Men Left On Base
by jscape2000 on Mar 16, 2010 10:00 AM EDT
in New York Yankees Analysis
Email
PrintI've always said that I don't care about Men Left On Base.
A few times every season, John Sterling or Michael Kay will go off on a tangent deriding the Yanks for leaving too many men on base. And it's true that several times a season the Yankees will score 3 runs, leave 10 men on base and lose 3-4, when one hit could have changed the game.
But Poz took the time to run the numbers:
If a team outperforms their opponent by getting more men on base in any way (hits, walks, HBP, errors, K passed ball), that more effective team will post a .827 winning percentage.
A team that leaves fewer men on base than their opponent, that team will win only .418.
Pretty incredible, no?
Interestingly, Total Bases (the basis for SLG%) is an even better indicator of winning than OBP (.847). Total bases, of course, doesn't just add up homers. For TB, 2 doubles in 2 ABs are just as valuable as a homer in one AB. And since TB is a counting stat, the extra ABs inherent in long innings (which result from getting on base and not making outs), are a part of the stat.
The Yanks led the league in OBP (.362) and TB (2703), and the next best team (Boston) wasn't that close (.352 and 2516). On the other side of the ball, the Yanks' pitching staff was the 3rd best in the AL, meaning the Yanks' opponents had as hard a time reaching base as against any team in the game.
Those are not insurmountable leads (the Yanks posted a .342 OBP in 2008), but man it gets me excited for the season.
34 comments
|
0 recs |
Do you like this story?
Comments
The Mel Brooks clip in Posnanski's column was great- I think there was something about baseball in it as well.
Actually, Posnanski’s column reinforces (for me, anyway) a point that I made (and took crap for) yesterday about sabermetrics. My point was that sabermetrics is worthless mumbo-jumbo. Sabermetrics is worthless because it takes statistics that are inherently fraught with error and then combines them in a way that multiplies the effects of the errors. More than that, the statistics don’t tell you anything that you can’t see if you watch a lot of games with a whit of intelligence. It’s predictive value is derived ina similar manner to, and with equal relaibility to a long range weather forecast- the further out it goes, the less reliable it becomes.
POsananski’s point, which is that producing baserunners produces wins, and that many baserunners combined with extra base hits produces even more wins, is the antithesis of sabermetrical analysis, and yet it makes a much more profound statement.
So there.
And Mel Brooks is a very funny guy. Take 5 minutes and play the clip.
by designatedquitter on Mar 16, 2010 10:33 AM EDT reply actions
Hang on
Sabermetrics is worthless because it takes statistics that are inherently fraught with error and then combines them in a way that multiplies the effects of the errors
First of all, you can’t make generalizations like that without providing any evidence to back them up, and then expect people to not to criticize you. If you want to talk about the pros and cons of wOBA+, then by all means let’s. But generalizations are going to be met with criticism.
Second of all, I think there is a constant effort being made to improve the quality and our understanding of the statistics we use. To me, “sabermetrics” does not set out to find one stat to trump them all, and anybody who tells you otherwise is full of it. Instead, it tries to fill in the gaps, to identify holes or flaws in current statistical measures and try to fix them.
The goal, then, is not for the new statistic to be perfect, but only for it to better (or less inaccurate, if you will) than the one it’s intended to substitute.
the statistics don’t tell you anything that you can’t see if you watch a lot of games with a whit of intelligence.
I saw a dozen games in person last season, and watched another two or three dozen on TV, so you could say that I saw “a lot of games”. But that’s still only about 2% of all the games that were actually played last season, and the ones I saw were heavily weighted towards the Phillies and the Yankees. Assuming that one’s eyes can’t deceive them (which, obviously, they can), am I only able to speak intelligibly about the Phillies and the Yankees?
My point is, with 30 teams of 25 players each, you can only see enough “with your eyes” to make sound judgments about perhaps a handful of teams and their respective players. Hence, love them or hate them, we have statistics.
