On the usage of statistics on the OH

[skanar]

skanar: How about this:

Assume the batter has already made one out, thereby creating a .000/.000/.000 line. A walk will make his OPS .500, a single 1.000, a double 1.500, a triple 2.000, and a home run 2.500, corresponding to the 1-5 values I listed above. Two previous outs would make the OPSs .333, .667, 1.000, 1.333, and 1.667. I would imagine that scale holds true for any number of previous outs made.

Yes, these calculations are true, for the first successful at-bat with any number of previous outs. But if the batter has, say, a single and an out as his previous at-bats (OPS 1.000), new OPS results will be 1.167, 1.333, 1.667, 2.000, 2.333.

Actually, I think this brand of analysis reveals one of the theoretical weaknesses of OPS: namely, it's a cobbled together estimation tool that adds two more-or-less completely unrelated statistics. OBP and SLG have different denominators, they measure different things, they're on different scales. The fact that, when added together, they produce a useful metric is mostly accidental and (in my view) quite nice.

OPS is useful statistical shorthand: it says a lot in one number, it's readily grokkable, and it can be explained to anyone who is familiar with baseball scorekeeping ("well, Grandpa, OBP is how often the batter doesn't make an out and SLG is like batting average, but extra-base hits count extra"). But it's not the be-all, end-all of offensive stats.

[TakebackOPACY]

eb, fantastic post!

A problem (in my opinion) is that we've used OPS so much that it comes across like we're treating it as the end all definitive stat for offensive prowess (even if we know better). Lately, I've cringed every time I've used it.

I'll often post it along with a full stat line (avg/obp/slg/ops) so that folks can see it in a more full context. And I've thought about leaving OPS out of that line because if we take a moment, the ingredients to OPS tell us more than OPS. But people want that one quick number. OPS is palatable, and better than batting average, so there it is.

Lately, I've passed on using wOBA simply because I know it's less familiar and less palatable for folks. On a separate note, I don't like the name of the stat, 'weighted on base average'. The way the mind completes things, it sounds too much like 'weighted on base percentage', and therefore the mind doesn't get hold of what this stat is going to tell you. I know, I know, the name wOBA is factually accurate; but my mind goes elsewhere when I hear it. Maybe something like 'weighted offensive production average' would be okay.

[SilentJames]

I agree wOBA is a much better measure than OPS. I even started a couple of threads where I swore I would use wOBA more often. (Like this one: http://forum.orioleshangout.com/forums/showthread.php?t=106755&highlight=wOBA) But, in many instances I just like using BB-ref better, and wOBA isn't available there. I'm guessing we'll use wOBA more and more often over the next few years, but I don't think the world will end if we use OPS as a readily available, easily calculable stat to estimate offensive prowess, despite its flaws, which you outlined correctly.

A primer on wOBA I found on fangraphs summarizes my frustration with some of these advanced metrics perfectly.

http://www.fangraphs.com/blogs/index.php/the-joy-of-woba/

In comparing Ludwick and H. Ramirez Ludwick had an OPS about .026 higher than Hanley. But Hanley got on base at a 400 clip while Ludwick's OPS was bolstered by much higher SLG.

using the wOBA method we find that Ramirez had a wOBA of .003 HIGHER than Ludwick.

OR - I could have just looked at the split of SLG and OBP and seen that Hanley gets on base at a much better rate and, if I put a lot of stock into that, would determine he is the better player.

It seems to me that a lot of advanced stats are trying to create one "golden number" to use as the standard evaluator.

I didn't need to get to wOBA, I just needed to look at the two numbers that make up his OPS and make a determination.

[Tony-OH]

The problem with asking people to use more and more sabermetrics is that most people want to enjoy the game of baseball and be able to talk about it without feeling like they have to go the calculus class.

OPS is a pretty good little quick indicator of a hitter's value. It's easy to calculate and it's better than just using the old average, home runs RBis of the pre-90s.

Little by little different stats will be accepted as more mainstream, but what some of you need to realize is that it doesn't make a person stupid if they choose to use stats like OPS or WAR to make a point.

There is value in Sabermetrics, but I've always kind of felt like the hard core sabermetrics and the arrogant attitude held by some that use them are the revenge of the nerds factor.

Anyone who uses math and thinks they've found the holy grail for predictive analysis forgets one thing. The game is played by human beings which mean there are too many things that can not be determined through statistics in a vacuum.

