Saturday, December 1, 2012

OBI% - A Simple Definition of Baseball Efficiency

When a baseball player comes up to bat, they have one main objective:  to produce runs.  Baseball, not unlike other sports, is a simple games where the winning team scores more runs than their opposition.   In most cases, when a batters come to the plate with a runner on base they will attempt any option they have as long as it maximizes their ability to drive in that runner.

Today, I am going to introduce you to a stat tracked by the good folks at Baseball Prospectus.  The statistic is OBI%.  OBI% stands for "others batted in percentage" which in layman's terms translates to the percentage of all runners who were on base that were batted in by a particular hitter.  This article will explore the results of 2012, the meaning of this data, and the obvious and not so obvious flaws of using this statistic to judge the success of a particular hitter.

In a lot of ways, OBI% is an efficiency statistic.  The efficiency of a particular hitter's at bat resulting in the direct production of a run.  If I had to guess I would expect players with high averages and high RBI totals to dominate this statistic in a simple correlation.  Without a doubt, I would anticipate AL Triple Crown winner Miguel Cabrera to have the highest OBI%.

Listed below is the top ten OBI% in major baseball who qualified for the batting title by securing 502 PAs:

As expected, Miguel Cabrera exceled in his OBI%, but what was unexpected to me was he beat out a player with an average almost 50 points lower in Josh Hamilton.  On the big picture Miguel Cabrera had a better season than Josh Hamilton enroute to the AL MVP Award by leading the league in average, homeruns, and runs batted in, but Hamilton's at bats were more efficient than Miguel Cabrera.  A lot of the other names on this list shocked me too.  I would expect to run into names like Ryan Braun, Prince Fielder, Buster Posey, and even Mike Trout, so count me surprised with the results. 

Let's take a specific look at the seasons of Hunter (an unexpected top ten) versus Braun.
There isn't a doubt in mind based on the usage of old school (runs, hitters, homeruns, RBIs, and average), new school (on based percentage, slugging, and OPS), and the ever popular sabermetrics' (wins above replacement player) stats that Ryan Braun was a better hitter in 2012 than Tori Hunter.  So why was Ryan Braun's OBI% more than 3 points worse than Tori Hunter.  Does this mean Hunter is more efficient?  Dare we say Hunter is more clutch than Braun?  For years, the mere mention of clutch baseball hitters has caused panic and widespread mockery through the statistics community by claiming the small sample sizes in October are not enough to demonstrate a player's ability to be clutch.  However, when a runner is a base, a player is demanded to be clutch and what better way to show if you are clutch in an at-bat than by driving in runs?  I believe OBI%, in its current form, best demonstrates the effective of driving in runners on base.

Yet with any hypothesis there are potential flaws, consider the following:
1) OBI% doesn't effectively judge the speed of a baseball player
    > Consider this scenario, Ryan Bruan & Tori Hunter hit identical singles down to right field with a runner on second base.  For Hunter that runner is Mike Trout and for Braun that runner is the pitcher.  Who is most likely to convert in their OBI% opportunity? 

2) OBI% doesn't consider what base the runner is on during the at bat
     > Simply put, if you have 20 at bats with a runner on third versus 20 at bats with a runner on first, what scenario do you expect will result in a higher OBI%?

3) Doesn't account for gameplay scenario
     > I am not a fan of bunting, because it gives up a free out, but if a team employs a small ball strategy, your OBI% will be punished despite what could be considered an effective out in the outcome of the game. 

4) Finally, doesn't account for the game's score
    > Is it clutch to increase your OBI% during a 10-1 game or a 3-2 game?  To have statistic truly identify batter efficiency or clutchness, the unique gameday scenario needs to be outlined.

As I outline in future articles, I hope to prove things like OBI% defines that clutch gene, but with all types of baseball data, it takes more data, more studying, and more watching to truly understand what you have.  In the end, there is few things better than engrossing yourself in baseball, so I am excited to embark on this journey.