It’s predictive value is derived in a similar manner to, and with equal relaibility to, a long range weather forecast- the further out it goes, the less reliable it becomes
I think this statement really depends on what you’re trying to do. Obviously, it’s foolish to say that based on X, Y, and Z, Derek Jeter is going to post a .317/.402/.466 batting line in 2010. But regardless of your affinity for stats, every baseball fan is likely to make some sort of prediction about the future, so why not have some method to the madness? We’ve got 125 years of baseball statistics at our fingertips, so is it completely unreasonable to sift through that data to try and find trends so that we can make better educated guesses? Is it unreasonable to say that, based on all the information we have at hand, Robinson Cano is a pretty good bet to be as good this year as he was last year? Because I don’t see any difference between doing that and looking at the forecast to decide if you should bring an umbrella to work on Thursday.
Posananski’s point, which is that producing baserunners produces wins, and that many baserunners combined with extra base hits produces even more wins, is the antithesis of sabermetrical analysis, and yet it makes a much more profound statement.
Again, I disagree. If Sabermetrics is all about correcting flaws in the way we analyze baseball, then he spot on. Kay and Sterling deride the Yankees for leaving runners on without acknowledging that it is a byproduct of having so many hitters reach base in the first place (which is a good thing).
I stand by my statements.
By the way, even though LOB is evidently a good problem to have, I think that their point is slightly more sohisticated than leaving men on, it’s things like failing to drive in runners from third with less than two out, or advance runners from second with none out with grounders to the right side, etc.. That point is valid.
by designatedquitter on Mar 16, 2010 12:11 PM EDT up reply actions
If you want to make blanket generalizations, refuse to discuss them, and simply insist that you’re right, so be it.
I don't argue with people about their religion
by designatedquitter on Mar 16, 2010 12:40 PM EDT up reply actions
Right
If that were true, you wouldn’t have said what you said in the first place.
sabermetrics isn't my religion- you said I wouldn't argue facts, and I won't.
I just don’t have the faith. Can’t/won’t argue with those who do.
by designatedquitter on Mar 16, 2010 1:13 PM EDT up reply actions
You missed the point of sabermetrics
Sabermetrics is worthless because it takes statistics that are inherently fraught with error and then combines them in a way that multiplies the effects of the errors.
It isn’t that long ago that OPS was a mysterious “advanced stat” shunned by sportwriters. Good sabermetricians work to clarify the pros and cons of the stats they use.
OPS is a better stat than batting average (which was the standard for most of the last 100 years), but it overvalues SLG in relation to OBP.
I’m not sure how you reached the conclusion that baserunners=success is antithetical to sabermetrics. Good sabermetrics, like good science, doesn’t set out to prove a point; it wants to explore the available data to clarify what works and why.
"Have faith in the Yankees, my son. Think of the great DiMaggio."
I reached that conclusion by noting that the stats used are those familiar to non-sabermetricians.
Left on base, wins, and other stats that show up in the box score as opposed to derived ratings. That’s what I meant. OK? As I posted elsewhere, I don’t think that all sabermetrics is useless, only that the value of the derived stats is overrated/ overstated. When derived stats are multiplied or combined, the inherent error factor is also multiplied. I don’t think batting average is a more “telling” stat than OPS, I just think that combining OPS with other numbers to produce ratings is not a reliable way to go.
by designatedquitter on Mar 16, 2010 1:22 PM EDT up reply actions
Ok, that’s a fair critique.
I agree with what you said when applied to the practices of certain sabermetricians.
"Have faith in the Yankees, my son. Think of the great DiMaggio."
I think quitter does have a point, though I think it’s pointed slightly in the wrong direction.
The general problem is a lot of people use sabermetric stats, especially summary ones, but aren’t exactly knowledgeable about the error bands of the statistic. False precision has always been a problem with a certain segment of statheads. How a team scores runs may be pretty easy to model, but it’s not so accurate that you can declare a player with a VORP of 52.3 is 12.3 runs more valuable offensively.
