The weaknesses of OPS

[MusicMan]

As someone around here once said... it's a good club, but why would you only play with one club in your bag?

FACT:
Pat Burrell just retired with the same career OPS as Roberto Clemente.

Anyone want to argue that Pat the Bat was as good a hitter as Clemente?

[subnuclear]

Who uses just OPS anymore? Clemente's career OPS+ was 131. Burrell's 116.

[Bench]

Who uses just OPS anymore? Clemente's career OPS+ was 131. Burrell's 116.

Exactly.  This isn't 2005, MM.  Wrong club.

[Mr. Happy]

Having had the privilege seeing both play, although admittedly much more of Burrell than Clemente, you have nailed a shortcoming of OPS. Burrell had more power; Clemente hit for a much higher average; the two were similar in OBP. However, there really was no comparison. Clemente was a solid four tool player-a tremendous outfielder who had a cannon. Burrell was a power hitter who was a bit of a liability in the field. If we were choosing on the schoolyard, I take Clemente 100 out of 100 times. Clemente is a HOFer; Burrell won't even get 10% of the vote.  

[Bench]

There's another fun OPS comparison that I've seen around here, probably from Jim or Noe.  I think the comparison was that Rob Deer and Pete Rose had substantially similar career OPS?

[MusicMan]

Who uses just OPS anymore? Clemente's career OPS+ was 131. Burrell's 116.

Your average telecast/fan discussion still leads with the holy triumvirate of AVG/HR/RBI.  OPS is barely into a lot of people's conciousness, and OPS+ would just blow them away.

[JimR]

As someone around here once said... it's a good club, but why would you only play with one club in your bag?

FACT:
Pat Burrell just retired with the same career OPS as Roberto Clemente.

Anyone want to argue that Pat the Bat was as good a hitter as Clemente?

duh

[Bench]

Your average telecast/fan discussion still leads with the holy triumvirate of AVG/HR/RBI.  OPS is barely into a lot of people's conciousness, and OPS+ would just blow them away.

And the "educated" sports writer is still largely stuck on OPS.  Have a look at this article about why the Red Sox can afford to "not care" about having a shortstop, which pretty much only discusses the OPS of shortstops on World Series winning teams over the last ten years.  I think defense is mentioned off hand once or twice.

Also, there are a couple of paragraphs on Jed Lowrie at the top which may be of more interest to the Astros fan.   

[subnuclear]

There's another fun OPS comparison that I've seen around here, probably from Jim or Noe.  I think the comparison was that Rob Deer and Pete Rose had substantially similar career OPS?

Career OPS+
Pete Rose 118
Rob Deer 109

Had Rose stopped playing when he was 38 (which was six years longer than Deer did) instead of 45, he would blow him out of the water.

[MusicMan]

Career OPS+
Pete Rose 118
Rob Deer 109

Had Rose stopped playing when he was 38 (which was six years longer than Deer did) instead of 45, he would blow him out of the water.

Rose was all about quantity not quality.

[Fredia]

arent most guys

[Noe]

There's another fun OPS comparison that I've seen around here, probably from Jim or Noe.  I think the comparison was that Rob Deer and Pete Rose had substantially similar career OPS?

Dave Kingman and Pete Rose. Same OPS, two very different type of players/hitters. (Don't know where Deer came from, but not from me)

ON-BASE PLUS SLUGGING (OPS) (OR PRODUCTION)

Origins?

      OPS was introduced by Pete Palmer in Total Baseball in the 1980's.

What Is It?

      The first sabermetric stat to go mainstream! Not long ago, OPS was confined to the pages of Total Baseball (when it was known as "Production"). Now, it's commonly found on websites, on scoreboards, even on the back of player cards.

      OPS is calculated by adding on-base average and slugging percentage.

      OPS = OBA + SLG

      OPS is a nice, simple way of measuring complete offensive performance, mixing the ability to get on base with the ability to hit for power.

Milestones?

      OPS hasn't been around very long; few people could tell you landmark OPS' from the past. As a general rule, an OPS over 1.000 is exceptional.

What Are Its Limitations?

      OPS doesn't really tell you anything; it's just two percentages added together. It's only used to give a quick and easy idea of who the best hitters in the league are.

      Also, the weights are uneven. A .400 On-Base Average is worth more than a .400 Slugging Percentage. OPS underrates players who get on base but don't hit for power (Pete Rose), and overrates players who hit for power but don't get on base (Dave Kingman).

      And then there are the usual problems with Slugging Percentage, mentioned above.

[Bench]

Dave Kingman and Pete Rose. Same OPS, two very different type of players/hitters. (Don't know where Deer came from, but not from me)

That was me confusing Kingman and Deer (to be fair, given enough passage of time, they're pretty similar).

[Noe]

That was me confusing Kingman and Deer (to be fair, given enough passage of time, they're pretty similar).

The comparison, if you will, is that since you're combining two stats, on-base and slugging, you could get an outstanding slugger and an outstanding on-base guy and have them measure out to be the same player... and of course, they're not.

[BizidyDizidy]

The comparison, if you will, is that since you're combining two stats, on-base and slugging, you could get an outstanding slugger and an outstanding on-base guy and have them measure out to be the same player... and of course, they're not.

I don't see that as a bad thing - the point of the metric isn't to figure out what players are most similar too each other, just which are more or less valuable to scoring runs. The other objections (OBP more important than slugging, doesn't capture fielding/baserunning, etc.) make more sense to me.

