T.R. Sullivan, statheads, and the RBI
by Adam J. Morris on May 5, 2008 10:00 AM CDT
Email
PrintT.R. Sullivan's blog entry from yesterday leads off with a lengthy paean to the RBI. And it got me thinking.
OBP seems to be the poster-stat for the ongoing stathead/traditionalist..."feud" is probably not the right word. Battle of paradigms? Or paradigae? Whatever the plural of paradigm is?
But even the anti-stathead types will acknowledge that OBP has some importance. OBP matters...there is just a differentiation of how important it is, how much weight should be put on it, compared to other stats.
But RBIs? The more extreme statheads call it a junk stat, irrelevant. The more extreme traditionalists use it to justify MVP and All Star votes. The RBI was at the heart of the very passionate disputes last season over the value of Sammy Sosa, which raged until late in the year when the Rangers just quit playing him. The RBI is a big part of the Jim Rice Hall of Fame debate, and the reason guys like Steve Garvey and Joe Carter have reputations that exceed their performances (or at least, their stathead-perceived performances).
Sullivan was one of the biggest Sosa fans last year -- even naming him the club's first half MVP -- and leaned heavily on RBI totals in his support. So it isn't surprising, I don't think, that Sullivan is presenting an ardent defense of the RBI (in response to Bill James being critical of the RBI on 60 Minutes).
The single most important offensive statistic in baseball is hitting with runners in scoring position. It's not any more complicated than that.
The Rangers hit .173 with runners in scoring position during their seven-game losing streak. They have hit .352 in their last eight games and have won six of those.
No question, hitting with RISP is crucial. The Rangers have had a pretty fair spread this season between the number of runs their stats suggest they should have scored, and what they've actually scored, and that's due almost entirely to poor performance with RISP.
(Although I would like to point out that, contrary to what many have suggested, this isn't a lengthy trend. There is a meme out there that the Rangers have tended to put up empty offensive numbers and then choke in the clutch, but the team has actually done better with RISP over the past few years than would be expected).
Is it the single most important offense statistic in baseball? I doubt it...but then, the "most important stat" argument is a little too esoteric for my tastes.
However...
The runners are going to get on base. The opposing pitcher will see to that. Some offensive players will get on base more than others and the value of that can't be denied. But it's all about getting them home and any Red Sox numbers cruncher is going to look like a genius when David Ortiz and Manny Ramirez are hitting 3-4 in a lineup.
Is this really logical reasoning? Couldn't you just as well say, the batters are going to hit with RISP, the law of averages will see to that, but it is all about getting runners on base so that batters will have guys to drive in?
I'd say that baseball history supports my statement moreso than the first statement. Some pitchers allow a lot more baserunners than others do. Some hitters get on base a lot more than others do. But how many pitchers have a dramatic differential between allowing hits with RISP and allowing hits normally? How many hitters have a dramatic differential between performance with RISP and overall?
Yes, David Ortiz and Manny Ramirez are great at driving runners home...but isn't that because they are great hitters? What reason is there to believe that they are making Bill James look good because of some innate run-driving-home ability that is separate from their offensive ability?
Bill James would say that Ramirez and Ortiz drive in lots of runs because they are great hitters who get lots of opportunities to hit with runners on base. And that, really, is all that RBIs are...a function of offensive ability, opportunity, and some luck (because whether Big Papi hits 25 points better or 25 worse than overall with RISP is, at the end of the day, a matter of luck).
Opportunity and luck are outside of the control of a player. And offensive ability can be measured more accurately by looking at how a player hits overall, in all situations. So if you are measuring a player, it doesn't seem to make a lot of sense to put any weight on RBIs, given that all it does is take something that can be more accurately determined with other methods, and then combines it with factors beyond a player's control to turn it into an irrelevant mash.
Look at David Ortiz. He has a career .287/.382/.554 line overall. With RISP, he has a career .303/.408/.523 line. Is he a "run producer", or just a great hitter? Ditto Manny...his overall line and RISP lines are similar. He drives in runs because he's a great hitter.
But nobody was better at getting them home than Juan Gonzalez when he was at his best. Will Clark and Rusty Greer were pretty good at it too and Michael Young has proven time and time again his ability to hit with runners in scoring position.