Any kind of decent stat can be a tool and it would be foolish to just discard them if you are a baseball executive, manager, coach or scout. But in the end, they are nothing more than a tool and although the more detailed data available certainly can be used for good, I reject the attitude that they should be the only thing used to discuss players' value.

This is the Hangout where all views are tolerated, but this board is not going to be turned into some Baseball Prospectus nerd club where people will be denigrated or looked down upon if they use OPS instead of wOBA.

[JERSEYORIOLE]

The problem with asking people to use more and more sabermetrics is that most people want to enjoy the game of baseball and be able to talk about it without feeling like they have to go the calculus class.

OPS is a pretty good little quick indicator of a hitter's value. It's easy to calculate and it's better than just using the old average, home runs RBis of the pre-90s.

Little by little different stats will be accepted as more mainstream, but what some of you need to realize is that it doesn't make a person stupid if they choose to use stats like OPS or WAR to make a point.

There is value in Sabermetrics, but I've always kind of felt like the hard core sabermetrics and the arrogant attitude held by some that use them are the revenge of the nerds factor.

Anyone who uses math and thinks they've found the holy grail for predictive analysis forgets one thing. The game is played by human beings which mean there are too many things that can not be determined through statistics in a vacuum.

Any kind of decent stat can be a tool and it would be foolish to just discard them if you are a baseball executive, manager, coach or scout. But in the end, they are nothing more than a tool and although the more detailed data available certainly can be used for good, I reject the attitude that they should be the only thing used to discuss players' value.

This is the Hangout where all views are tolerated, but this board is not going to be turned into some Baseball Prospectus nerd club where people will be denigrated or looked down upon if they use OPS instead of wOBA.

Tony, this could be the best post I have ever read on this site.

Sabermertics are a tool. One that any good executive should utilize. However, they are not a perfect science in dermining a players value, as much as some people on this site would like you to believe. The disturbing part is when a poster states an opinion that he thinks player A is better then Player B. That is his opinon by watching them play. This post is almost certainly followed by a poster that is sitting in front of his computer, pocket protector and all, working and reworking numbers to try and "prove" the original poster wrong. He then vomits up his stats and presents it as if it is clear that his player is superior. It takes the joy out of this board. The nerds have taken over.

IMO if you asked any scout or executive which he would rather have, a 20 game accumulation of a players stats or WATCH that player play for 20 games, they would rather watch the player play.

Lets say a country that does not have baseball

[Hank Scorpio]

Tony, this could be the best post I have ever read on this site.

Sabermertics are a tool. One that any good executive should utilize. However, they are not a perfect science in dermining a players value, as much as some people on this site would like you to believe. The disturbing part is when a poster states an opinion that he thinks player A is better then Player B. That is his opinon by watching them play. This post is almost certainly followed by a poster that is sitting in front of his computer, pocket protector and all, working and reworking numbers to try and "prove" the original poster wrong. He then vomits up his stats and presents it as if it is clear that his player is superior. It takes the joy out of this board. The nerds have taken over.

IMO if you asked any scout or executive which he would rather have, a 20 game accumulation of a players stats or WATCH that player play for 20 games, they would rather watch the player play.

Lets say a country that does not have baseball

I don't disagree with you or Tony. I just don't get why it's gotta be so confrontational. The "nerd" has a right to his or her opinion just like the casual fan does, no?

It's actually possible to disagree without getting into an internet fight about it. That gets lost around here way too much, IMO.

[SilentJames]

I don't disagree with you or Tony. I just don't get why it's gotta be so confrontational. The "nerd" has a right to his or her opinion just like the casual fan does, no?

It's actually possible to disagree without getting into an internet fight about it. That gets lost around here way too much, IMO.

Being the self-appointed Duke of Nerds in my personal life I must speak on this.

Sometimes saber-guys get REALLY defensive about their stats and to disagree with them automatically means that you are somehow ignorant of their greater knowledge.

I know that I have had many interactions like this with some of the more cultish-devotees of saber.

I for one am glad that these interactions seem to be few and far between here.

And you and I both know that the internet fights usually are instigated by two, maybe three posters on a regular basis.

[ChrisAF79]

Tony, this could be the best post I have ever read on this site.