[HudsonHawk]

I don't see that as a bad thing - the point of the metric isn't to figure out what players are most similar too each other, just which are more or less valuable to scoring runs. The other objections (OBP more important than slugging, doesn't capture fielding/baserunning, etc.) make more sense to me.

My biggest gripe with OPS is it serves no purpose for anything other than arranging names on paper according to OPS.  Why would you add OBA and SLG?  Why not simply look at both and get more information?  OPS gives you less information.

[BizidyDizidy]

My biggest gripe with OPS is it serves no purpose for anything other than arranging names on paper according to OPS.  Why would you add OBA and SLG?  Why not simply look at both and get more information?

What's the point of looking at OBP and SLG? Why not just have a pitch-by-pitch list of every at-bat of the season?

[EasTexAstro]

What's the point of looking at OBP and SLG? Why not just have a pitch-by-pitch list of every at-bat of the season?

You don't?

[HudsonHawk]

What's the point of looking at OBP and SLG? Why not just have a pitch-by-pitch list of every at-bat of the season?

OBP and SLG are not obscur stats.  They are readily available for every player at any time.  They are not cumbersome or difficult to interpret.  Furthermore, you cannot calculate OPS *without* first knowing OBP and SLG.  Why add them together to give you less clarity of a player's performance?

[BizidyDizidy]

OBP and SLG are not obscur stats.  They are readily available for every player at any time.  They are not cumbersome or difficult to interpret.  Furthermore, you cannot calculate OPS *without* first knowing OBP and SLG.  Why add them together to give you less clarity of a player's performance?

Doesn't all of the above apply to the play by play stats

[Ron Brand]

When Thorn and Palmer came up with it in the early 80s, OPS was a handy little thing to use in the hope that it would get people away from using RBIs and AVG to measure importance with respect to scoring runs. Joe Fan in 1984 didn't get any of this exotic stuff. It's outdated and has been improved upon with several different metrics. It's not entirely worthless, but there's much better ways to go.

[HudsonHawk]

Doesn't all of the above apply to the play by play stats

No because average stats measure a particular thing...how often does he reach base...what's the relative value of his base hits...and the counting stats measure...well counts through a particular number of at bats. 

Sure you could look at every stat, and baseball professionals do.  They analyze a whole hell of a lot more than just OBP and SLG.  Yet statgeeks piss and moan about people who don't use OPS somehow being behind the curve.  Which is comical.

[Ron Brand]

There's statgeeks relying on and pimping OPS? That's pretty sad.

[EasTexAstro]

I thought HH led in Opinions per Subject?

[Bench]

I thought HH led in Opinions per Subject?

Is infallibility an average stat or a counting stat?

[Noe]

My biggest gripe with OPS is it serves no purpose for anything other than arranging names on paper according to OPS.  Why would you add OBA and SLG?  Why not simply look at both and get more information?  OPS gives you less information.

*DING, DING, DING*! Con-freaking-cur! I like what the opinion is on the link provided says: "OPS doesn't really tell you anything; it's just two percentages added together."  Eggszactly!

[Arky Vaughan]

It's funny how people arguing how misleading OPS is tend themselves to mislead about how misleading OPS is.

Maybe I just don't read as much about baseball statistics as others here, but I don't recall reading anything about OPS being a way to determine whether two guys who both have an .800 OPS are similar players. One guy may have a .400 OBP and .400 slugging, the other guy may have a .300 OBP and .500 slugging, and those would obviously be different kinds of hitters. What OPS indicates is that they may be of comparable value, not that they are similar or interchangeable players. Nor have I read any articles using raw OPS to compare the values of two players, one of whom spent the bulk of his career playing in a neutral home park in the '60s and the other of whom spent the bulk of his career in the 2000s, with half that time in an extreme hitter's park. If you want to use a driver to complete a six-foot putt, be my guest, but don't blame the driver when you miss.

[Arky Vaughan]

A house cost $20,000 in 1980. A compact car costs $20,000 now. Anyone want to argue that a car is worth as much as house? Does this make dollars an inappropriate medium for comparing prices?

[Arky Vaughan]

No because average stats measure a particular thing...how often does he reach base...what's the relative value of his base hits...and the counting stats measure...well counts through a particular number of at bats.  

Sure you could look at every stat, and baseball professionals do.  They analyze a whole hell of a lot more than just OBP and SLG.  Yet statgeeks piss and moan about people who don't use OPS somehow being behind the curve.  Which is comical.

I could give a damn whether people do or don't use OPS, but it sure would be nice if the Astros on the scoreboards at the ballpark and in the graphics during their broadcasts gave me more useful information about the lead-off man than how many home runs and runs batted in he has. Of course, teams have been doing that for decades, but few if any of the people exercised about the zombie army of stat geeks trying to pollute the world with OPS don't seem to have much of a problem with that useless practice, or at least they don't vocalize it.

[Arky Vaughan]

*DING, DING, DING*! Con-freaking-cur! I like what the opinion is on the link provided says: "OPS doesn't really tell you anything; it's just two percentages added together."  Eggszactly!

Well, if you told me a full-time player had a 1.000 OPS, that would tell me something. It would tell me he seems to be a valuable offensive player. It wouldn't tell me where that value comes from. But it's false to see that it doesn't really tell you anything.

