OPS Discussion 2004 [Mergred Threads]

[TreAnt985]

I know some other people have probably noticed this, but, while OPS is probably one of the best indicators of players' overall batting performances, it is slightly biased towards sluggers.
Although most will argue that slugging helps the team more than getting on base in general, I'd like to see a stat that adds the two categories equally: since OBP is on a scale from 0 to 1.000 and SLG is from 0 to 4.000, those players with higher slugging marks would end up with disproportionally high OPS.
Like I said, I'm not trying to argue whether slugging or getting on base is more important, but I'd like to see something that uses OBP and SLG in a proportional manner (there may be something like this out there, but I'm not big into sabermetrics).
To make the two 'proportional' to each other when adding them I think one should either multiply the OBP by 4 (to put everything on a 0 to 4 scale) or divide SLG by 4 (to put everything on a 0 to 1 scale), and then add the two.
I thought of this because it seems like the guys who get on-base often don't get nearly as much 'statistical respect' as those who hit lots of homers with lower OBP (of course players like Bonds, Ruth, and Williams don't count ;) )
Does anyone have any thoughts about how to show a player's on-base and slugging performances with equal weight in one stat without using complicated formulas?

[TreAnt985]

Also, sorry if this has been discussed before, but I couldn't search for OPS (which would be the only real way I could find another thread on this) since it's less than 4 letters long. :)

[Skip]

While SLG is on a scale 4 times larger than OBP, practical numbers dont differ by that much. As ballpark figures, .400 OBP and .650 SLG are considered excellent, so the difference really isnt a factor of 4.

I believe many have suggested altering the OPS formula to weight OBP higher, the one I remember most (Neyer's I think) is to use 1.4*OBP + SLG.

[TreAnt985]

While SLG is on a scale 4 times larger than OBP, practical numbers dont differ by that much. As ballpark figures, .400 OBP and .650 SLG are considered excellent, so the difference really isnt a factor of 4.

I believe many have suggested altering the OPS formula to weight OBP higher, the one I remember most (Neyer's I think) is to use 1.4*OBP + SLG.
I was hoping someone would bring that up--I knew that SLG wasn't anywhere near OBP*4, but I wasn't sure what the 'average' ratio was....

[TreAnt985]

Using the SBE (1876 through 2002), the ML OBP is .327 while slugging is .379. However, the difference between the two is much greater than this for recent years:
1997: .419 .337 (1.24:1 SLG:OBP)
1998: .420 .335 (1.25:1 SLG:OBP)
1999: .434 .345 (1.26:1 SLG:OBP)
2000: .437 .345 (1.27:1 SLG:OBP)
2001: .427 .332 (1.29:1 SLG:OBP)
2002: .417 .331 (1.26:1 SLG:OBP)

Maybe something like 1.25*OBP + SLG would be a good indicator for recent seasons?

[Skip]

While the 1.25 factor seems to fit recent historical overall averages, I expect the 1.4 number from Neyer or whoever comes from the wider statistical range (and without doing the math, variance) in SLG scores. Certainly for the best players (top quartile measured somehow) I think the ratio is higher.

I dont recall seeing a version of this sort of weighted OPS scaled down to the standard OPS values that (at least saber-esque) fans have accepted. That is, 1.000 OPS is accepted as a pretty good threshold of mucho quality with traditional OPS. Changing the formula changes the scale. 1.100 (or whatever) as the threshold isn't as sexy - just like anything but .300 for batting, 300 wins for pitchers, 20 wins for pitchers, 500 HR (going going gone) don't hit that nice even number type milestone.

[Ytown Tribe fan]

Considering that the Runs Created formula is based more closely on OBA TIMES SLG, maybe that would be a truer expression of batting ability.

The reason OPS is so popular is simple laziness. It is much easier for most folks to simply add the OBA and SLG than to multiply them.

Now, OPS minus BA might be a little better, since it doesn't double-count BA (as OPS does), but it involves a third step of subtraction and doesn't work for the slothful either.

Nope -- OPS is easy, if biased and inaccurate. And flawed.

[sweaver]

And so is everything else, to one degree or another.

Aaron Gleeman uses his GPA, the "Gleeman Performance Average," which is OBA*1.7 + SLG divided by 4. It comes out on a scale comparable to batting average. Of course, it's got the extra math, too.

[Wolf Hopper]

Here's an old discussion on this, sorta: http://www.netshrine.com/vbulletin2/showthread.php?t=4489

[huskerdru]

Nope -- OPS is easy, if biased and inaccurate. And flawed.

It is true, but the flaw is not that great. The increase in precision from multiplying weighted OBP by SLG is insightful, but for typical in-season or recent season evaluation, OPS does a pretty damn good job. There is no question that it favors sluggers more than OB guys, which is why we look at the individual numbers, too.

