PDA

View Full Version : Digging a little deeper into BABIP

12-31-2008, 12:49 AM
dougdirt's post on the Tavaras thread got me thinking. I'm not a big fan of the BABIP/luck theory, but he made a good point:

Whether it was earned or not, the likelihood of him repeating a .371 average on balls
in play is very unlikely. Between 2000 and 2007 there were 1747 people with at least 400 PA
in a season. The number of people with a .371 BABIP or higher over that period of time was
a grand total of 36. Let me repeat that, 36 of 1747 people were able to post a .371 BABIP or
higher over a 400 PA season. Thats barely 2%. Thats pure luck. Whether you are willing to admit
that or not is another thing, but it is what it is.

Then I remembered WestOfYou's post about Tavaras bunting for base hits. He does it a lot, and is quite successful at it.

In 2007, Tavaras was 37 for 52 when trying to bunt for a base hit.

So let's recalculate the BABIP number, when we back out the bunt hits and bunt attempts.
I think this is valid, because bunting for a base hit is different than the typical attempt to get a basehit.
Obviously, for Tavaras, when the time was right, it was much more successful than swinging away.
I hope the BABIP fans agree that there is a different set of conditions associated with bunting for a basehit vs
just swinging away and hoping it falls between fielders.

Since I am not sure how many of Tavaras' failed bunts were considered in play (vs a pop up out of bounds),
I will give a range of the adjusted BABIP.

I am going to show my math, just in case I make a mistake.

Wily had 119 hits in 2007. If his BABIP was .371, that means he put 321 balls into play.
Backing out his bunts, he had 82 non bunt hits, his adjusted "in play" is between 269 and 284

That makes his revised BABIP between 288 and 305. I am guessing that this BABIP is not considered lucky.

Note, this is not an attempt to say that Tavares is a good or bad acquision. Just another look at BABIP.

Edit: I hope doug's quote was refering to Tavaras. But maybe I was mistaken and he was talking about Dickerson.
In any event, part of the discussion involved Tavaras being lucky in getting hits in 2007. IMO, even if I used the wrong
BABIP number for 2007, the point about backing out bunts is still valid, IMO.

dougdirt
12-31-2008, 01:09 AM
To quote SteelSD (http://www.redszone.com/forums/showpost.php?p=1787524&postcount=954) from the same thread on the bunts, and I hope he doesn't mind me doing so

Yep. The BUH% overage was pretty close to being the entire difference between Taveras' 2007 and 2008 RC values. In 2007, his RC/600 PA was 80.48 with those extra 14 singles. Remove them and his RC/600 PA drops to 67.04. Now, knowing that those additional OB events also allowed for more SB attempts, his actual RC/600 PA value was pretty close to his 2008 result (62.36 RC/600 PA).

Here are the players since 2004 (tracking doesn't go back past that) who've produced a BUH% above 50% and with at least 10 Bunt Singles (Min. 400 PA):

2008- Jacoby Ellsbury (11 BUH, 57.9%)
2007- Willy Taveras (38 BUH, 64.4%)
2006- Ryan Zimmerman (10 BUH, 83.3%), Corey Patterson (21 BUH, 51.2%)
2005- No qualifiers
2004- Luis Castillo (10 BUH, 62.5%)

Now, here are the hitters who were able to repeat that feat:

<crickets>

Castillo hasn't produced anything higher than 48.5% since 2004. Zimmerman hit 100% in 2008 (3 measly bunts). Patterson's BUH rate dropped to 35.1% and a respectable 44.4% (in line with his career) in 2008 and look what happened to him anyway. Taveras' BUH dropped to around 40% in 2008 and his performance tanked in large part due to the BABIP hit he saw from it even though he was reaching base at a good BUH rate for anyone.

Overall, 50% BUH appears to be the high water mark for what we could normally expect for a premier speed bunter, with low to mid-40% ranges being far more likely from even the fastest guys. But that's not what we saw, of course, from Taveras in 2007 and it's completely unreasonable to expect him to be able to replicate that 64.4% rate again. In fact, it's most likely about 18 to 20% higher than he'll ever see from him again.

So while Taveras may be an extremely good bunter, to expect him to get anywhere near that amount of bunt hits again is crazy unless he bunts even more.

So even if he gets to that magical 50% rate, that would still have cut off 11 hits of his, leaving him as a .290/.339/.352.