[HudsonHawk]

I could give a damn whether people do or don't use OPS, but it sure would be nice if the Astros on the scoreboards at the ballpark and in the graphics during their broadcasts gave me more useful information about the lead-off man than how many home runs and runs batted in he has. Of course, teams have been doing that for decades, but few if any of the people exercised about the zombie army of stat geeks trying to pollute the world with OPS don't seem to have much of a problem with that useless practice, or at least they don't vocalize it.

Because no one with a thimble full of baseball sense uses HRs and RBIs as a measure of a leadoff hitter. Complete strawman.

[HudsonHawk]

Well, if you told me a full-time player had a 1.000 OPS, that would tell me something. It would tell me he seems to be a valuable offensive player. It wouldn't tell me where that value comes from. But it's false to see that it doesn't really tell you anything.

What does it tell you that. .400 OBP and a .600 SLG don't?

[subnuclear]

What does it tell you that. .400 OBP and a .600 SLG don't?

Nothing (and you're correct that its less information), but its one number versus two which makes it a convenient of communicating whether a batter can hit or not. There are no good hitters in any era with .600 OPS. There are no bad hitters in any era with a .900 OPS.

[JimR]

What does it tell you that. .400 OBP and a .600 SLG don't?

preach, Preacher! adding the two together for one number tells less.

[Ron Brand]

Maybe I'm wrong, and I haven't gone back to Palmer's book to verify it, but my recollection is that this was created as a shorthand way to pry people away from using AVG and RBI when assessing individual values with respect to scoring runs. It's just a snapshot kind of thing, with better correlation to creating runs. That's all. To tout it now as some Great Metric is, to me, missing the point completely.

[Limey]

I can see that OPS is valuable when evaluating players for a fantasy league draft.  Other than that, it's mostly harmless.

[HudsonHawk]

Nothing (and you're correct that its less information), but its one number versus two which makes it a convenient of communicating whether a batter can hit or not. There are no good hitters in any era with .600 OPS. There are no bad hitters in any era with a .900 OPS.

I just don't see why you would use one number for anything. It's a pointless exercise, particularly for evaluating baseball talent.

[HudsonHawk]

Maybe I'm wrong, and I haven't gone back to Palmer's book to verify it, but my recollection is that this was created as a shorthand way to pry people away from using AVG and RBI when assessing individual values with respect to scoring runs.

I think this is the fallacy. No one does this.

[Ron Brand]

I think this is the fallacy. No one does this.

They did before OPS came along. Now OPS is outdated, and fans are more educated to using better metrics.

[HudsonHawk]

They did before OPS came along. Now OPS is outdated, and fans are more educated to using better metrics.

No one I knew did. Cerrtainly no one who was charged with evaluating baseball players did.

[Mr. Happy]

No one I knew did. Cerrtainly no one who was charged with evaluating baseball players did.

Same here. I've known several scouts and cross-checkers over the years. Never heard a single one ever mention OPS.

[Ron Brand]

No one I knew did. Cerrtainly no one who was charged with evaluating baseball players did.

This. Was. Created. For. Joe. Fan. Who. Did. Not. Know. Any. Better.

Also for dimwit sportswriters. Not for scouts, serious talent evaluators, etc. This was created because any yokel you talked to would value RBI and AVG disproportionally.

[Noe]

I think Limey hit on a very unique truth... the value is found in the usage and not in the information it provides. If I were wanting to put together a lineup, I'm going with OBP and SLG to see if this is a good leadoff man or a cleanup hitter. If I am putting together a list of offensive prowess for whatever reason (fantasy baseball, salary negotiation, etc.) and it's purely a ranking system, hey... gimme OPS (I guess).

[juliogotay]

I guess this wouldn't be the crowd to suggest how great a stat WHIP is.

[HudsonHawk]

This. Was. Created. For. Joe. Fan. Who. Did. Not. Know. Any. Better.

Also for dimwit sportswriters. Not for scouts, serious talent evaluators, etc. This was created because any yokel you talked to would value RBI and AVG disproportionally.

In other words, for rotogeeks for their fantasy drafts. And that's cool. I just object to the notion that those who don't embrace it are somehow missing something.  That is has any value in the real baseball world.

[Ron Brand]

In other words, for rotogeeks for their fantasy drafts. And that's cool. I just object to the notion that those who don't embrace it are somehow missing something.  That is has any value in the real baseball world.

I don't know that it works for fantasy drafts. It's value was in coming up with something that had a better correlation to creating runs, back when 'creating runs' was met with "What?" and a dazed expression.

[Bench]

I guess this wouldn't be the crowd to suggest how great a stat WHIP is.

Why ignore HBP?

[Reuben]

It's funny. Pretty much anybody with a really high OPS, say .900 or above, you know its a very good hitter (even if you don't know whether he's a leadoff man or a cleanup man). But say you're looking at a .650 OPS. That could be a good little leadoff man (.350 OBP, .300 SLG) or a #7 or 8 hitter with lots of holes in his swing(.250 OBP, .400 SLG). One is worth a LOT more than the other.

[Ron Brand]

It's funny. Pretty much anybody with a really high OPS, say .900 or above, you know its a very good hitter (even if you don't know whether he's a leadoff man or a cleanup man). But say you're looking at a .650 OPS. That could be a good little leadoff man (.350 OBP, .300 SLG) or a #7 or 8 hitter with lots of holes in his swing(.250 OBP, .400 SLG). One is worth a LOT more than the other.