Juan Gonzalez, in his career, overall: .295/.343/.561
Juan Gonzalez, in his career, w/RISP: .292/.358/.543
Or look at Juan's 1998 season, the year he had 30-something RBIs in April and people were talking about him breaking Hack Wilson's RBI record:
Overall: .318/.366/.630
W/RISP: .315/.381/.612
Juan Gone had 157 RBIs that season, despite hitting worse with RISP than he did overall. Why did he have 157 RBIs that year? Because he had a whopping 372 plate appearances with runners on base. (The only player with as many PAs with men on in 2007 was ARod. He had 156 RBIs last year.)
Juan hit cleanup every game except 1 in 1998. Ranger leadoff hitters had a .369 OBP that season. Ranger #2 hitters had a .376 OBP. Ranger #3 hitters had a .373 OBP.
Juan didn't have a ton of RBIs in 1998 because he had a special knack for getting runners home. He had a ton of RBIs in 1998 because he had a great offensive year, and had a whole bunch of runners on base in front of him.
Pitching and defense are important, no doubt about it. But without spending a lot of time in a San Francisco hotel crunching numbers, experience says most close games are determined by which team is better at getting runners home.
Well, yeah. But that begs the question...do some teams have a particular knack for getting runners home? Or are the teams that are better at getting runners home simply the teams that have the better hitters?
Every year, you hear people say about bad offensive teams, "They need to win with timely hitting." That seems to be another way of saying, "They need to win by getting hits disproportionately often when they have runners on base."
And sure, in the abstract, that's fine. In reality, though, that works about as well as saying, "I'm going to provide for my retirement by having the winning lotto numbers disproportionately often." It simply isn't an executable strategy.
The RBI still lives as an important number and if Nate Gold is going to drive in 103 runs in two straight Minor League seasons then somebody ought to be paying attention.
Nate Gold hit .290/.346./.516 last season for Oklahoma. Pretty pedestrian numbers for a 27 year old AAA first baseman, and not the sort of line that is going to get you on the prospect radar screen.
The implication in the above quote, though, is that Gold deserves a shot because he's a Gonzalezian "run producer," based on two years in a row of 103 RBIs.
Gold's performance with RISP in 2007: .284/.335/.512. Basically, the same as his overall line. But he had 297 PAs with men on base...2.43 per game played. Basically the same rate as Juan Gonzalez in 1998. Gold had a bunch of RBIs in AAA last year because he had a lot of guys get on base ahead of him.
In fairness to Gold (and Sullivan), Gold did much better with runners on than not in the Texas League in 2006. But I'd want to see more evidence that that to believe he warrants a major league job based on his run producing ability.
And the reality is, while baseball people may pay lip service to that sort of skill, they don't really believe in it. Not really. If they did, someone would have snagged Nate Gold in the Rule 5 draft, or would have given up a Michael Hernandez or Scott Shoemaker to get him from the Rangers. If they did, someone would have signed
Sammy Sosa this year, based on his great run producing numbers from 2007.
The reality is, the more you look at the data, the more you see that clutch hitting and run producing correlates with overall performance. If you can hit, then over time, you'll drive in runs if the opportunities are there. If you can't hit, you aren't going to drive in the runs, even if you are considered to be a run producer.
And at the end of the day, this debate will continue to be, I think, the real Rubicon for the stathead/traditionalist splits. The traditionalists will look at the Rangers performance vis-a-vis bringing home runners in 2007 and see it as a failing of character, of will, of coaching, of approach, of...of something tangible, definitive, correctable. And they will look at RBIs as a meaningful proxy for whether a player is a run producer, or whether he's just an empty numbers guy who can't get guys home.
And guys like me will see it as a small sample size aberration, shrug our shoulders, and wait for the inevitable reversion to the mean. And will see RBIs as a distraction, rather than anything meaningful.
60 comments
|
2 recs |
Do you like this story?
Comments
That was a Posnanski length post
I cringe every time Sullivan mentions Gold. And I forsee a time when Gac gets the same mentions if he slowly climbs the ladder over the next few seasons instead of coming quickly.