RedsManRick
12-31-2008, 01:38 AM
Yes and no. Yes, Taveras' inflated BABIP in 2007 was a product of his efforts bunting for base hits. And yes, bunting for a base hit carries a different success rate than swinging away. And yes, Taveras is among the, if not the best bunters in baseball.

But let's look at how successful he's been. Fangraphs has a stat called BUH%. This is simply the percentage of times in which a bunt results in a hit. Here's Taveras' career:

2005: 47.7%
2006: 42.9%
2007: 64.4%
2008: 40.7%

One of these things is not like the others. Was Taveras a better bunter for base hits in 2007? Well, he set the modern record in 2007 with those 37 bunt hits. So what happened in the other three years? He wasn't bunting more, so it's not like he was extra picky in 2007.

In fact, let's look at all seasons in the last 4 in which a player had at least 10 bunt hits. And let's sort those by BUH%

Year Team Hits Bunts BUH%
Ryan Zimmerman 2006 Nats 10 12 83%
Willy Taveras 2007 Rockies 38 59 64%
Jacoby Ellsbury 2008 Red Sox 11 19 58%
Norris Hopper 2007 Reds 18 32 56%
Corey Patterson 2006 Orioles 21 41 51%
Melvin Mora 2005 Orioles 12 24 50%
Jason Ellison 2005 Giants 11 22 50%
Coco Crisp 2008 Red Sox 11 22 50%
Jose Reyes 2008 Mets 10 20 50%
Luis Castillo 2007 - - - 16 33 49%
Endy Chavez 2006 Mets 14 29 48%
Willy Taveras 2005 Astros 31 65 48%
Gregor Blanco 2008 Braves 17 36 47%
Nick Punto 2005 Twins 11 24 46%
Carlos Gomez 2008 Twins 30 66 46%
Julio Lugo 2007 Red Sox 10 22 46%
Scott Podsednik 2005 Whi Sox 17 38 45%
Corey Patterson 2008 Reds 12 27 44%
Corey Patterson 2005 Cubs 13 30 43%
Alexi Casilla 2008 Twins 16 37 43%
Juan Pierre 2006 Cubs 21 49 43%
Willy Taveras 2006 Astros 21 49 43%
Juan Pierre 2005 Marlins 29 71 41%
Willy Taveras 2008 Rockies 24 59 41%
Gerald Laird 2007 Rangers 10 25 40%
Michael Bourn 2008 Astros 13 34 38%
Jose Reyes 2007 Mets 11 30 37%
Cory Sullivan 2005 Rockies 10 28 36%
Dave Roberts 2005 Padres 12 34 35%
Corey Patterson 2007 Orioles 13 37 35%
Rafael Furcal 2006 Dodgers 10 29 35%
Juan Pierre 2008 Dodgers 11 32 34%
Dave Roberts 2006 Padres 10 33 30%
Rafael Furcal 2005 Braves 12 41 29%
Juan Pierre 2007 Dodgers 19 66 29%
Joey Gathright 2008 Royals 15 54 28%
Joey Gathright 2006 - - - 11 42 26%
Nook Logan 2005 Tigers 11 45 24%

The list is littered with really fast guys who can bunt: Juan Pierre, Coco Crisp, Jose Reyes. In the last 4 years, there have been 5 player seasons with a 50% or greater BUH%. Require 20 attempts and it's 3. The point? It's not a sustainable rate of success. For his career, Taveras is at 48.9%. Pierre 36.9%. Crisp 33.3% Reyes 38.4%. Ichiro 50.0%.

You are absolutely on point to call out the use of BABIP when it comes to Tavares. For most players, the BABIP of bunts is not an issue because they bunt in less than 2% of their PA. However, Tavares bunts often enough to have a significant impact on his batted ball statistics and that was especially true in 2007 when bunted nearly 15% of the time. Thus the normal assumptions fail.

That said, the general conclusion holds. Both his body of work and that of his peers suggest that no amount of bunting skill can sustain a rate of success north of 50%. Yes, Taveras is a great bunter. But 2007 was a fluke; he was lucky. No, it wasn't balls squeaking under gloves and blooping behind SS like normal BABIP "luck". But it was probably a few extra throws pulling the guy off the bag. It was a probably few extra bobbled barehands. It may very well have been perfectly tailored grass.