[KCBOOMER]

I think you are wrong say people are being lazy in their current calculation of OPS. The addition method of OPS is user friendly and it generates a number that humans can relate to (the closer you are to one or over it is good). Yes, it could be tweaked for accuracy, but you rarely are going to give any additional meaning. BP's EqA is great but mere mortals aren't going to calculate that quickly in their heads any time soon. Same with the GPA.

And neither of those "micrometers" really improve most people understanding of a hitter's value over OPS.

[nyy26wc]

I think you are wrong say people are being lazy in their current calculation of OPS. The addition method of OPS is user friendly and it generates a number that humans can relate to (the closer you are to one or over it is good). Yes, it could be tweaked for accuracy, but you rarely are going to give any additional meaning. BP's EqA is great but mere mortals aren't going to calculate that quickly in their heads any time soon. Same with the GPA.

People not calculating things in their heads isn't a valid reason.

Pick any stat and nobody's calculating it in their heads. For example, AVG isn't entrenched in the world because everyone is going around dividing H by AB on their own. It's entrenched because that's the number people are given and people are given the results for them.

And neither of those "micrometers" really improve most people understanding of a hitter's value over OPS.

Agreed.

[Rajah]

I think it came up in Moneyball that the A's measured a point of On Base Percentage being 3 times as valuable as a point of Slugging Percentage.

I quote:
Not long after he arrived in Oakland, Paul (DePodesta, Beane's assistant and chief stat guy for the few who don't know) asked himself a question: what was the relative importance of on-base and slugging percentage? His answer began with a thought experiment: if a team had an on-base percentage of 1.000--that is, every hitter got on base--how many runs would it score? An infinite number of runs, since the team would never make an out. If a team had a slugging percentage of 1.000--meaning it gained a base for each hitter that came to the plate--how many runs would it score? That depended on how it was achieved, but it would typically be a lot less than an infinite number. A team might send 4 hitters to the plate in an inning, for instance. The first man hits a home run, the next three make outs. Four plate appearances have produced four total bases and thus a slugging percentage of 1.000 and yet have scored only one run in the inning.
(skipping ahead)
...OPS was the simple addition of on-base and slugging percentages. Crude as it was, it was a much better indicator than any other offensive statistic of the number of runs a team would score. Simply adding the two statistics together, however, implied they were of equal importance. If the goal was to raise a team's OPS, an extra point of on-base was as good as an extra point of slugging.
Before his thought experiment Paul had felt uneasy about this crude assumption; Now he saw that the assumption was absurd. An extra point of on-base percentage was clearly more valuable than an extra point of slugging percentage--but by how much? He proceeded to tinker with his own version of Bill James's "Runs Created" formula. When he was finished, he had a model for predicting run production that was more accurate than any he knew of. In his model an extra point of on-base percentage was worth three times an extra point of slugging percentage.
Paul's argument was radical even by sabermetric standards. Bill James and others had stressed the importance of on-base percentage, but even they didn't think it was worth three times as much as slugging. Most offensive models assumed that an extra point of on-base percentage was worth, at most, one and a half times an extra point of slugging percentage. In major league baseball itself, where on-base percentage was not nearly so highly valued as it was by sabermeticians, Paul's argument was practically heresy.

End quote.

Hence the A's not always successful search for guys who don't make outs. They actually posted a below league average on base percentage last season. On the other hand, with a little luck, they should greatly improve that figure based on last season:
Durazo .374 OBP, Chavez .350, Kotsay .343, Hatteberg .342, Byrnes .333, Melhuse .372.

They still have lineup holes like .313 Ellis, . 261 Dye and Damian Miller of a .310. But now it looks like the A's want good defensive guys who don't make outs.

[Wolf Hopper]

Just came across this: http://mysite.verizon.net/vze1yy5u/

[gyb13]

it's funny that he didn't bother to change the title of the html file....

[Rajah]

That article makes sense to me. 3 seemed like too high a multiplier. Not sure if 1.64 works (I only read the executive summary. High level stats still gives me nightmares), seems a bit low still.

[Fatwater Fewl]

According to an email I received from Richard Lederer, in response to a question that I asked him about a funky version of OPS that I was playing with, he found that (in looking at the numbers from 1992-2001) the additive version of OPS had a .953 correlation to runs scored while the multiplicative version had a .956 correlation to runs scored (mine came in at .955, so I've dropped it as it provides no improvement). So, neither are perfect but both are pretty damn good in terms of assessing a player's ability to produce. Lederer discusses OPS in his latest Weekend Baseball BEAT (http://www.baseballbeat.blogspot.com/2004_02_01_baseballbeat_archive.html#1076135761923 67733).

