Saturday, March 28, 2009

An Old School Saber Stat

Analysts are always trying to invent the perfect statistic for defining total offensive production - a technical version of Runs Created, Batting Runs, Base Runs, VORP, etc. I like those types of advanced formulas because they are more accurate measures of offensive contribution than simpler statistics. However, many people find them complicated and avoid them. The simplicity of the OPS statistic is the main reason it has been able to penetrate the mainstream audiences to an extent.

OPS has some limitations. It gives too little weight to walks and to extra base hits and too much weight to singles. It also gives no credit to players for stolen bases. I think the biggest presentation problem is that it does not measure anything concrete. On Base Percentage is the percentage of plate appearances which result in a batter reaching base. Slugging Percentage (which is misnamed because it's really an average rather than a percentage) is the average number of total bases per at bat. So, OPS is the sum of a percentage and an average which is kind of crazy. It works reasonably well in ranking hitters on total hitting production but it has no real definition.

We can not get around the inaccurate weighting of walks and hits without using a more advanced statistic. However there is relatively simple measure which addresses stolen bases and which is more concrete than OPS. The Total Average (TA) statistic presented by sports writer Tom Boswell in 1981 is the ratio of total bases gained to outs. It is calculated as follows:

TA = (TB + BB + HBP + SB)/(AB-H+CS+GIDP)

(It should be noted that Barry Codell introduced a Base-out percentage in the Baseball Research Journal in 1979 which was the same as TA except that it included sacrifice flies and bunts in both the numerator and denominator).

Jim Albert and Jay Bennet in their excellent book Curve Ball showed that Total Average is slightly more highly correlated with runs than OPS is. That is because of the inclusion of stolen bases and caught stealing. It has the opposite weighting problem of OPS in that it gives to much credit for walks and extra base hits. Still, it is at least as accurate as OPS and more concrete without much if any loss in simplicity.

The Total Average leader between 2007-2008 was Albert Pujols at1.142. In other words, Pujols gained 1.142 bases for every out. Chipper Jones (1.136) and Alex Rodriguez (1.131) were second and third. We would get the same three leaders using OPS. The top three current Tigers were Magglio Ordonez (.937), Miguel Cabrera (.929) and Curtis Granderson (.923). Ordonez ranked 22nd in the Majors, Cabrera 26th and Granderson 28th. The Table below shows the Total Averages in 2007-2008 for players in the 2009 Tigers starting line-up.

Table 1: Totals Averages in 2007-2008 for Tigers 2009 Starting Line-up

Player

PA

OPS

TA

Ordonez

1301

.952

.937

Cabrera

1364

.926

.929

Granderson

1305

.887

.923

Guillen

1119

.839

.816

Sheffield

1075

.788

.795

Polanco

1270

.807

.748

Inge

984

.682

.640

Laird

829

.673

.591

Everett

386

.600

.515

No comments:

Post a Comment

Twitter

Blog Archive

Subscribe

My Sabermetrics Book

My Sabermetrics Book
One of Baseball America's top ten books of 2010

Other Sabermetrics Books

Stat Counter