The correlation between OBP and runs scored is stronger than the correlation between OPS and runs scored, IIRC. Nevertheless, as has been stated and overstated, there's better ways to get there than by using OPS. If that's the metric you're using, you are behind the times.

[Limey]

I don't know that it works for fantasy drafts. It's value was in coming up with something that had a better correlation to creating runs, back when 'creating runs' was met with "What?" and a dazed expression.

I assume that preparation for fantasy drafts would include ranking the offensive value of players at each position.  Best way to do that is by using a single value, and OPS is the best single value stat for ranking offensive prowess.

[Arky Vaughan]

Because no one with a thimble full of baseball sense uses HRs and RBIs as a measure of a leadoff hitter. Complete strawman.

Nor does anyone with a thimble full of baseball sense think OPS means Roberto Clemente and Pat Burrell are comparable as hitters or is a good way to tell whether two players are similar or interchangeable. This whole thread is a strawman. And you should now, given your proclivity for using them.

Of course, just about every major-league team uses Avg, HR and RBI as the short-hand statistics at ball games and on TV. Why?

[Arky Vaughan]

What does it tell you that. .400 OBP and a .600 SLG don't?

It tells you less.

[Arky Vaughan]

I just don't see why you would use one number for anything. It's a pointless exercise, particularly for evaluating baseball talent.

It doesn't evaluate talent.

[Arky Vaughan]

I think Limey hit on a very unique truth... the value is found in the usage and not in the information it provides. If I were wanting to put together a lineup, I'm going with OBP and SLG to see if this is a good leadoff man or a cleanup hitter. If I am putting together a list of offensive prowess for whatever reason (fantasy baseball, salary negotiation, etc.) and it's purely a ranking system, hey... gimme OPS (I guess).

I concur with Noe.

[Arky Vaughan]

This. Was. Created. For. Joe. Fan. Who. Did. Not. Know. Any. Better.

Also for dimwit sportswriters. Not for scouts, serious talent evaluators, etc. This was created because any yokel you talked to would value RBI and AVG disproportionally.

I don't think Joe Fan Who Did Not Know Any Better was the target audience for a $50, 2,000-page book full of baseball statistics with lots of calculations involved. Especially not in 1984.

[Arky Vaughan]

Same here. I've known several scouts and cross-checkers over the years. Never heard a single one ever mention OPS.

Because it doesn't have much if anything to do with what teams employ scouts for.

[Arky Vaughan]

I assume that preparation for fantasy drafts would include ranking the offensive value of players at each position.  Best way to do that is by using a single value, and OPS is the best single value stat for ranking offensive prowess.

I wouldn't say best, but I would say easiest or simplest, or even laziest.

[NeilT]

What does it tell you that. .400 OBP and a .600 SLG don't?

Course what does OBP tell you that the number of hits, ABs, BB, HBPs & SFs doesn't? What does SLG tell you that the number of singles, doubles, triples, HRs & ABs doesn't?  And frankly, knowing those individual numbers is probably more informative.  The greatest value of a derived number is comparison, and no derived number (and no number) gives a complete picture of the player on the field.  No offensive number indicates whether a player can play second base. It doesn't mean that offensive numbers aren't useful.

My sense of OPS was that it originated in the shift from Avg as a dominant offensive metric to OBP, so this argument is dead and done with.  If I want to get a sense of how a player is playing, what I want to see is a whole bunch of individual numbers.  If I want to get a sense of an individual player's value relative to other players, composite numbers are useful.  Is OPS useful?  To me it is, but it's just a way to start thinking about a player.  Certainly nobody's ever going to make decisions about player acquisition based solely on OPS, or even on OPS at all.  Doesn't mean somebody out there isn't finding some OPS based metric useful.  Shoot, I bet even OPS+ is seen these days as quaint.

[NeilT]

Wikipedia had the following list of players with more than 3000 ABs ranked by OPS+.

  1. Babe Ruth, 206
  2. Ted Williams, 190
  3. Barry Bonds, 181
  4. Lou Gehrig, 178
  5. Rogers Hornsby, 175
  6. Mickey Mantle, 172
  7. Dan Brouthers, 170
  7. Albert Pujols, 170
  9. Joe Jackson, 169
10. Ty Cobb, 168
11. Jimmie Foxx, 163

I had no clue who Dan Brouthers was, but I'm fascinated by the problem of trying to compare players over time.  The lists have no real value, of course, but does anyone not find a list like that and its rationale pretty compelling? 

[Waldo]

While we're replaying decade-old arguments, anybody up for a lively debate on Zone Rating?

[Ron Brand]

I don't think Joe Fan Who Did Not Know Any Better was the target audience for a $50, 2,000-page book full of baseball statistics with lots of calculations involved. Especially not in 1984.

Shit man, what kind of copy did you get? Mine's about 250-300 pages, probably ran about $12.95 to buy. OPS was just a part of that book, not the whole enchilada.

ETA: Amazon lists the book at 419 pages.

[HudsonHawk]

It tells you less.

So in what application is it preferable to have less information?

[Bench]

I had no clue who Dan Brouthers was... 

You should have seen the headlines when he lead the league in position player WAR in 1891 and 92.

[Bench]

So in what application is it preferable to have less information?

Some people prefer pretty numbers to actual people. 

[NeilT]

So in what application is it preferable to have less information?

When your wife is shopping for shoes.