...and curse Sir Walter Raleigh, he was such a stupid git.
Yeah
I was working on it last night during the Stars game, and it just sort of kept growing.
by Adam J. Morris on May 5, 2008 1:58 PM CDT up reply actions
RBIs will be
the most important stat for handicapping the MVP race for the forseeable future. If a guy is leading the league in RBIs, and is on a first place team, he becomes the overwhelming favorite for MVP.
Other than that, I don’t have much use for it.
"Before I leave, I once again condemn the despicable buffoonery of D.J. Cahill." - Huck
are you just too busy or
does this post overwhelm you?
by Randy Richardson on May 5, 2008 2:23 PM CDT up reply actions
Toby Harrah
what? why are you looking at me like that?
Grieve: The Yanks have struggled so far. - Lewin: Yeah, cry me a bag of money.
becuase you're crazy? i kidd...
i’m ‘bout to refer both you and slc to baseball reference. Maybe that will edgeumagate yourselves a little bit…
just got ownaged
heh
that made me chuckle.
Picking a random blog comment and wielding it as a club to bash "blogs" is like picking a random romance novel off an airport bookstore shelf and saying, "This book sucks. Fuck you, Tolstoy -- your medium is worthless!" - Ken Tremendous
Nice
Gotta be Pudge, though. A-Rod’s the best to ever play for the Rangers, but I don’t think he contributed more in his three years than Pudge did.
Upon some quick thought,
The greatest player to play for the Rangers has to be ARod. His seasons were better than Juan’s, and Alex played a more demanding position. Other than being pissed that he played for a crappy team, there’s not much wrong what ARod did during his Ranger years.
The best player to play for the Rangers that came up through their farm system has to be either Pudge or JuanGone.
RBI and Runs scored...
I’m pretty sure that OPS encompases those two stats.
After all it’s hard to knock in runs and score runs without getting on base and hitting for some power.
"The only good is knowledge and the only evil is ignorance."-Socrates
yeah, I view runs the same as RBIs
you get RBIs because someone got on base and you hit the ball.
you get RS because you got on base and someone hit the ball.
Certainly not individual stats.
Grieve: The Yanks have struggled so far. - Lewin: Yeah, cry me a bag of money.
Botts and RBIs
you get RBIs because someone got on base and you hit the ball.
I’d like to take a look at Botts and his RBIs. Here is his line with RISP .333/.571/.667. This looks pretty good, but, as we all know, this is an incredibly small sample size. He had 9 ABs with RISP, but had 14 PAs. Botts drew five walks (none of which came with the bases loaded). If Botts got a hit instead of drawing a walk, he would have driven a few more runs (despite that amazing line, Botts only drove in 3 runs). I remember when Evan Grant harped on Botts during ST because he did not drive in any runs in a particular AB. In fact, he referred to that situation more than once during his ST coverage. In his very limited time this season, Botts drew a walk more often than getting a hit with RISP. Of course, this has value in its own way (though not nearly as much as a hit). I have to wonder, was Botts the casualty of the RBI way of thinking? Would one or two hits with RISP been enough to keep his job?
by Excel Hearts Choi on May 5, 2008 10:42 AM CDT up reply actions
I have ardently believed
that Botts was a victim of not being perceived as clutch, or not clutch enough, more likely. It seems silly, but when you look at Washington, who is more than willing to throw down a sac bunt, even in games where we are getting to the pitcher, it is apparent that Washington does not take the Bill James approach that sacrifice bunts are not the way to go, ever. While James’s approach may be extreme to most managers, Washington’s (over?) use of the sacrifice bunt should tell you that he is interested in one thing: producing runs. I am certain that in his mind a walk is just not anywhere close to a sacrifice or a hit.
In the end, RoBotts is gone and I am still bitter.
Oh...
