### Slugging and On Base

You’ll have to bear with me today, as 1) my subject matter is just an idea that popped in my head, not a tested formula, 2) I have plenty of ire directed towards the Houston Astros in general, Ezequiel Astacio in specific, and 3) you always have to bear with me, as I tend to be slow at times.

I just finished reading

Nor is one really needed. Baseball is inundated with so many meaningless stats that adding one more – even if it has meaning – would be sheer overkill. But I must say that I have noticed a flaw of sorts in the OPS statistic.

Here’s a disclaimer before I get into this: I didn’t do full-out research in this category. I scoured glossaries of sabermetric terms, and did not find the statistic that I am about to describe anywhere. However, I concede that it is a logical stat, and that it may very well exist, either in more obscurity than I unraveled or in language that I didn’t understand.

There are two flaws in adding together on base percentage and slugging percentage. First, by adding them together with no modifiers, it assumes that each statistic carries equal weight, which they do not. In fact, no two statistics in baseball are equal in value. Second, one value is out of a possible 1.000 while the other is out of a possible 4.000, making their addition a bit lopsided.

The reason OPS is so highly regarded is because of the newly realized importance of on base percentage and slugging percentage over batting average. I guess the thinking is that if you slap these numbers together, you get the guys who are good at both rising to the top, with guys who are apt at one and not the other right behind them (obviously, due to the nature of the numbers, power hitters are rated higher because of their slugging percentages, while guys who get on base often are right behind).

Isn’t there a more fitting way to put together an OPS-like stat? I realize what I am about to say also has its share of flaws, but I think it is more efficient than just slapping the numbers together. Here are the 2005 AL OPS leaders, with their adjusted stat (On Base Slugging, for lack of a better term) listed next to it.

As you can see, the mingling of OBP and Slugging in this manner shakes up the list a little but, most notably at the end with Brian Roberts. Ostensibly, this would suggest that guys with higher slugging percentages and lower on base percentages would also be favored by this system. But look at Giambi v. Manny. There is a wide discrepancy in their stats, since Giambi had an OBP .052 points higher than Manny, while Manny’s slugging was .059 points higher. Yet Manny’s superior slugging numbers don’t give him an enormous edge in OBS.

Also, a gander at David Dellucci is also telling: .367 OBP, .513 Slg., .879 OPS (14th in the AL), .587 OBS (slips in right behind Konerko, way ahead of Roberts).

Another experiment is with the 11th and 12th ranks in OPS, Michael Young and Gary Sheffield, respectively. Young had an OBP of .385 and a Slg. of .513. Sheff was worse in both categories, posting an OBP of .379 and a Slg. of .512. Yet, when the average is taken, Young ends up with an OBS of .552 while Sheff towers at .570. How can this be?

Basically, this scrutinizes every plate appearance. If on-base percentage is considered more important than batting average (because you don’t give up an out and give your team a chance to score a run), then why aren’t walks implemented in slugging percentage? You get first base, just like a single, which is the idea behind OBP.

Of course, there are flaws here, just like any stat. For instance, Young is being penalized for being a leadoff hitter, and therefore having more plate appearances. So possibly organizing these stats based on plate appearances is an option. This makes more sense when evaluating more leadoff hitters. Jeter’s OBS is .508; Chone Figgins’s is .433; David DeJesus sits at .489; and monster leadoff man Grady Sizemore is .523.

As always, I’ll be looking for ways to effectively implement this and other non-traditional stats. Up tomorrow: non-out percentage vs. on base percentage.

I just finished reading

*Moneyball*for the second time (and it may become a yearly or bi-yearly reading, considering how much more I absorbed this time around), so I’m sitting around thinking of alternative ways to evaluate baseball. There has been such an influx of new baseball knowledge in the last 20 years that it would, on the surface, seem a difficult task to find another significant offensive statistic.Nor is one really needed. Baseball is inundated with so many meaningless stats that adding one more – even if it has meaning – would be sheer overkill. But I must say that I have noticed a flaw of sorts in the OPS statistic.

Here’s a disclaimer before I get into this: I didn’t do full-out research in this category. I scoured glossaries of sabermetric terms, and did not find the statistic that I am about to describe anywhere. However, I concede that it is a logical stat, and that it may very well exist, either in more obscurity than I unraveled or in language that I didn’t understand.

There are two flaws in adding together on base percentage and slugging percentage. First, by adding them together with no modifiers, it assumes that each statistic carries equal weight, which they do not. In fact, no two statistics in baseball are equal in value. Second, one value is out of a possible 1.000 while the other is out of a possible 4.000, making their addition a bit lopsided.

The reason OPS is so highly regarded is because of the newly realized importance of on base percentage and slugging percentage over batting average. I guess the thinking is that if you slap these numbers together, you get the guys who are good at both rising to the top, with guys who are apt at one and not the other right behind them (obviously, due to the nature of the numbers, power hitters are rated higher because of their slugging percentages, while guys who get on base often are right behind).

Isn’t there a more fitting way to put together an OPS-like stat? I realize what I am about to say also has its share of flaws, but I think it is more efficient than just slapping the numbers together. Here are the 2005 AL OPS leaders, with their adjusted stat (On Base Slugging, for lack of a better term) listed next to it.

- Player OPS OBS
- A-Rod 1.031 .666
- Hafner 1.003 .652
- Ortiz 1.001 .654
- Manny .982 .645
- Giambi .975 .642
- Vlad .959 .611
- Teixiera .954 .621
- Sexson .910 .605
- Konerko .909 .592
- Roberts .903 .561

As you can see, the mingling of OBP and Slugging in this manner shakes up the list a little but, most notably at the end with Brian Roberts. Ostensibly, this would suggest that guys with higher slugging percentages and lower on base percentages would also be favored by this system. But look at Giambi v. Manny. There is a wide discrepancy in their stats, since Giambi had an OBP .052 points higher than Manny, while Manny’s slugging was .059 points higher. Yet Manny’s superior slugging numbers don’t give him an enormous edge in OBS.

Also, a gander at David Dellucci is also telling: .367 OBP, .513 Slg., .879 OPS (14th in the AL), .587 OBS (slips in right behind Konerko, way ahead of Roberts).

Another experiment is with the 11th and 12th ranks in OPS, Michael Young and Gary Sheffield, respectively. Young had an OBP of .385 and a Slg. of .513. Sheff was worse in both categories, posting an OBP of .379 and a Slg. of .512. Yet, when the average is taken, Young ends up with an OBS of .552 while Sheff towers at .570. How can this be?

Basically, this scrutinizes every plate appearance. If on-base percentage is considered more important than batting average (because you don’t give up an out and give your team a chance to score a run), then why aren’t walks implemented in slugging percentage? You get first base, just like a single, which is the idea behind OBP.

Of course, there are flaws here, just like any stat. For instance, Young is being penalized for being a leadoff hitter, and therefore having more plate appearances. So possibly organizing these stats based on plate appearances is an option. This makes more sense when evaluating more leadoff hitters. Jeter’s OBS is .508; Chone Figgins’s is .433; David DeJesus sits at .489; and monster leadoff man Grady Sizemore is .523.

As always, I’ll be looking for ways to effectively implement this and other non-traditional stats. Up tomorrow: non-out percentage vs. on base percentage.

<< Home