Taveras will almost always have what appears to be a "lucky" BABIP when compared to his LD%. But even as one of the best bunters in baseball, and even bunting more than anybody else, a BABIP north of .335-.340 is likely due to a nice dose of "luck" and not just his rare skill.

BuckeyeRedleg
12-31-2008, 01:49 AM
As I brought up in another thread, I wonder how much the infield grass in GABP will help his bunting for hits.

The dude can't hit his way out of a wet paper bag, so let's hope he's bunting all the time.

12-31-2008, 10:48 AM
So while Taveras may be an extremely good bunter, to expect him to get anywhere near that amount of bunt hits again is crazy unless he bunts even more.

So even if he gets to that magical 50% rate, that would still have cut off 11 hits of his, leaving him as a .290/.339/.352.

Baseball is a game of adjustments though. If teams consistently defend against the bunt hit vs Tavares, that will bring the infield in and give him a greater BABIP when he tries to get a normal hit.

Tavares has shown a very good skill for bunting for a base hit when the opportunity arises. True, he may not get 50 good opportunities next year to bunt for a base hit but it is still a skill that can raise his total BABIP.

Doug, this thread wasn't an attempt to start an argument about Tavaras or to call you out. The point is that you can't just look at BABIP and declare someone lucky or unlucky. Last offseason, people were using Belisle's BABIP allowed as evidence that Belisle was due for a breakout. That logic was clearly flawed, yet many people refused to dig deeper than the raw numbers. Belisle allowed a high BABIP because he stunk.

Cooper
12-31-2008, 10:51 AM
I realize there is a break even point when bunting doesn't make sense...but if you can't hit and you are bunting for a base hit at a .450 clip....wouldn't you try to push the number of bunts you take upwards? I say Willie needs to bunt 150 times this year -with the hope that 75 of them are successful.

12-31-2008, 10:56 AM
The list is littered with really fast guys who can bunt: Juan Pierre, Coco Crisp, Jose Reyes. In the last 4 years, there have been 5 player seasons with a 50% or greater BUH%. Require 20 attempts and it's 3. The point? It's not a sustainable rate of success. For his career, Taveras is at 48.9%. Pierre 36.9%. Crisp 33.3% Reyes 38.4%. Ichiro 50.0%.

I disagree. I think a 50% success bunt rate could be sustainable easily if the player only bunted at the optimal time (defense playing back).
However, in the real game of baseball, the situation sometimes dictates that maybe Tavares bunts in a non optimal situation. Let's say there's a really tough pitcher on the mound. If Tavaras has a 30% chance of bunting for a basehit but only a 20% chance of getting a regular hit, then it makes sense to try to bunt for a basehit, even though the odds are less than 50%.

In other words, no player makes it his goal to bunt at an over 50% success rate. Also, sometimes a player will bunt with a man on base and it is ruled "trying for a base hit" instead of a sacrifice. Obviously, any time Tavares is sacrificing, of course he will try to get a hit out of it.
Maybe Tavares didn't attempt many sacrifices.. the point is that no player ever has a goal of having a bunt hit success rate of a certain threshold, because that's not what wins games.

Tavares probably could've had a bunt hit success rate of over 50% in 2005 and 2006 if he was a little more selective.

You are absolutely on point to call out the use of BABIP when it comes to Tavares. For most players, the BABIP of bunts is not an issue because they bunt in less than 2% of their PA. However, Tavares bunts often enough to have a significant impact on his batted ball statistics and that was especially true in 2007 when bunted nearly 15% of the time. Thus the normal assumptions fail.

thanks, that's what I was trying to say

kpresidente
12-31-2008, 01:20 PM
I disagree. I think a 50% success bunt rate could be sustainable easily if the player only bunted at the optimal time (defense playing back).
However, in the real game of baseball, the situation sometimes dictates that maybe Tavares bunts in a non optimal situation. Let's say there's a really tough pitcher on the mound. If Tavaras has a 30% chance of bunting for a basehit but only a 20% chance of getting a regular hit, then it makes sense to try to bunt for a basehit, even though the odds are less than 50%.

That's a good point. The threshold should be you're OBP on non-bunt PAs. So if Willy has a .310 OBP when he doesn't bunt, he should be bunting up until defensive adjustments bring his BA on bunts down to .310 as well. That might equal quite a few bunt hits over the course of a season.

Of course, bunts aren't normal singles. They can only advance a runner one base instead of two, so you might be less inclined to drag bunt with men on base.