However, combining stats with different error bands and variance is not a problem from a mathematical standpoint. If you’ve ever made a budget or planned anything ever, you’ve combined measures with inherently different errors (and it’s certainly not multiplicative). And basic stats are frequently just as “derived” as some sabermetric stats, but just happen to be more familiar to a wider audience. BA and ERA have components that have different error margins as well – for example, in the former, one aspect (whether something is a single or an error) has a subjective aspect while another (whether something is a homer), has a more objective definition. Something like RBI would never make headway today (and was roundly mocked for being ridiculous back when it was introduced).
--
Dan Szymborski
http://www.bbtf.org
http://insider.espn.go.com/blog/tmi-mlb/category?name=dan-szymborski-contributors
by D.Szymborski on Mar 16, 2010 8:33 PM EDT up reply actions
Men Left on Base
So is your point that John, Michael et al.’s compaint about leaving more men on base is irrelevant because getting more men on base in the first place is much more important than leaving fewer men on base at the end of an inning? That does make sense.
On the other hand, the sportscasters’ point is that the team is missing a lot of opportunities to convert baserunners into runs. (i.e., they are not being “clutch”). In the long run, putting more men on base than your opponent will result in more total runs, and therefore more wins. But over the course of one game—-an inherently small sample size—-you will still need to convert your baserunners into more runs than your opponent does.
Agreed.
I suppose a corollary to this post is the statement that I don’t believe in clutch (I do, however, believe in choke).
"Have faith in the Yankees, my son. Think of the great DiMaggio."
They scored 69% of the time with runner on 3rd< 2 out.
They hit for a higher batting average after the 7th. They hit .246 RISP late. I guess these numbers are mid-pack. The question is whether Sterling, Kay et als did a mid- pack amount of whining about their moving/ scoring runners.
by designatedquitter on Mar 16, 2010 12:49 PM EDT reply actions
RISP...
Runners left in scoring position is the more important stat than men LOB. You can get
a two out single (with nobody on) three times in a game for example, the next guy
makes an out all three times, 3 LOB, in that scenario, not an issue…Just a quick example off top of my head.
As I understand the post, the real question is whether the criticism of the Yankees leaving too many on base is justified.
The answer appears to be no, because lots of LOB correlates to winning. Their batting average with runners in scoring position, especially late in the game, looks low, although the team average in general late in the game is very high (over .300. Does this account for the comeback wins? Probably). But there are other ways to score runs than by base hits. Sterling in particular bemoans the lack of A-B-C baseball by the Yanks, that is ground ball to right side plus sac fly to score a runner from second.
On the other hand, if you score 4 runs but leave 2 on base in an inning, the 2 LOB is nothing to be concerned about. If you are losing by a run, and someone hits a home run, you have zero left on base. But Posnanski makes a strong case for LOB as being a sort of collateral damage resulting from a lot of offense in general.
by designatedquitter on Mar 16, 2010 3:27 PM EDT reply actions
Couple points...
1) I agree with Jscape in that there is no such thing as “clutch”. It’s like what Jeter said when the Yankees were down 0-2 in the ALDS one year, “There is no trying to win more” It’s not like guys are only half trying when there’s nobody on base, and then amp it up when there’s a runner on second. It’s not like the Yankees were only half trying to win the first two games of the ALDS and can now turn it up a notch.
In every at bat, it is reasonable to assume that you’re getting at least 90% effort from the pitcher and batter, and clutch is a byproduct of luck rather than skill. However, psychological factors definitely can lead to gripping the bat too tightly and choking. In clutch situations, I think that the ceiling is the players natural (non-clutch situation) ability. The love-fest that springs around “clutch” players like Derek Jeter and David Ortiz is a media and fan creation and isn’t tangible.
