Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Monday, March 03, 2014

Baseball Prospectus | Framing and Blocking Pitches: A Regressed, Probablistic Model

Rather than identifying a single strike zone and giving binary credit for each pitch relative to that strike zone’s borders (i.e., strike or no strike), our model gives partial credit for each pitch based on that pitch’s likelihood of being called a ball or a strike. To determine that, we created a probability map of likely calls… To reflect what is best known about the way the size and position of the strike zone shifts from count to count and batter to batter, we ran individual models for each set of batter and pitcher handedness as well as [type of pitch]. The smoothing parameters of each model were allowed to vary by count, so that while the general shape of the strike zone derived for each variable combination did not change, the width and height of it did (reflecting, for example, a larger strike zone on 3-0 counts than on 1-2 or 0-2 counts). We also accounted for the changing size of the strike zone from season to season (although these yearly changes are much smaller than the other changes we measured).

We also corrected the data in several ways before running these models. First, all pitch classifications were hand-labeled by Pitch Info to eliminate variability in pitch labels… To account for batter height differences, we normalized the height of each pitch by the batter’s height using what is now the standard formula (first published by Mike Fast). We also used the correction scheme that Mike published at BP for correcting the X and Y location of each pitch based on the likely distribution of pitch locations that each pitcher would use against left-handed hitters and right-handed hitters…

Rather than simply give a single credit for each pitch (~.14 runs) as has been done in many previous models, we looked at the count in which each pitch was framed and gave credit equal to the difference in runs between framing or not framing that pitch. For example, a frame in an 0-2 count was counted as more valuable than a frame in an 0-0 count, because a frame in an 0-2 count can result in a large change in run expectancy while a frame in an 0-0 count does not have quite the same impact… The run value for a framed pitch is the run value differential for that count… multiplied by the residual of the probability—in other words, if an 0-0 pitch is called a strike in a spot where it’s normally called a strike just 80 percent of the time, the catcher will get 20 percent of the available value (.08) for a total of .0004 runs credited (which will later be adjusted based on the pitcher and umpire impact). Failing to get a strike on the same pitch would result in a .0016 run deduction…

We empirically determined each pitcher’s value—to isolate it from each catcher’s value—by performing a WOWY (“With or Without You”) analysis… We also made systematic but small changes to the data based on the umpire who was calling each game…

we have regressed career totals to the league average… Because seasonal variability is different from career variability, we also regressed seasonal totals to career totals based on a similar formula…

You can find all of this new framing and blocking information in a couple place on the Baseball Prospectus site.

The District Attorney Posted: March 03, 2014 at 11:00 AM | 30 comment(s) Login to Bookmark
  Tags: baseball prospectus, defense, pitch framing, sabermetrics

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. DL from MN Posted: March 03, 2014 at 02:02 PM (#4665426)
Makes me wonder how much of the Twins poor strikeout totals to attribute to Ryan Doumit.
   2. DL from MN Posted: March 03, 2014 at 02:20 PM (#4665448)
Add 6 WAR/year to Johnny Bench's totals and how good does he look?
   3. The District Attorney Posted: March 03, 2014 at 03:00 PM (#4665509)
As you can imagine from the "regressed, probabilistic" part, this system gives less dramatic (although still certainly "large") results than the 50 runs/season estimate we saw in earlier research. The league leader by this system is usually around 25 runs/season.

In the six-year span from 2008-13, the top gainers are:

McCann 127 runs, J. Molina 116, Lucroy 94, Martin 91, Hanigan 74, D. Ross 65, Y. Molina/M. Montero 62, C. Stewart 43, Posey 41

And top losers are:

Doumit -124, Laird -83, Iannetta -75, Buck/Hundley -55, C. Santana -54, R. Johnson/K. Suzuki -47, Marson -42, A.J. Ellis -40
   4. cardsfanboy Posted: March 03, 2014 at 03:11 PM (#4665521)
I like the portion that they added on blocking, wish they would have gone into it a bit more. As a Cardinal fan, I assume that Yadier is the best in the game at it, but didn't see (or missed) the chart that showed the leaders of that particular stat.
   5. Jim Furtado Posted: March 03, 2014 at 03:33 PM (#4665555)
This is good stuff. I may finally pay Baseball Prospectus money for access to their web site.
   6. The District Attorney Posted: March 03, 2014 at 03:53 PM (#4665574)
Blocking is in the sortable stats. 2013 leaders were:

Lucroy/Yadier 14 runs, Martin 13, Wieters/W. Castillo 12, Posey 10, McCann/K. Suzuki/Ruiz 9, Hundley 8

And worst were:

J. Molina -3, Jaso/H. Sanchez/Quiroz/Shoppach/Romine/Brantly/Avila/T. Sanchez/R. Johnson -1, everyone else 0 or better. Huh, odd.
   7. cercopithecus aethiops Posted: March 03, 2014 at 04:03 PM (#4665584)
In the six-year span from 2008-13, the top gainers are:

McCann 127 runs, J. Molina 116, Lucroy 94, Martin 91, Hanigan 74, D. Ross 65, Y. Molina/M. Montero 62, C. Stewart 43, Posey 41


Chris Stewart must be outrageously good at this, since he only caught about 1700 innings during this span, while McCann had over 6000 and Molina had more than 7500. It really is too bad he hits like a pitcher.
   8. The District Attorney Posted: March 03, 2014 at 04:09 PM (#4665590)
Chris Stewart must be outrageously good at this, since he only caught about 1700 innings during this span, while McCann had over 6000 and Molina had more than 7500.
The "per 7000 opportunities" leaders are:

J. Molina 36, D. Ross 33, Grandal 32, Lucroy 31, C. Stewart 28, Zaun/Zunino 24, Hanigan 23, Corporan/McCann 22
   9. dr. scott Posted: March 03, 2014 at 05:25 PM (#4665666)
Chris Stewart must be outrageously good at this, since he only caught about 1700 innings during this span, while McCann had over 6000 and Molina had more than 7500. It really is too bad he hits like a pitcher.


I was thinking the same about Posey, but he did not show up in the most runs saved per 7000 ops in #8. He has barely played 3 seasons '10, '12, and '13.
   10. Zach Posted: March 03, 2014 at 05:35 PM (#4665671)
Like I said the other day, the methodology seems sound, but the magnitude of the effects they're getting seem unbelievable to me.

Look at Jose Molina:

36 runs per 7000 opportunities is half a run per 100 opportunities. So if every pitch were borderline, Jose Molina all by himself would be subtracting about .75 runs per game from the staff ERA.

I believe the authors when they say they've found a repeatable skill, and the players they identify as being good do have good defensive reputations. But I don't think they've completely nailed down the magnitude. Given the choice between Jose Molina and Mike Trout, I'm still going with Trout.
   11. The District Attorney Posted: March 03, 2014 at 07:04 PM (#4665728)
Given the choice between Jose Molina and Mike Trout, I'm still going with Trout.
Yeah, obviously.

if every pitch were borderline, Jose Molina all by himself would be subtracting about .75 runs per game from the staff ERA.
What does that mean, though? I mean, the hypothetical where "every pitch is borderline" is so removed from reality that I don't know how you can say what numbers would make sense in that world.

In this world, 25 runs per year is less than some "runs saved" numbers we've seen cited for corner outfielders who probably only get a few dozen non-"routine" opportunities per season... never mind for the guy catching every pitch. I don't know that it's the correct magnitude, but I certainly don't think it's prima facie impossible.
   12. villageidiom Posted: March 03, 2014 at 07:39 PM (#4665744)
I believe the authors when they say they've found a repeatable skill, and the players they identify as being good do have good defensive reputations. But I don't think they've completely nailed down the magnitude.
As mentioned in the other thread, we're now giving credit to catchers for changes in pitch-by-pitch outcomes that we do not give to hitters or pitchers. If a hitter grounds out to short, we don't adjust his WAR or whatever based on working the count. We just say he grounded out to short, which is a value of -N runs, or -N/10 wins.

For example, let's take a RHP and RHB. The pitcher throws a curveball that is just outside the low outside corner. Let's say the league distribution for that pitch is

50% take for a ball
25% swing and miss
10% foul out of play
10% ball in play or HR
5% take for a strike.

Right now, if a catcher converts the taken-ball outcome to a taken-strike outcome we credit the catcher with the linear weight difference between one-fourth a walk and one-third a strikeout. (Or a different amount, depending on if it's BPro or others doing the adjustment.) Yet the 35% of times the batter swings for a strike (25% swing and miss + 10% foul) we don't debit him for having converted a likely ball call into a strike, even though what he did is the equivalent of the pitch-framing catcher's achievement. Likewise, if the pitcher has a really nasty curveball that skews the outcome toward strikes, we don't credit the pitcher for having done the equivalent of the pitch-framing catcher. Credit is only given based on the PA outcome. If the nasty curve, or the poor pitch selection, produces Ks or BIP outs, then they get credit for the Ks and outs. For the catcher, credit is given irrespective of the PA outcome. Getting an extra called strike in a PA where the batter eventually hits a HR still gets the catcher credit for having produced as much value as if the called strike had been in a PA ending in a strikeout.