Johnny Footstool
12-31-2008, 01:46 PM
Yay, bunts.

bucksfan2
12-31-2008, 01:55 PM
That's a good point. The threshold should be you're OBP on non-bunt PAs. So if Willy has a .310 OBP when he doesn't bunt, he should be bunting up until defensive adjustments bring his BA on bunts down to .310 as well. That might equal quite a few bunt hits over the course of a season.

Of course, bunts aren't normal singles. They can only advance a runner one base instead of two, so you might be less inclined to drag bunt with men on base.

Bunts aren't normal singles but they can be productive with runners on early in the order. A bunt single that moves a runner over to third is much more beneficial if you have 3-4-5 hitters coming up. A bunt single to move a runner over to third when you have the 9 spot due up early in the game doesn't make much sense.

Caveat Emperor
12-31-2008, 01:57 PM
I realize there is a break even point when bunting doesn't make sense...but if you can't hit and you are bunting for a base hit at a .450 clip....wouldn't you try to push the number of bunts you take upwards? I say Willie needs to bunt 150 times this year -with the hope that 75 of them are successful.

If you bunted that often for a base hit, teams would eventually construct a defensive shift to cut off the drag bunt and just pull the 3b in close to stop the traditional bunt.

A judy-hitter like Taveras more than likely would be unable to make a defense pay for doing that.

mth123
12-31-2008, 02:21 PM
If you bunted that often for a base hit, teams would eventually construct a defensive shift to cut off the drag bunt and just pull the 3b in close to stop the traditional bunt.

A judy-hitter like Taveras more than likely would be unable to make a defense pay for doing that.

Exactly. The only option would be for the team to make a real fast track on the IF and hope his rollers would maintain enough steam to find the OF often enough to keep the defense honest. Of course, if they do that, then the whole EdE/Keppinger IF defense problems get even more damaging.

Ltlabner
12-31-2008, 02:24 PM
If you bunted that often for a base hit, teams would eventually construct a defensive shift to cut off the drag bunt and just pull the 3b in close to stop the traditional bunt.

A judy-hitter like Taveras more than likely would be unable to make a defense pay for doing that.

Which is a point that is often overlooked when people just want him to bunt more to boost his output.

Even a guy who's known for bunting has to maintain the illusion that he could pop one over the third baseman's head now-and-again to keep people honest.

Just strolling up and daring the other team by bunting every time won't work.

SteelSD
12-31-2008, 02:59 PM
I disagree. I think a 50% success bunt rate could be sustainable easily if the player only bunted at the optimal time (defense playing back).

However, in the real game of baseball, the situation sometimes dictates that maybe Tavares bunts in a non optimal situation. Let's say there's a really tough pitcher on the mound. If Tavaras has a 30% chance of bunting for a basehit but only a 20% chance of getting a regular hit, then it makes sense to try to bunt for a basehit, even though the odds are less than 50%.

In other words, no player makes it his goal to bunt at an over 50% success rate. Also, sometimes a player will bunt with a man on base and it is ruled "trying for a base hit" instead of a sacrifice. Obviously, any time Tavares is sacrificing, of course he will try to get a hit out of it.

Maybe Tavares didn't attempt many sacrifices.. the point is that no player ever has a goal of having a bunt hit success rate of a certain threshold, because that's not what wins games.

Tavares probably could've had a bunt hit success rate of over 50% in 2005 and 2006 if he was a little more selective.

The problem is that being more "selective" wouldn't necessarily be helpful to the player's overall value. When we're talking about a player who's actually pushing his overall BABIP up via BUH% (which is basically "Batting Average on Bunt Hit Attempts), reducing the number of bunt attempts may actually lower the hitter's value even if we see an increase in BUH%.

In 2006, 11.09% (49) of Taveras' 442 BIP events were non-sac bunt attempts. His BUH% was 42.9% (again, BABIP for Bunt Attempts). If we move a chunk of his bunt attempts out of that sample, they're going to be moved into his non-bunt event outcome sample. In 2006, his BABIP for non-bunt events was .310. I'd suggest there's nothing to be gained by swapping 42.9% BABIP probability for 31.0% probability simply to push his BUH% to 50% or higher. And considering how high his rates are already (only player in MLB to be over 40% from 2005-2008 with a min 10 BUH), I'd suggest that he's already optimizing his opportunities. I don't see this as a reasonable area for growth, which leaves us with the most probable result being an expectation of a BUH% of between 40 and 45% for 2009.