...and I said that Washing is interested in producing runs in that yes, all managers want runs produced. What I meant was that Washington is interested in doing what he believes to be “manufacturing” runs, a la a sacrifice bunt. Not necessarily something I agree with or that the numbers can say actually produce runs, but I’ve typed enough.
walks worth "not nearly as much as a hit"
this is the problem. The state of the art in measuring performance is to look at the impact of each at bat on the expected number of runs the team will score. You can get that by looking at how many runs on average are scored after each “state” of an inning (say 1 out runners on 1st and third). Then the value of a plate appearance is the change in the expected number. At the start of an inning, the expected number of runs is about .5, (if the average runs per 9 is 4.5). If the first batter reaches the expected number goes up, otherwise it goes down. If some one scores but the state doesn’t change (a solo home run, double with man on second) the even is worth exactly one run etc.
If you average the value of each event by the frequency of each state, you get a pretty good average value. When you do this you find that a single is worth on average about half a run, and a walk is worth about 1/3 of a run. Extra base hits are worth even more, a home run on average is worth about 1.4 runs. So a hit is worth more than a walk.
But we don’t compare players that way. Look at your Botts stats. In 9 AB he had 3 hits and 6 TB and made 6 outs. The hits added an expected 3 runs for the team, the outs cost the team about a run and a half. Net gain about 1.5 runs, or a little less than .2 runs per AB. The walks were only worth about 1/3 of a run each, but that is the net value, we don’t have to worry about the outs created by putting the ball in play. So even though Botts was hitting at a fantastic level when he put the ball in play, those plate appearances were worth on average, only a little more than half what the walks were. Put another way, if Botts had managed to put the ball in play in every PA and still kept his .333 BA and .667 slugging instead of drawing those 5 walks, the expected number of runs scored would have gone down, by about 2/3ds of a run.
I think there is a perception that players who walk with RISP are also striking out more, and that is probably true, but we don’t separate out the PA’s where they walk or strike out from the ones where they put the ball in play. If Botts made 5 extra outs to go with his 5 walks, it would be a legitimate criticism, that he is too picky with runners on, but think about what that means- he hit .333 in his 9 AB and slugged .667 but if he gave up 5 of those outs trying to draw walks then he really only tried to put the ball in play 4 times to get 3 hits. I don’t think that the folks who decided to get rid of Botts think he is a .750 hitter who takes too many pitches, but I’m not sure what their reasoning is.
by DaveRichardson on May 5, 2008 12:29 PM CDT up reply actions
My original statement
I realize that providing alternate scenarios than what actually happened is quite foolish (nobody really know what would have happened had Botts put a ball in play). I am just struggling to find a reason as to why Botts was DFA’d. Since Grant touched on a similar situation in ST, I just thought the same criticism was applicable here.
by Excel Hearts Choi on May 5, 2008 12:47 PM CDT up reply actions
Why was Botts DFA'd?
Probably because they think he will get through waivers, and then the organization is spared making a decision on a guy like Broussard. I think its a silly plan, since Broussard is so replaceable, but then again, I do think he will slide right through waivers.
I don’t think it had much to do with RISP, I think it had a lot more to do with his total performance.
"Before I leave, I once again condemn the despicable buffoonery of D.J. Cahill." - Huck
Great post, AJM....
I’ve been watching baseball for a long time, but I’m only now learning more about the science and statistics of the great game.
It’s information like this that helps me learn more about the game and appreciate it more.
Thanks!!
For all that you missed the real problem with RBIs
The passion that James brings to the RBI issue is because of Joe Carter (and a few others like him). You point out that the number of RBIs is largely a function of how many times a player hits with men on base, but, for James the key word there is “hits”. When James was establishing his reputation, the most overlooked contribution to winning baseball games was the number of walks a player takes, and consequently the number of times he is on base and keeps the inning going for the next batter. Carter consistently was among the league leaders in RBI because he batted 4th in a good lineup, hit a lot of home runs and most importantly, never walked (a typical Carter season was .289/.312/.497). Carter’s high RBI counts were not because he performed better with runners in scoring position. He just got credit for a larger share of the runs he helped generate than anyone else.
If a player walks with men in scoring position he creates an even better scoring chance for the guy behind him. Moreover, it doesn’t really matter to the team if you drive yourself in with a home run, or score because someone else drives you in later. Carter drove himself in a lot, but rarely was on base for someone else to drive in. He also made huge numbers of outs with runners in scoring position (we now see that in the box as lob), so that he often left the team worse off after his at bat.