[subnuclear]

So in what application is it preferable to have less information?

All statistics simplify information. You choose the numbers that best represent what you want to say.

e.g. "Carlos started slow in the first half with a 0.740 OPS, but improved to 0.850 in the second half."

[Arky Vaughan]

So in what application is it preferable to have less information?

It might be useful to know if I am comparing two players and want to know who appears to have created more value overall, regardless of whether that value comes from getting on base or moving runners along.

[HudsonHawk]

It might be useful to know if I am comparing two players and want to know who appears to have created more value overall, regardless of whether that value comes from getting on base or moving runners along.

Why can't you simply look at their respective OBP and SLG?  What does adding them together do that they don't do on their own?

Secondly, why would you not want to know how that value was created?

[NeilT]

Why can't you simply look at their respective OBP and SLG?  What does adding them together do that they don't do on their own?

Secondly, why would you not want to know how that value was created?

Two examples.  you tell me player x is having a great season.  I look at his ops and it's 1000.  I have a pretty good idea that he's having a good season, and that his slg is somewhere close to 700.  How is it helpful to have more of a breakdown?

you tell me player y is having a lousy season.  I look at his ops and it's 400.  I have a pretty good idea he's having a lousy season.

Why in either case do I really want more information, and how is it helpful to know more?

[HudsonHawk]

Wikipedia had the following list of players with more than 3000 ABs ranked by OPS+.

  1. Babe Ruth, 206
  2. Ted Williams, 190
  3. Barry Bonds, 181
  4. Lou Gehrig, 178
  5. Rogers Hornsby, 175
  6. Mickey Mantle, 172
  7. Dan Brouthers, 170
  7. Albert Pujols, 170
  9. Joe Jackson, 169
10. Ty Cobb, 168
11. Jimmie Foxx, 163

I had no clue who Dan Brouthers was, but I'm fascinated by the problem of trying to compare players over time.  The lists have no real value, of course, but does anyone not find a list like that and its rationale pretty compelling? 

Yes, more is better. I imagine if you add HRs and RBIs your list would look substantially similar. That doesn't mean adding HRs and RBIs provides any particular insight.

[HudsonHawk]

Two examples.  you tell me player x is having a great season.  I look at his ops and it's 1000.  I have a pretty good idea that he's having a good season, and that his slg is somewhere close to 700.  How is it helpful to have more of a breakdown?

you tell me player y is having a lousy season.  I look at his ops and it's 400.  I have a pretty good idea he's having a lousy season.

Why in either case do I really want more information, and how is it helpful to know more?

Why would I need the extra effort of calculating OPS?  If I told you Player X is having a great year with a .400 OBP and .600 SLG, why wouldn't you be able to get it?

[JimR]

Two examples.  you tell me player x is having a great season.  I look at his ops and it's 1000.  I have a pretty good idea that he's having a good season, and that his slg is somewhere close to 700.  How is it helpful to have more of a breakdown?

you tell me player y is having a lousy season.  I look at his ops and it's 400.  I have a pretty good idea he's having a lousy season.

Why in either case do I really want more information, and how is it helpful to know more?

does he get a ton of walks or hit a lot of doubles, triples and homers?

[NeilT]

HRs gives you this:

1. Barry Bonds (22) 762 L HR Log
2. Hank Aaron+ (23) 755 R HR Log
3. Babe Ruth+ (22) 714 L HR Log
4. Willie Mays+ (22) 660 R HR Log
5. Ken Griffey (22) 630 L HR Log
6. Alex Rodriguez (18, 35) 629 R HR Log
7. Sammy Sosa (18) 609 R HR Log
8. Jim Thome (21, 40) 604 L HR Log
9. Frank Robinson+ (21) 586 R HR Log
10. Mark McGwire (16) 583 R HR Log


RBIs gives you this:

1. Hank Aaron+ (23) 2297 R
2. Babe Ruth+ (22) 2213 L
3. Cap Anson+ (27) 2075 R
4. Barry Bonds (22) 1996 L
5. Lou Gehrig+ (17) 1995 L
6. Stan Musial+ (22) 1951 L
7. Ty Cobb+ (24) 1938 L
8. Jimmie Foxx+ (20) 1922 R
9. Eddie Murray+ (21) 1917 B
10. Willie Mays+ (22) 1903


without going through and adding people up, you can't tell how the final list would add up, but I'd guess you'd lose Anson, Jackson, Brouthers & Cobb.  You'd certainly gain Mays and Aaron.

[NeilT]

does he get a ton of walks or hit a lot of doubles, triples and homers?

But that's exactly right, OBP doesn't tell you that, and neither does slg.  If someone gives me OPS, I know something.  If someone gives me OPB and Slg I know something.  In neither case do I know this.

[Bench]

But that's exactly right, OBP doesn't tell you that, and neither does slg.  If someone gives me OPS, I know something.  If someone gives me OPB and Slg I know something.  In neither case do I know this.

If the player got a lot of walks his OBP would be higher, but if he hit a lot of extra base hits, his SLG would be higher. 

[HudsonHawk]

But that's exactly right, OBP doesn't tell you that, and neither does slg.  If someone gives me OPS, I know something.  If someone gives me OPB and Slg I know something.  In neither case do I know this.

You can't have OPS without first knowing OBP and SLG. And the latter two *do* tell something about whether he walks a lot, hits a lot of singles, Hrs etc. it's not complete, but it's some information. Way more than you get from OPS.