2. Small ball is so vastly overrated it isn’t even funny. The Yankees should never bunt, they should never try and hit a sacrifice fly, and they should rarely worry about stealing bases. When was the last time that the perennial contenders were small ball teams? 1920? Giving outs to the other team when you have the type of bats the Yankees have is just stupid.
Sterling, or any other fan, bemoaning the lack of ABC baseball, like not bunting when there’s a runner on 2nd and no outs, is clinging to an archaic idea of how baseball should be played. The guys are on the juice, the parks are smaller, and there’s a DH, making the NL small-ball style inferior. Get over it.
This sticker is dangerous and inconvenient, but I do love Fig Newtons.
Correction
2. Small ball is so vastly overrated it isn’t even funny. The Yankees should NEVER EVER EVER EVER EVER EVER EVER EVER EVER EVER EVER bunt, they should never try and hit a sacrifice fly, and they should rarely worry about stealing bases. When was the last time that the perennial contenders were small ball teams? 1920? Giving outs to the other team when you have the type of bats the Yankees have is just stupid.
Not ever doing those things mentioned above make your team quite exploitable and predictable, no?
by Scooby Snacks on Mar 16, 2010 6:40 PM EDT via mobile up reply actions
That's like saying that the Colts offense is "exploitable and predictable"
since on every drive they try and score a touchdown.
Giving the other team outs is bad and should not be done by any major league level batter.
This sticker is dangerous and inconvenient, but I do love Fig Newtons.
Obv. there’s methodology involved. Otherwise, Peyton Manning should be lobbing hail mary’s on every play.
by Scooby Snacks on Mar 16, 2010 11:42 PM EDT up reply actions
Bunting
I agree that most of the Yankees should never ever ever…ever bunt. But what about GGBG? He’s not exactly a power hitter, and as I recall, had reasonable success getting to base on bunts.
I think the never ever ever ever ever ever ever ever ever ever ever bunt admonition only applies to sacrifice bunts. It’s giving away outs that is the problem.
Usqueadbaugham! Anam muck an dhoul ! Did ye drink me doornail?
There are in fact occasions where the sacrifice bunt does make mathematical sense. For one, in very low scoring environments, the relative value of a run, base, and out change to favor the former two greatly over the latter. There’s also a break-even point for particularly skilled/fast bunters – at 15%, a bunt attempt becomes profitable in a generic situation.
--
Dan Szymborski
http://www.bbtf.org
http://insider.espn.go.com/blog/tmi-mlb/category?name=dan-szymborski-contributors
by D.Szymborski on Mar 16, 2010 8:38 PM EDT up reply actions
+1
In extra innings, with a tie score, a sac bunt from a low contact and/or low power hitter is acceptable.
A bunt for a hit attempt is acceptable, especially when it takes advantage of a misalignment in the defense, or if ‘failure’ results in a sac bunt advancing the runners.
"Have faith in the Yankees, my son. Think of the great DiMaggio."
Of course, there are exceptions to never ever*10. jscape2000 points out one good example in his response to you. I’d imagine there are others, but I also imagine there is a point in the game before which it never makes sense to sac bunt. It’s pretty clear that it wouldn’t make sense in the first inning for any realistic game; I’m guessing that’s probably true for the first several innings. Any idea when that point might be?
Also, you say that 15% is a break-even point for bunters, but 15% of what?
Usqueadbaugham! Anam muck an dhoul ! Did ye drink me doornail?
Sorry, 15% chance of reaching base on a bunt.
--
Dan Szymborski
http://www.bbtf.org
http://insider.espn.go.com/blog/tmi-mlb/category?name=dan-szymborski-contributors
by D.Szymborski on Mar 18, 2010 8:59 AM EDT up reply actions
So let me get this straight...
A team that gets more men on base than their opponent and or has more extra base hits than their opponent is more likely to win?
Good to know.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
Community Guidelines
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.