Sabermertics are a tool. One that any good executive should utilize. However, they are not a perfect science in dermining a players value, as much as some people on this site would like you to believe. The disturbing part is when a poster states an opinion that he thinks player A is better then Player B. That is his opinon by watching them play. This post is almost certainly followed by a poster that is sitting in front of his computer, pocket protector and all, working and reworking numbers to try and "prove" the original poster wrong. He then vomits up his stats and presents it as if it is clear that his player is superior. It takes the joy out of this board. The nerds have taken over.

IMO if you asked any scout or executive which he would rather have, a 20 game accumulation of a players stats or WATCH that player play for 20 games, they would rather watch the player play.

Lets say a country that does not have baseball

Joe Morgan, is that you?

[Tony-OH]

I don't disagree with you or Tony. I just don't get why it's gotta be so confrontational. The "nerd" has a right to his or her opinion just like the casual fan does, no?

It's actually possible to disagree without getting into an internet fight about it. That gets lost around here way too much, IMO.

Yep,I agree. That's why I said all views here should be tolerated and discussed. I use and enjoy a lot of different stats and I've learned a lot from the sabermetric crowd. They have some real value. However, some in that crowd have that arrogant attitude towards those who might not spend the time or want to spend the time to learn every new fangle stat they can get their hands on. There's nothing wrong with those who do, nor would I ever suggest the "nerd" crowd doesn't have a place here, but in the end, the Hangout should be a place where everyone can discuss players without having someone else tell them what stats they must use to be taken seriously.

[waroriole]

I don't disagree with you or Tony. I just don't get why it's gotta be so confrontational. The "nerd" has a right to his or her opinion just like the casual fan does, no?

It's actually possible to disagree without getting into an internet fight about it. That gets lost around here way too much, IMO.

I was going to say the same thing, but not as well.

Also, I haven't really see any sabermetric guys on OH that act Tony was describing.

[Hank Scorpio]

I was going to say the same thing, but not as well.

Also, I haven't really see any sabermetric guys on OH that act Tony was describing.

Me not speak good either.

[Phantom]

Stats are not only a tool, they are the BEST tool for player evaluation. The eye test is a tool as well, but a far less useful one (as always, I challenge scouts or any "stat basher" to predict performance better than stat engines). Stats should not be taken as gospel, especially without context, but in the end I'd rather place too much emphasis on stats than too little.

For those who want to up the discourse, you should help out those posting BA or OPS, by quoting their post, and posting a more relevant stat. I think that would be a great way to get better information out there without requiring users to use stats they don't want to.

[KAZ97]

First, nice post, and some valid points. But there's a mistake buried here: slugging percentage is calculated with At-Bats as the denominator, not Plate Appearances. So walks and HBPs just aren't a factor for slugging - they don't bring it down. That is, it's not accurate to say that a player's SLG for a plate appearance that results in a walk is .000.

This is spot on. The major flaw in OPS is that it is not defined for certain events on an individual level. OPS is a summation of two aggregate level statistics. Might seem like a minor point, but its not.

If one was to describe the distribution of the batting average statistic, it is tempting to say that it varies between .200 and .400 for the most part. This is incorrect. Batting average (and OBP) are distributed binomially. Slugging follows a Poisson distribution. There are plenty of statistic tools to deal with those well behaved distributions. But things fall apart when you add two distributions based on different observational sets (PA and At-bats).

You don't really need to get very high tech in developing a new statistic to correct this flaw. Just figure out how much a walk represents one's ability to "slug" (do the same for sac flys and all the other differences between an at-bat and plate appearance). If its zero, then its zero. get OBP and SLG to use the same set of observations and then the combined stat is back to a Poisson distribution based upon a single set of observations.

Simply put, it all boils down to this:

Plenty of people have hit .300. No one in history has ever gotten 3/10ths of a hit.

Whether one stat is "better" than any other ultimately should be based on its strength of relationship with the ultimate stat: winning. Or maybe bi-winning, but I haven't been able to figure out what that is yet.

[MurphDogg]

IMO if you asked any scout or executive which he would rather have, a 20 game accumulation of a players stats or WATCH that player play for 20 games, they would rather watch the player play.

The problem lies with the fact that scouts can't watch a player play for 20 games and then add up all the positive and negative aspects about that player's ability in his head and reliably place him on a continuum of other Major League players. Beyond that, even if he could, 20 games is an insufficient sample size to reliably predict how good a player is. There are 20-game stretches where Luke Scott looks like a AA hitter and there are other 20-game stretches when he looks like a Hall of Famer. Statistics allow for all of this to be equalized over time.

[Hooded Viper]

This is still a game right?????