As an FYI- Willy Taveras led all of MLB position players (min 400 PA) in Sac Bunts in 2008 and finished 6th in 2006. I hardly think that official scorers are confused and wrongly attributing Sac attempts as Base Hit attempts.

RedsManRick
12-31-2008, 03:47 PM
FWIW Steel, in a 2007 article on BP, Dan Fox says that BUH% includes all bunt attempts, including those in which the batter was credited with a sacrifice.

http://www.baseballprospectus.com/article.php?articleid=6446

jojo
12-31-2008, 03:50 PM
What are the odds that Taveras will even need to bunt over one of the four likely guys that will bat in front of him?

SteelSD
12-31-2008, 04:11 PM
FWIW Steel, in a 2007 article on BP, Dan Fox says that BUH% includes all bunt attempts, including those in which the batter was credited with a sacrifice.

http://www.baseballprospectus.com/article.php?articleid=6446

Well, crap. Then it gets even worse as it means that Taveras actually produced a 73% BUH rate for non-sacrifice attempts in 2007. That takes us from highly unlikely to reproduce to pretty much impossible. It appears that only two players in history have every produced two seasons with an adjusted BUH rate above 70% (Matty Alou, Frank Taveras) with as many as 20 non-sac bunts. It also means that Willy Taveras is the only player among thousands to post a rate that high with 30 or more non-sac bunt attempts.

Thanks very much for the info.

01-01-2009, 11:20 PM
If you bunted that often for a base hit, teams would eventually construct a defensive shift to cut off the drag bunt and just pull the 3b in close to stop the traditional bunt.

A judy-hitter like Taveras more than likely would be unable to make a defense pay for doing that.

But he'd have a greater chance of getting a regular single with the 3b defending the bunt every time he came to the plate. Sure, that increased chance wouldn't make him a HOF player, but he should be trying to maximize his chances of avoiding an out.

If he feels he's got a 35-40% of reaching base by bunting at a given time, he should do it. Obviously, in real life, he's got to estimate that percentage.

A bunt single is at least as good as a walk, every time.

01-01-2009, 11:25 PM
The problem is that being more "selective" wouldn't necessarily be helpful to the player's overall value.

I agree with this. His job is to avoid outs by whatever manner possible.
I didn't say that clearly, but that was part of my point. Tavares shouldn't start the year with the goal of having a bunt hit success rate of %X. His goal should be to maximize his OBP by whatever means possible.

As an FYI- Willy Taveras led all of MLB position players (min 400 PA) in Sac Bunts in 2008 and finished 6th in 2006. I hardly think that official scorers are confused and wrongly attributing Sac attempts as Base Hit attempts.

Interesting. Thanks for the information. I was just speaking ancedotally. Sometimes I've seen hitters sacrifice, but since they run out the bunt, the scorer says "he was trying for a hit" and doesn't credit him with a sacrfice.
It's something to muddy the waters on the stats. Personally, I think any bunt that advances the runners should be counted as a sacrifice, and it shouldn't be a judgement call at all. But I agree, it's probably not a significant effect on any player.

mth123
01-01-2009, 11:36 PM
But he'd have a greater chance of getting a regular single with the 3b defending the bunt every time he came to the plate. Sure, that increased chance wouldn't make him a HOF player, but he should be trying to maximize his chances of avoiding an out.
If he feels he's got a 35-40% of reaching base by bunting at a given time, he should do it. Obviously, in real life, he's got to estimate that percentage.

A bunt single is at least as good as a walk, every time.

As CE said, he's such a judy hitter that he probably couldn't take advantage of it. The Reds would need to help him keep the IF back by making a hard fast surface. That would be a disaster for the EdE/Keppinger left side.

The bolded part of your quote is key. The Reds should be targeting guys who already do that. Taveras is a double whammy. His OBP says he doesn't do a good job of avoiding outs. Then, to make up for not being able to slug, he needs to try to steal bases to get more than 1B once in a while and that provides the defense with an additional chance to get him out after one of his less than frequent successes in getting on base in the first place. Make a bunch of outs at the plate and when you don't make some more on the bases. Its why a guy with world class speed only scores 60 to 80 runs or so at the top of the order in the small park at Houston or Coors field of all places. That ain't good.