Bottom line, Carter was a fairly average player with some power who never walked. The result was that he was often among the league leaders in AB, and so his counting stats (HR, RBI) were inflated relative to similar players who walked at an average rate. The missing walks meant that Carter contributed little to his teams’ ability to score runs compared to an average replacement (he was not a very good player), but by themselves explained the high RBI total for which he got MVP consideration.
It is worth noting that the relative importance of walks depends on the relative quality of players. In high school, the best player on a team may be nearly ML quality, while the second best couldn’t walk on at a decent college program. If you are hitting .560 with a HR every 5 AB, and the guy hitting behind you is hitting .230 with no power, the benefits of taking a walk and letting the next guy have a chance are a lot smaller. I think a lot of the RBI bias comes from this, the fact that players learn baseball in a context were the talent distribution is a lot wider than it is in the majors.
Carter...
I’d say he’s a little better than “fairly average”.
I do agree with the rest of your argument though.
"The only good is knowledge and the only evil is ignorance."-Socrates
why?
In 1992 when he had his best year in the MVP vote (3rd), Carter was a below average RF (8 errors, below average RF). He hit .264/.309/.498, for an OPS of .807. That is a fairly average RF, most of his value comes from HRs (he had 34) but even his .498 slugging isn’t a huge strength. He got 622 AB because he drew only 36 walks in 683 PA. He made 491 outs, which is why he is only a fairly average player.
I guess the other thing going for Carter is he looked like a ball player, Big strong fast, great arm. He wasn’t average in terms of the skills he brought to the park, only in the actual impact he had on his team’s record.
by DaveRichardson on May 5, 2008 11:35 AM CDT up reply actions
I wish I had your baseball brain.....
Can I read a book or something to learn the stuff you know??
Great posts….thanks!!
Places to learn:
www.baseballprospectus.com
www.hardballtimes.com
www.tangotiger.com
And pick up Moneyball, Baseball Between the Numbers, and anything you can find from Bill James.
so rob neyer said
you use begs the question correctly and I thought you might have been in that last diary but are you sure it is correct here? there usually isn’t a question after the correct usage of begs the question.
Rare Gnats Sex
MY
How does one explain the phenomenon of Michael Young? His career OPS with RISP is almost 100 points higher than his career OPS in general. He has consistently performed better with runners in scoring position throughout his career, and I’m not sure that can be explained exclusively by “luck”. In his case, there must be something else at play.
But that river of tears has dried for all of us.
MY approach
I’m guessing it’s because MY has a different approach to hitting with 2 strikes or in a “clutch” situation. At least that’s what it seems to me anyway. Maybe he would be a better hitter if he had that same approach through the entire at bat for each at bat, instead of just those specific situations.
Derek
Signature! I don't need no stinking signature!!
by DerekSTheRed on May 5, 2008 11:47 AM CDT up reply actions
2 Strikes
For his career, MY has a line of .227/.283/.306 with two strikes.
by Excel Hearts Choi on May 5, 2008 11:56 AM CDT up reply actions
And that is better than most.
MLB as a whole is hitting .185/.260/.276 so far in 2008.
...and curse Sir Walter Raleigh, he was such a stupid git.
The other thing thats odd about MY
is that in close and late games, his OPS is like 120 points lower.
"Before I leave, I once again condemn the despicable buffoonery of D.J. Cahill." - Huck
Dammit Cahill ...
you’re killing the good vibes I was getting from this thread.
RBIs
are a function of good hitting. Something missing in this argument: A player’s RISP average relative to the score in the game. That where you get non-stat heads (like me) throwing around the word clutch and players like Jeter in the same sentance. Exhibit A: Manny’s grandslam against Texas. While RBIs are still relative in my mind, the timing of them in the game is far more important and, for some reason, not worthy of measure.
I've seen no evidence...
...that certain players perform better in closer games than in blowouts.
by Adam J. Morris on May 5, 2008 12:18 PM CDT up reply actions
This is something that is now getting attention.
