Friday, November 12, 2010

What's a Good OPS?

Many fans like the traditional statistic batting average because they are so familiar with its values.  They know right away that a .300 average is excellent, a .220 average is poor and a .400 batting average is both fantastic and rare.  One problem many fans have when they are introduced to a new measure is that they don't know what values of the statistic are good and bad.  When they hear about OPS for the first time, for example, they might ask: "What's a good OPS?"  

One of the most popular features of my book Beyond Batting Average is a series of simple percentile charts which help give an idea as to which values of a statistic are good and bad by comparing them to batting average.  The table below shows the percentiles for  some of the less traditional statistics and puts them next to the equivalent percentiles for batting average.  This makes it easier for fans to grasp some of the newer statistics.  The chart includes all MLB regulars or semi-regulars with 400 or more plate appearances in 2010.

One can see from the table that the 75th percentile for batting average was .286.  This means that 75 percent of these players hit below .286 and 25 percent hit better than .286.  A player with an Isolated Power (ISO) of .197 would also be at the 75th percentile.  So, we can say that an ISO of .197 is as good as a batting average of .286. Similarly, an ISO of .081 would be bad because it's equivalent to a .243 batting average (10th percentile).  One might also call this a batting average equivalency chart.   

The statistics in the chart can be found on FanGraphs.com. They are defined as follows:

BA = Batting Average
OBP = On-Base Percentage
SLG= Slugging Average
OPS = On-Base Percentage Plus Slugging Average
ISO = Isolated Power = Number of Extra Bases Per AB = (2B + 2 x 3B + 3 x HR)/AB
BB%= Base on Balls Percentage = % of PA resulting in walks
K%= Strikeout Percentage = % of AB resulting in strikeouts

I'll provide batting average equivalency charts for more statistics in later posts.  

Table 1: Batting Average Equivalencies for 2010

5 comments:

  1. Do you have anything in your book that revolves around analyzing the value of the negative events? For example, do you analyze why a high strikeout % is bad by showing the difference in contributions to a team if you have 2 equal hitters, where one hitter has an inordinate number of strikeouts and another makes those outs by putting the ball in play?

    We have so many people and sources analyzing the positive events to show methods to calculate how valuable those events are, but there is less work done on the converse and that's where baseball is really won and lost, in that area where the game is misunderstood and neglected.

    -The Strategy Expert

    ReplyDelete
  2. I've seen linear weights studies looking at the value of strikeouts versus putting the ball in play. Putting the ball is sometimes better than a strikeout, but then doubleplays negate a lot of the advantage. If a player can consistently put the ball in play and not hit into double plays, then he would certainly have somewhat of an advantage over one who strikes out all the time. I don't think many players can do that on a consistent basis though.

    I don't think many games are won and lost on that basis, because most players who put the ball in play also hit into double plays a lot. However, it's an interesting area to investigate further. If a team can get any advantage at all, then it's worth examining.

    ReplyDelete
  3. Well a sliver is a sliver and not much at all, but across a long spectrum of games it can matter. What's a 1% improvement in your half glass of milk? Inconsequential. But 1% in baseball is worth 1.6 games, sometimes that's the difference between winning a division and missing the playoffs. And analyzing outs may not be worth 10%, whose to say what amount it would be, but if you can get a percent or two better by understanding one game of the area better and then do the same for others, you can pick up a couple percent or up to 10% or what not. 10% is a small ratio yet a powerful 16 games, so anything beyond 1% is really significant.

    Let's look at good hitting events, what's better a player that has a slight BA average to SLG or SLG versus BA? Well if the difference in number sets is 5 points or 20 points, you have a tiny difference versus a large one, but nevertheless a person can come up with an analysis of all those events and show how its important to pick correctly if you can. And if a batter hits .333, well then 1/3 times he gets a hit and you are evaluating him on that score, but if he is out 2/3 times then that's TWICE the frequency, so the value in analyzing how well a player fails only has to be half as relevant to equate to the same. So you can't look at the difference of strikeouts versus double plays and say the events aren't noticeably different, cause whatever the difference is you double it due to the FREQUENCY of those events being twice as plentiful as the hitting events.

    And EVERY player consistently does avoid DPs or Ks versus other events or vice versa. It doesn't mean a player has to always trade one for the other, we are looking at stats and some guys strike out 9% of the time, and some 20% of the time, and those other percents go to other types of outs. Every single player has their own unique signature of distribution and some guys compared against others have vastly different outlooks.

    What if you had two identical hitters that had the same BA and SLG and OPS? Whose better? Well most people would say the same, but analyzing the negative events would tell you the real answer unless they have identical signatures in the way they fail.

    ReplyDelete
  4. Of course, a ground out is better than strikeout. I'm questioning this statement: "that's where baseball is really won and lost"

    Do you any real data from real games to back up your theory?

    ReplyDelete
  5. meaning areas that GMs pay no attention to. Every GM reads the BA stats, those are already factored in. I'm talking about key areas that aren't being paid quality attention to.

    ReplyDelete

Sabermetrics Book

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

Blog Archive

Subscribe

501 Baseball Books

501 Baseball Books
Recommended by Tiger Tales

Stat Counter

Site Meter