[NeilT]

If the player got a lot of walks his OBP would be higher, but if he hit a lot of extra base hits, his SLG would be higher. 

But you know that because you have a pretty good idea of what makes up a SLG percentage, and what makes up an OBP.   It's less precise, but you also have a pretty good idea of what makes up an OPS.  If you tell me someone's got an OPS of 1000, I know that he's probably hitting with some power.  Maybe down towards the middle that's less true, but the higher the number, the more you can guess, and the lower the number, the less you care.

[Noe]

It might be useful to know if I am comparing two players and want to know who appears to have created more value overall, regardless of whether that value comes from getting on base or moving runners along.

Agreed. It has a use if, say, I was negotiating a salary as either a GM or an agent. If one side says "He strikes out too much (see: Berkman, Lance), OPS can help win an argument on value of the player... offensively. I don't think it's meant for scouting or a manager to fill our a lineup card per se. That is stretching the value of the stat to where it should never be stretched.

[JimR]

But you know that because you have a pretty good idea of what makes up a SLG percentage, and what makes up an OBP.   It's less precise, but you also have a pretty good idea of what makes up an OPS.  If you tell me someone's got an OPS of 1000, I know that he's probably hitting with some power.  Maybe down towards the middle that's less true, but the higher the number, the more you can guess, and the lower the number, the less you care.

how do you know that?

[Noe]

Course what does OBP tell you that the number of hits, ABs, BB, HBPs & SFs doesn't?

Not the same. OBP is a cumulative stat while OPS is a combination stat. One shows a player's ability to get on base while the other allows you to *rank* players in terms of pure offensive value (however they provide said value, by on-base or slugging or and equal amount of both is not the intent).

[NeilT]

how do you know that?

Because of you see an ops of 1000, you can be pretty certain that the slg is going to be higher than 500. You can have a strong hunch that it's going to be around 600.  In neither case is that certain, but it's not useless to me as a fan.  In the same way, if I see an OBP of 400, I can guess that the player knows how to take a walk.  What would an abnormally high OBP be?  500?  If a player had a 250 OBP and a 1000 OPS, I wouldn't have to look at either number because he'd be leading the league in HRs. 

[NeilT]

Not the same. OBP is a cumulative stat while OPS is a combination stat. One shows a player's ability to get on base while the other allows you to *rank* players in terms of pure offensive value (however they provide said value, by on-base or slugging or and equal amount of both is not the intent).

Noe, my comment was that neither gives perfect information.  If I really want to see what a player is doing, knowing his OBP may indicate how he's hitting, but without knowing his HBP or BB or hits it's imperfect information, and without knowing his ABs it's useless.

[Noe]

Noe, my comment was that neither gives perfect information.  If I really want to see what a player is doing, knowing his OBP may indicate how he's hitting, but without knowing his HBP or BB or hits it's imperfect information, and without knowing his ABs it's useless.

Actually, it's not useless. OBP tells me the guy can get on-base and if I am filling out a lineup card, guess what? I have very valuable information who I can chose from my 25 guys to bat leadoff. OPS is a ranking system and that is not the same type of use for me. Your argument seems to be very heavily slanted towards fans who need quick consumption stats to place value of a hitter. That's all good and well, but if you stay with OPS, you'll never get to the detail you need to understand why one guy bats leadoff and the other doesn't. Isn't that what you'd want as a fan and not some ranking system stat?

[NeilT]

It occurs to me that this discussion really misses the point of OPS, and frankly it's a point that nobody is arguing.  If you want to look at the value of a player, the quickest way to get there is to look at the player's OBP and Slg.  Add 'em up, disect them into their components, it doesn't matter.  No one is really arguing what the point of OPS was: that Avg and RBI really don't tell you much.  Do I look at Avg and RBI?  All the time.  I love those numbers.  Do I look at Slg and OBP?  It's the first thing I'd look at after I looked at OPS, and then I'd probably look at HR and BB. 

I wouldn't look at range factor though.

[austro]

I wouldn't look at range factor though.

See, you're an AL fan already.

[NeilT]

Actually, it's not useless. OBP tells me the guy can get on-base and if I am filling out a lineup card, guess what? I have very valuable information who I can chose from my 25 guys to bat leadoff. OPS is a ranking system and that is not the same type of use for me. Your argument seems to be very heavily slanted towards fans who need quick consumption stats to place value of a hitter. That's all good and well, but if you stay with OPS, you'll never get to the detail you need to understand why one guy bats leadoff and the other doesn't. Isn't that what you'd want as a fan and not some ranking system stat?

Just to be clear, what I said was that OBP was useless without knowing ABs.  You'd get an OBP of 500 with one SO and a HBP.  That may not be the guy you want leading off.

[Waldo]

You can't have OPS without first knowing OBP and SLG. And the latter two *do* tell something about whether he walks a lot, hits a lot of singles, Hrs etc. it's not complete, but it's some information. Way more than you get from OPS.

I see OPS as kind of like Body Mass Index.  Both are conglomerate stats (so to speak) and neither is perfect, but both are generally good indicators in their own areas.

If I wanted to make a spreadsheet with a list of men and women and sort them from "overweightedness" to "underweightedness", doing it by both height and weight would be nearly impossible.  But BMI takes in those two numbers and spits out another, single number.  It's not accurate in some cases, but it would do a pretty good job of accomplishing what I want.

