roy Halladay (7.4).  We can certainly begin to form ideas and make comparisons, but the analyst must be thorough before drawing any concrete conclusions.  We will continue to examine Kershaw’s success in 2011 in the next few posts by looking at the effect he had on the Dodgers as a whole and how he used his repertoire of pitches to his advantage.

*There are two accepted forms of WAR out in the baseball universe, fWAR and rWAR.  fWAR uses the Fangraphs methodology while rWAR is the version used by Baseball-Reference.  Values are very similar but use slightly different formulas.

Sabermetrics

Viewing Clayton Kershaw’s 2011 with Sabermetrics and Traditional Stats

Pitching at any level of baseball is a constant test of physical and mental strength.  At the major league level, only the catcher sees a more demanding workload.  If you have pitched at any level, then you know of the continuous focus required to navigate a pitch, a batter, an inning, a lineup, a game, a season.  For big leaguers, off days are far from “off” – conditioning, side bullpen sessions, film study.  Surely the rituals will vary for relievers and starters but you get the idea. Pitching is hard.

Every once in a while, a pitcher will have a breakout year and fool us into thinking – just for a few moments each time he takes the mound – that pitching looks easy.  Enter the 2011 version of Clayton Kershaw.  Every fifth game for the whole season, Kershaw gave the Dodgers a chance to win.  His traditional stat line looked like this:

21-5 with a 2.28 ERA and 248 K

which, by the way, was good enough for the pitching Triple Crown in the National League (he was tied with Ian Kennedy for the NL lead in wins).  This is a nice way to sum up a season, but as I mentioned in the overview post about Kershaw, we are much more capable now of identifying more meaningful metrics when it comes to analyzing and predicting a player’s performance than we were even 20 years ago.  These traditional stats are nice, but do they properly characterize a player and his performance?  You guessed it, the answer is no.

For example, pitcher wins has come under scrutiny in the sabermetrics world as being a somewhat meaningless stat.  Sure, it tells us that the pitcher’s team won the game and he was qualified to tally the win, but can we tell how the pitcher performed just by his number of wins?  Take a look at these two scenarios:

-Pitcher A throws a no-hitter and picks up a win in a 1-0 contest

-Pitcher B gives up 6 runs and walks 5 batters in 5 innings but his team scores 7 runs to qualify him for the win

Both pitchers earn a win, but Pitcher A obviously had the better performance.  The downfall is that the traditional win stat does not differentiate between these two performances.  Now stretch this over an entire season and it will quickly become clear that a traditional stat like the win can be a misleading characterization of a pitcher’s aptitude.  This is just one example of the way that traditional stats are being scrutinized in today’s baseball culture.  Not only is it becoming more practical, but it’s fun (I know, nerd alert!) to explore in more detail the more advanced stats that have been created to better understand the game.  I am still learning many of these new stats from the sabermetric world.

Now that we have established Kershaw’s dominance of the traditional stats, let’s take a look at some of his advanced metrics and compare him to the rest of the league.  Using the Fangraphs Leaderboard we can see that Kershaw is in the top 3 in the National League for:

Strikeouts per 9 innings (K/9) with 9.57 – 2nd
Walks plus hits per inning pitched (WHIP) with 0.98 – 1st
Fielding Independent Pitching (FIP) with 2.47 – 2nd

The point is this: Clayton Kershaw had a fantastic year on the mound any way you slice it.  You might even reason that his traditional stats are in fact telling an accurate story since his advance metrics were also superb.

Now that we have established (briefly) that Kershaw had a stellar 2011, there are a couple directions to go in defining his success.  One avenue is to explore his season in historical context – this opens up the classic baseball conundrum of comparing players across eras.  The other is to begin exploring the nuts and bolts of how he was able to be so successful in 2011.  We will go one of these directions the next time we meet.  Until then, let me know if there are any sabermetric principles you would like to discuss.