We know that some players match up better with some pitchers. In particular guys with short swings and quick bats should handle the best pitchers better than guys who have “slider speed” bats. Since teams are more likely to run out the good stuff (top starters, closers, lefty/righty specialists) with the game on the line, it makes sense that some players would consistently handle those situations (relatively) better. The basic aggregate stats are pretty crude for getting at that level of detail, but now that games are recorded on a pitch by pitch level, it is a lot easier to look at. I know Bill James is actively reconsidering his position that “clutch hitting” doesn’t exist.
by DaveRichardson on May 5, 2008 12:45 PM CDT up reply actions
recent James quote on clutch hitting
Q: Is clutch hitting a repeatable/retain-able skill?
A: I don’t know.
by Randy Richardson on May 5, 2008 2:18 PM CDT up reply actions
So in a 3-2 game
with 1 on and 2 out , you have a choice between 2 right handed batters that have identical stats, you dont think you 1 will come through for you more often than the other?
I don't get it
every other profession has clutch. Why can’t baseball?
You think trial lawyers don’t occasionally pull a cross examination out of their ass when 99% of the time they or another lawyer would have wiffed on the witness?
I think your assumption
that every other profession has clutch is wrong. There are some attorneys who are just better than others, thus, more able to pull something out of their asses. Others will occasionally hit on all cylinders but that doesn’t make them clutch.
People that are good in the clutch are just people who are more talented, have worked harder, and/or are better prepared for that particular task, some combination of those three things.
...and curse Sir Walter Raleigh, he was such a stupid git.
I agree Tball about those 3 things
but not that some people have the ability to put it into over drive and do something that 99% of the time most cant do. I’ve seen it too many times in my life. But I hear ya, most of the time it can be explained. Clutch can’t or at least can’t be measured. That must mean clutch doesn’t exist?
Ugh...
Jeter + clutch in the same sentence = me rolling my eyes.
"The only good is knowledge and the only evil is ignorance."-Socrates
WTF????????
Wash Fired?
Norm just said that, and went to commercial. I can’t find any links…..
...it's the weekend, so why the hell not?
Probably
just a tease to get you to stay through the commercial break.
Brandon Boggs 2008 Texas Rangers ROY
F'ing Norm, What a tease!
Rat bastard….he had me!
Sorry, I jumped the gun.
...it's the weekend, so why the hell not?
norm
“Ron Washington no longer manages the Rangers, he manages Oklahoma City”
Man, i’d like to punch him in the face right now.
just got ownaged
Nice analysis Adam.
Very easy to follow. I think you’ve made an excellent point. I agree with your position. But I do think karma plays a role in baseball. For example, the way some players just murder certain teams (like Vlad and the Rangers, or Cust and the Rangers). There are certain guys that come to bat late in games where you just say to yourself, “here comes a game-changing hit.” That’s something I don’t think you can really quantify with stats, but I’m certain it exists.
Perception
It would make sense to me that some pitchers get just as scared as we fans do, if not more, of certain batters coming up simply because they are perceived as “clutch” hitters, or “ranger killers”. This could very well cause pitchers to pitch in a way they normally wouldn’t and give the hitter an advantage in the match-up. That is not in the stats and exists in the minds of players.
"Is this really logical reasoning?"
Nope.
“Couldn’t you just as well say, the batters are going to hit with RISP, the law of averages will see to that, but it is all about getting runners on base so that batters will have guys to drive in?”
Yep
Thank you for writing this, Adam. I’m glad to see someone that others listen to taking RBIs to task. I just do not understand how it isn’t apparent to people that there’s no reason to look at them, because 1) they’re a virtue of the lineup, not just the player, and 2) there are stats that tell you so much more, so why bother hampering yourself?
I also found this pretty funny:
Pitching and defense are important, no doubt about it. But without spending a lot of time in a San Francisco hotel crunching numbers, experience says most close games are determined by which team is better at getting runners home.
And when he’s talking about Gold, that’s what gets my blood to boil. We have enough problems building a winning team without someone crying that baseball research and analysis needs to be set back in time 15 years or more.
I would dump Ben Brossard in a NY minute
and bring up Gold, every day Brossard takes up space is another day less we learn about the Golds, the Botts et al. Every day Brossard is on this team, is a another bad management decision.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