I assume the counterargument is that any gains by the hitter in a PA - working the count in his favor - is negated if he doesn't capitalize on it. A grounder to short in a 3-1 count returns the same value as an 0-2 grounder to short. But let's look at the flip side, a single to center after which the next batter hits a run-scoring double vs. a single to center after which the next batter hits an inning-ending DP. In the former case the outcome is a run; in the latter, no run. Yet in sabermetrics we treat the single to center as having produced the same run "value" regardless of the actual run outcome. This is the same concept.

The big difference is that in the double vs. DP example the run depends on another player, and thus to remove the context of the other hitter we need to look at the outcomes by the batter who hit the single to center. In the grounder to short example, both the pitch outcomes and the PA outcome are not influenced by any other hitter.

But they are influenced by the catcher. And the pitcher. And possibly the baserunners. And the umpire. If we're going to credit the catchers for these intermediate changes in value, but not the other players for doing the same thing, it's apples vs. oranges.
   13. PreservedFish Posted: March 03, 2014 at 08:15 PM (#4665759)
The thing that's interesting about this to me is that every part of the league will be adjusting to the research. There are probably catchers right now in spring training watching video of Jose Molina vs Ryan Doumit. And there are probably umpires that are doing the same. The magnitude of this effect might change dramatically in the next few years.
   14. cardsfanboy Posted: March 03, 2014 at 08:27 PM (#4665763)
As mentioned in the other thread, we're now giving credit to catchers for changes in pitch-by-pitch outcomes that we do not give to hitters or pitchers. If a hitter grounds out to short, we don't adjust his WAR or whatever based on working the count. We just say he grounded out to short, which is a value of -N runs, or -N/10 wins.


I'm glad people understood what I was saying there. It took me a while to get that point across because I was mis-explaining it and needed prompting to get it to mean what I was thinking.
   15. Walt Davis Posted: March 03, 2014 at 08:43 PM (#4665768)
In this world, 25 runs per year is less than some "runs saved" numbers we've seen cited for corner outfielders who probably only get a few dozen non-"routine" opportunities per season... never mind for the guy catching every pitch. I don't know that it's the correct magnitude, but I certainly don't think it's prima facie impossible.

I don't necessarily disagree with you but this is not a great analogy. The effect of an extra play by an OF is rather massive, on the order of .8 runs. The average value of a strike is .14 runs but in this system the C is only getting credit for a proportion of that. It might require 10-12 framed pitches to create a run value of .8. (On the other hand, maybe the run value for the OF isn't being properly weighted by the probability that a play is made on that ball?)

#12: a fair point but you may have missed this from the excerpt: "We empirically determined each pitcher’s value—to isolate it from each catcher’s value—by performing a WOWY (“With or Without You”) analysis."

I pointed out the need to adjust for the pitcher effect the first time I saw one of Fast's articles and Tango ran a Q&D analysis and found some huge pitcher effects that put Molina to shame (Lowe was a burglar that season).

And I could have sworn those early analyses showed McCann to be a bit of a butcher. Am I misremembering?

#6: yeah, those blocking numbers don't seem to make sense. Maybe the really terrible guys don't get to catch enough innings to get listed on the leaderboard but are so bad they drag the "average" way down. If so that might suggest that the range for "starting" Cs is about -7 to +7.
   16. Walt Davis Posted: March 03, 2014 at 08:46 PM (#4665769)
Oops, I skipped the big picture.

Hooray! I'm not going to dig into the precise methodology or data but these folks are definitely trying to do it correctly and address some of the basic concerns I've raised here in the past.
   17. Zach Posted: March 03, 2014 at 08:48 PM (#4665770)
I don't know that it's the correct magnitude, but I certainly don't think it's prima facie impossible.

Last year, per 100 pitches thrown, Clayton Kershaw's fastball was 1.81 runs better than average, his slider 0.63 runs, and his curveball 2.69. If we take these numbers literally, that means that Clayton Kershaw throwing a frameable fastball to Jose Molina would be 30% more effective, 81% more effective with the slider, 20% more effective with his curveball.