If I wanted to make a spreadsheet with a list of baseball players and sort them based on their overall offensive contribution, doing it by both OBP and SLG would be difficult.  In comes OPS, and while it wouldn't tell the whole story, it would give you a general idea.

[HudsonHawk]

If I wanted to make a spreadsheet with a list of baseball players and sort them based on their overall offensive contribution, doing it by both OBP and SLG would be difficult.  In comes OPS, and while it wouldn't tell the whole story, it would give you a general idea.

This is it in a nutshell.  OPS is good for arranging players on a list according to OPS.  But that's it.  When it comes to evaluating baseball skill, it doesn't really bring anything to the table and is pretty much pointless.

[Noe]

I see OPS as kind of like Body Mass Index.  Both are conglomerate stats (so to speak) and neither is perfect, but both are generally good indicators in their own areas.

If I wanted to make a spreadsheet with a list of men and women and sort them from "overweightedness" to "underweightedness", doing it by both height and weight would be nearly impossible.  But BMI takes in those two numbers and spits out another, single number.  It's not accurate in some cases, but it would do a pretty good job of accomplishing what I want.

If I wanted to make a spreadsheet with a list of baseball players and sort them based on their overall offensive contribution, doing it by both OBP and SLG would be difficult.  In comes OPS, and while it wouldn't tell the whole story, it would give you a general idea.

Perfect! Thanks.

[Noe]

This is it in a nutshell.  OPS is good for arranging players on a list according to OPS.  But that's it.  When it comes to evaluating baseball skill, it doesn't really bring anything to the table and is pretty much pointless.

If I were running a club and I were presented a list like this, first thing I'd want to do is find out why the guys that are my top ten performers are ranked so high. In other words, it leads me in the right direction to ask the right questions. But to evaluate, as you say, I wouldn't make any judgment call based on OPS.

[Noe]

It occurs to me that this discussion really misses the point of OPS, and frankly it's a point that nobody is arguing. 

I believe we do understand the point of OPS. We're really just asking or discussing the right framework for it to be effectively used.

[Noe]

Just to be clear, what I said was that OBP was useless without knowing ABs.  You'd get an OBP of 500 with one SO and a HBP.  That may not be the guy you want leading off.

Nobody makes judgments on low sample sizes... do you really believe they do? Heck, unless you're a radio sports talk show host who wants to say a player "sucks" after 10 ABs, very rarely do baseball people pull such a trigger. I remember how Michael Bourn was said to be a huge bust after about 50 ABs in April by some guy on KMBE (Davies?). Called him a "AAAA" player. Michael Bourn, for anyone with eyes, had much more talent than being cast as a "AAAA" player, so the baseball men knew to wait for a fair amount of data (observation and stat based) to make a judgment.

[Ron Brand]

Somewhat interesting. I thumbed through Thorn (ha) and OPS is mentioned as a shadow stat of the real point of the book, his Linear Weights system, devised to show how many runs were created by a player in a normalized manner across decades. I pulled out all my Abstracts from 1984 forward and James doesn't mention OPS in the glossaries / formulas parts of the books - he refers to Palmer's stuff in aggregate, as 'Palmer Method.'

I'm surprised that a causal link to what you guys believe to be a universal reliance on OPS wasn't uncovered in either of those references, but I haven't hit them hard, or the subsequent James publications. And since I'm off to work now, that may wait for another time...

[Arky Vaughan]

You can't have OPS without first knowing OBP and SLG. And the latter two *do* tell something about whether he walks a lot, hits a lot of singles, Hrs etc. it's not complete, but it's some information. Way more than you get from OPS.

OPS provides a means to compare the relative performance of different kinds of hitters, such as a Dawson-like hitter with a .300 OBP and .500 slugging average and a Rose-type hitter with a .400 OBP and .400 slugging average.

[Arky Vaughan]

Somewhat interesting. I thumbed through Thorn (ha) and OPS is mentioned as a shadow stat of the real point of the book, his Linear Weights system, devised to show how many runs were created by a player in a normalized manner across decades. I pulled out all my Abstracts from 1984 forward and James doesn't mention OPS in the glossaries / formulas parts of the books - he refers to Palmer's stuff in aggregate, as 'Palmer Method.'

I'm surprised that a causal link to what you guys believe to be a universal reliance on OPS wasn't uncovered in either of those references, but I haven't hit them hard, or the subsequent James publications. And since I'm off to work now, that may wait for another time...

You're barking up the wrong tree with James. He wasn't a fan of OPS. He figured runs created was simple enough to use before he created 20-something versions of it. Linear weights is fairly simple too: .31 for walks and HBP, .46 for a single, .78 for a double, 1.09 for a triple, 1.40 for a home run, minus something like 0.10 for outs depending on the league average.

[Arky Vaughan]

This is it in a nutshell.  OPS is good for arranging players on a list according to OPS.  But that's it.  When it comes to evaluating baseball skill, it doesn't really bring anything to the table and is pretty much pointless.

But it doesn't measure skill. It measures performance. Faulting it for not measuring skill is, how do you put it, a straw man?

[Arky Vaughan]

I believe we do understand the point of OPS. We're really just asking or discussing the right framework for it to be effectively used.

I concur with Noe again. In a thread about stats. Go figure.

[Ron Brand]

You're barking up the wrong tree with James. He wasn't a fan of OPS.