On seeing this thread, I was going to post the very link that Wolf Hopper did. Glad to see it's up there. Here's another bit (http://www.corante.com/transition/archives/001684.html) on the same sort of thing (Rivers responds to it near the bottom of the page). Also, tangotiger of Baseball Primer has two articles on OPS here (http://baseballprimer.com/articles/tangotigre_2003-05-20_0.shtml) and here (http://baseballprimer.com/articles/tangotigre_2003-05-27_0.shtml).

And, finally, an item of historical interest: Keith Woolner found Branch Rickey was looking at OBA in 1954. The First Stathead (http://www.stathead.com/bbeng/woolner/brickey.htm)

[steffens]

Chris Kahrl says we shouldn't use it (http://premium.baseballprospectus.com/article.php?articleid=2640) -- there's better stuff out there, like EqA and MLVr.

I don't know, I think I would have a hard time giving it up. While it's true better metrics exist, OPS still has significant value, and there's nothing wrong with stating things simply sometimes, as long as you realize the limits of what you are saying.

[sweaver]

Some of what I have to say about Chris Kahrl can't be printed on this site. :D But, OPS is valuable because it is a "quicker" metric, adding OBP and SLG. Equivalent runs and all that stuff are cool, but harder to calculate.

[lonelybrewerfan]

Interesting, I guess it depends on what you are looking for. I think we owe it to ourselves as dedicated fans we always should be looking for better means of analysis. OPS may be good, but it doesnt tell the whole story.

I will side with Chris Kahrl on this one, I guess Sweaver and I are just on different wavelenghts.

[SmedIndy]

Nothing tells the whole story - but OPS is a good, quick metric. I don't have time to sit around and calculate equivalent runs based when discussing players on the fly. Kahrl needs to chill. It's not like we're basing everything on BA and RBI.

[Jim Rice]

Chris Kahrl says we shouldn't use it (http://premium.baseballprospectus.com/article.php?articleid=2640) -- there's better stuff out there, like EqA and MLVr.

I don't know, I think I would have a hard time giving it up. While it's true better metrics exist, OPS still has significant value, and there's nothing wrong with stating things simply sometimes, as long as you realize the limits of what you are saying.

I wouldn't pretend to know enough about all the bells and whistles inherent in EqA , et al, to dispute whether or not they are, in fact, as wonderful as Kahrl says they are. I'm sure they're wonderful. But, at the same time, they are not the type of stats that regular fans can grasp readily or that announcers can discuss intelligently. They seem to be much better suited to the brainiacs of SABR who love to get into the theoretical and statistical minutiae. OPS is something a normal fan can understand easily, while still not being horribly misled about a player's ability by looking at just his average or RBI total. That's far more functional than the esoteric mumbo-jumbo BP spins out.

Besides, Kahrl has an obvious bias I have to discount - he writes for BP, the people who invented, publish, and make money from the very stats he says are the best. I wouldn't expect him to say otherwise - no good capitalist would - but I do have to take into account that he has a vested interest in the stats he says are so wonderful.

Funny how BP feels that OPS is outdated the very year Topps started putting OPS on the back of their baseball cards. "Hey, if we can discount the validity of OPS and make EqA more mainstream, maybe we can cut a deal with Topps too!" If that's their strategy, it's not a bad one, but they've got a lot of work in front of them. When I read Kahrl's work, my brain translates it all into something that sounds like Charlies Brown's teacher - a string of mumbled words that do nothing but annoyingly interrupt something I had hoped would be entertaining.

[Craig S.]

Take it easy, Chris - it's not like we're basing a player's ability on whether or not he hits the 100-RBI mark. Using OPS might not always be the best measure, but it's usually a decent indicator of a player's ability.

So is EqA an absolutely perfect measure, or will a similar article be written about its shortcomings a year from now?

[lonelybrewerfan]

Jim Rice: If you think that Chris Kahrl is promoting other stats in lieu of OPS so that it can go on a baseball card, then I have some swampland I would loooooove to sell you.

Prospectus is about challenging the status quo. OPS is a good metric but the people at Prospectus thought they could find a better one, and in my opinion they did. That isn't sell agrandizement, it's good research.

[Jim Rice]

Jim Rice: If you think that Chris Kahrl is promoting other stats in lieu of OPS so that it can go on a baseball card, then I have some swampland I would loooooove to sell you.

Prospectus is about challenging the status quo. OPS is a good metric but the people at Prospectus thought they could find a better one, and in my opinion they did. That isn't sell agrandizement, it's good research.

I don't think that's necessarily the case, I just said the timing was an interesting coincidence. For years, stat-heads have argued that most of the mainstream stats are crap and that OPS specifically is a much better alternative for evaluating a hitter. Now that they have what they've been begging for for a few years - OPS on just about the most mainstream thing in baseball, the back of Topps cards - Kahrl comes along and bashes it.