I know, but I thought he'd at least mention it somewhere, if not criticize it, and it would help me nail down a time when it vaulted into general discussion.  No soap.

[Waldo]

This is it in a nutshell.  OPS is good for arranging players on a list according to OPS.  But that's it.  When it comes to evaluating baseball skill, it doesn't really bring anything to the table and is pretty much pointless.

Would you evaluate "skill" solely based on OBP and SLG?  Of course not.

[HudsonHawk]

OPS provides a means to compare the relative performance of different kinds of hitters, such as a Dawson-like hitter with a .300 OBP and .500 slugging average and a Rose-type hitter with a .400 OBP and .400 slugging average.

No it doesn't. It completely obscurs the relative performance and instead lumps them together as if they were similar hitters. That's the very problem with OPS, not its value. And again, you can't know the OPS of those playes without first knowing their respective OBP and SLG. You know their poduction and the differences in how they arrived at it *before* you ever calculate OPS. OPS is is pointless and rendundant when comparing those two players.

[HudsonHawk]

But it doesn't measure skill. It measures performance. Faulting it for not measuring skill is, how do you put it, a straw man?

You've already measured performance before you calculate OPS. it's pointless. It brings nothing to an evaluation, either of skill or performance.

[HudsonHawk]

Would you evaluate "skill" solely based on OBP and SLG?  Of course not.

What's your point?

[NeilT]

You've already measured performance before you calculate OPS. it's pointless. It brings nothing to an evaluation, either of skill or performance.

It brings this to the evaluation:  If you want to use numbers to evaluate players, start with Slg and OBP.  You aren't suggesting a different set of numbers to start from, are you?

[HudsonHawk]

It brings this to the evaluation:  If you want to use numbers to evaluate players, start with Slg and OBP.  You aren't suggesting a different set of numbers to start from, are you?

I'm not suggesting you use any particular number and certainly not a single number. I'm saying that if you want to look at the value of a player's contribution using his ability to get on base and hit for power, then use OBP and SLG. OPS is less informative and counterproductive.

[Ron Brand]

I'm not suggesting you use any particular number and certainly not a single number. I'm saying that if you want to look at the value of a player's contribution using his ability to get on base and hit for power, then use OBP and SLG. OPS is less informative and counterproductive.

OPS was not created to give you that information. It was created as a sum of the two, a crude tool irrespective of the makeup of its parts, deliberate in its aggregate. Finer distinction was to be determined by other tools. Other peoples' bastarded usage of it is the problem.

[Arky Vaughan]

You've already measured performance before you calculate OPS. it's pointless. It brings nothing to an evaluation, either of skill or performance.

OPS has a higher correlation with run production than either OBP or slugging average separately. If I want to know which player had the more valuable performance, I can get a pretty good idea of that by comparing their OPS, regardless of whether the excel in getting on base, hitting for power, neither or both. If that does nothing for you, fine, but do not presume to judge whether it has meaning to somebody else.

[Arky Vaughan]

OPS was not created to give you that information. It was created as a sum of the two, a crude tool irrespective of the makeup of its parts, deliberate in its aggregate. Finer distinction was to be determined by other tools. Other peoples' bastarded usage of it is the problem.

Precisely. Sometimes you want specificity, sometimes you want an aggregate. OPS provides a relatively accurate measure of both getting on base and hitting for power.

[HudsonHawk]

OPS has a higher correlation with run production than either OBP or slugging average separately.

You don't look at them separately.  That's the point.

[HudsonHawk]

OPS was not created to give you that information. It was created as a sum of the two, a crude tool irrespective of the makeup of its parts, deliberate in its aggregate. Finer distinction was to be determined by other tools. Other peoples' bastarded usage of it is the problem.

I know what it is.  It has no practical use or value other than for arranging players on a list according to OPS.

[Arky Vaughan]

You don't look at them separately.  That's the point.

How would you rank players in a single list using two statistics?

[Arky Vaughan]

I know what it is.  It has no practical use or value other than for arranging players on a list according to OPS.

So would this be of no more practical value than arranging them alphabetically, or randomly?

By this logic, slugging average has no practical value other than for arranging players on a list according to slugging average. You really know nothing unless you break them down into singles, doubles, triples and home runs.

[Ron Brand]

I know what it is.  It has no practical use or value other than for arranging players on a list according to OPS.

Its practical use is to compare a player's contribution to creating runs. Again, it is crude, but it does have a good correlation to that use.

[Ebby Calvin]

Rinse, repeat.

[Ron Brand]

Rinse, repeat.

Head, meet wall.

[HudsonHawk]

How would you rank players in a single list using two statistics?

Why would you want to rank them on a single list?  But yes, if you want to list players according to their OPS, then OPS is the best tool for the job.

[HudsonHawk]

So would this be of no more practical value than arranging them alphabetically, or randomly?

By this logic, slugging average has no practical value other than for arranging players on a list according to slugging average. You really know nothing unless you break them down into singles, doubles, triples and home runs.

No single stat has much value other than listing players according to that stat.  That's my point.

[sporadic]

OPS has a higher correlation with run production than either OBP or slugging average separately.

if this is indeed true

It has no practical use or value other than for arranging players on a list according to OPS.

then this cannot be

[Ron Brand]

Fuck it, I'm out. My head hurts.

[subnuclear]

No single stat has much value other than listing players according to that stat.  That's my point.

Do you work for a scatter plot software company now?