My point is that most fans, like me, follow baseball for enjoyment, and guys like Kahrl are missing that point. About 95% of the people who follow baseball and read stats want something easy to understand that will help them enjoy the game. Reading the latest argument between the stats geeks ("My stat is best." "No, MY stat is best!"), when all of those stats require a LOT more math than the average fan cares to bother with, is not what fans want. It's the very reason I refuse to subscribe to BP - they take themselves and what is supposed to be a game WAY too seriously. They suck the fun out of baseball and turn it into not just a math class, but a petty one where the teachers all argue about which of their formulas is best.

[lonelybrewerfan]

I couldn't disagree more. BP is pushing the envelope by bringing other disiplines (math, economics etc.) into baseball. By saying that BP is sucking the fun out of baseball is a purely subjective opinion, one you are entitled to but I think is misinformed. How can extra information as well as providing another way to look at the game not fun?

Understanding all of the math that goes into these statistics is not critical to understanding the metric itself. All you need to know is what the metric is measuring and then find a baseline for what is a good number and a bad number. Once that has been established you are good to go.

[steffens]

I think what BP does is great, and I love reading their material and find it worth subscribing to. And it's great to come up with better, more meaningful metrics. But do we have to discard OPS in the process? I don't think so. It, too, is a meaningful measure, and an easy one to use.

[rejtable]

Plain old-vanilla OPS is fine and dandy for the big bulk of arguments at your local watering hole. Heck, if all/most (any?) of the arguments at the local watering hole used OPS as the base we'd be a much further advanced race of beings. ;)

Nothing wrong with the next level up metrics either. If the granularity of the given discussion warrants it, great. I tend to take a peak at a variety of metrics in most discussions, but plain old vanilla OPS will get you much/most of the information you require a lot of the time. I think that just flat out dismissing OPS because it doesn't correlate quite as good as, say, EqA, or because at face value it doesn't account for parks or playing time is just silly. Of course it doesn't do all those things, but that by itself doesn't make it useless.

I've yet to see a full consensus on any perfect variable in measuring baseball production. Until one comes up, I'm just glad that we have so many to choose from, and I make an effort to understand the pros and cons of each.

At the end of the day, if we could just get Joe Fan converted to OPS, and to understand a line that looks like .300/.400/.500 the world would be a better place.

[Jim Rice]

I couldn't disagree more. BP is pushing the envelope by bringing other disiplines (math, economics etc.) into baseball. By saying that BP is sucking the fun out of baseball is a purely subjective opinion, one you are entitled to but I think is misinformed. How can extra information as well as providing another way to look at the game not fun?
...

I've read BP. I understand the points they are making. I agree, they are pushing the envelope. But they are points that either A) don't interest me, or B) I feel are made in such a self-serving, condescending fashion that I choose not to give them my money for the privilege of reading them. I don't happen to like the manner in which they push that envelope, and I'll agree with you again - that's a purely subjective view on my part. As such, I really can't be misinformed about the level of fun I feel BP-type analysis takes from my personal enjoyment of baseball, now can I?

I find skydiving fun, too. If you tried skydiving and came away scared to death, would it be okay for me to say, "Well, that's your right and I respect it, but you must be misinformed about how much fun skydiving really is. How can you not enjoy it? You must not have understood it." That would be pretty arrogant.

[Skip]

I like what BP has to offer most of the time and like that they are a fairly accessible group striving to improve the understanding of the game. As with anything, including NetShrine, when I find topics pushed too vociferously or too often, topics that are not of personal interest, or a tone taken that I find inappropriate, I can simply choose not to pay attention and focus on what's important to my enjoyment and understanding of baseball. This is true whether you swim in the shallow end or deep end of the baseball pool - or somewhere in between.

As with most here, I find OPS useful and am glad that it has become more mainstream. I like the more advanced stats too; in my place in the pool they work for me. There is no need to disregard OPS though, it is what it is and it's an improvement in general understanding of the game. Hence, this article is one I ignore.

[gyb13]

i think a lot of you are missing the point here. Kahrl isn't saying that OPS doesn't have a place. Like he says, it'll tell you that Bonds is great and Walker is good. But this hardly says much.

As with any statistic, you want to be able to compare two players and say who is better by that measure. Thus, if X has a .350 OBP and Y has a .349 OBP, you can say X was better at getting onbase.

The problem is that, as Kahrl showed through several examples, a comparison of close OPS figures can be misleading because OPS's are built differently - in some players, by more OBP and in others, by more SLG. Thus, one of the things he's saying is that, if you want to work with simplifications, it is better to look at a player as a 250/350/400 guy, not as a 750 OPS guy, because OPS will give you a condensed picture.

But if you want to be able to say, confidently, that X is better than Y, you can't escape not using an advanced metric, because no simple metric adjusts for a player's context (ballpark, league, etc).