I just don't think an effect that large could go unnoticed. You would expect to see lots of pitchers --even star pitchers -- going to pieces on days when they have to pitch to the backup catcher.

Put it this way: for an average pitcher on a frameable pitch, throwing to Jose Molina is as big a change as having Clayton Kershaw's slider. Do you buy it?
   18. Zach Posted: March 03, 2014 at 08:52 PM (#4665771)
Like I say, I think these guys are doing this the right way. I'm just not sure that the thing they're measuring corresponds to show-up-on-the-scoreboard runs in the same way that a play made or not made by a shortstop. I'm willing to say that Jose Molina generates 36 "Molinas" per 7000 chances, I'm just not convinced that 1 Molina equals 1 run.
   19. Der-K and the statistical werewolves. Posted: March 03, 2014 at 10:08 PM (#4665793)
As previously noted, something's up with the blocking numbers. They need (IMO) to have it be zero-sum on an MLB level.
   20. cardsfanboy Posted: March 03, 2014 at 10:53 PM (#4665811)
As previously noted, something's up with the blocking numbers. They need (IMO) to have it be zero-sum on an MLB level.


Why? I understand other things being zero sum, but with blocking pitches I don't think it has to be zero sum. I guess it depends on how they do it, but in theory if they are rating where the ball is thrown and what an average catcher does with that particular pitch in regards to blocking it or not, then there is no reason to think that the distribution of "blockable" pitches will equal out.

   21. villageidiom Posted: March 03, 2014 at 11:48 PM (#4665824)
#12: a fair point but you may have missed this from the excerpt: "We empirically determined each pitcher’s value—to isolate it from each catcher’s value—by performing a WOWY (“With or Without You”) analysis."
No, I saw that. The way I read TFA is that they are attributing value to the pitcher solely for the sake of adjusting the catcher value from blocking/framing for the mix of pitchers they have caught; they are not actually adjusting pitcher WAR to reflect the additional value attributable to them. IOW, if without the adjustment they had a catcher at +10 runs, but based on the WOWY work they have the "true" catcher effect at +2 runs, then the catcher is given +2 runs but the pitchers don't get an additional +8 runs among them.

It's still apples and oranges, but BPro has painted the oranges red.

EDIT: I don't mean for that to sound as disparaging as it does. BPro is doing some really good stuff here. I'm just saying it doesn't address the differences I cite; it only mitigates the appearance of those differences.
   22. Sunday silence Posted: March 04, 2014 at 12:13 AM (#4665831)

I'm glad people understood what I was saying there. It took me a while to get that point across because I was mis-explaining it and needed prompting to get it to mean what I was thinking...


Right. I think that's the advantage of having a forum like this where people remain a constant and the arguments re-appear again and again with slight modifications. Rather than other institutions like courts or politics, where someone makes some dumb argument, runs away and then is never to account for it again. Like the Domino Theory or The Glove Doesnt Fit or [insert favorite crap sound bite].

I consider this place like some ancient Greek forum where we respect each other because we have a long history and we can recall back our previous positions to see if we are consistent or not.
   23. cardsfanboy Posted: March 04, 2014 at 12:18 AM (#4665833)
No, I saw that. The way I read TFA is that they are attributing value to the pitcher solely for the sake of adjusting the catcher value from blocking/framing for the mix of pitchers they have caught;


That is the way I read it also.
   24. Der-K and the statistical werewolves. Posted: March 04, 2014 at 12:20 AM (#4665834)
I'm with you, vi. (And do want to praise what they're doing.)

cfb, as of now, the blocking stats are devoid of context. For instance, is saving six runs through pitch blocking good or bad? (Better than five, worse than seven - sure.) And, if your answer is 'we don't know without knowing how many opportunities they had,' then it's not yet a very useful metric.
   25. cardsfanboy Posted: March 04, 2014 at 01:42 AM (#4665872)
cfb, as of now, the blocking stats are devoid of context. For instance, is saving six runs through pitch blocking good or bad? (Better than five, worse than seven - sure.) And, if your answer is 'we don't know without knowing how many opportunities they had,' then it's not yet a very useful metric.


I think it's still an infancy stat, if you look at how they introduced it on the site (a side bar with no real article on it) it appears to me to be an early version of the stat and that they are still fine tuning.

I agree with the context, but I was just pointing out that it doesn't have to be a sum zero stat. Let's say you have a very good catcher at blocking balls, and he happens to be the catcher for a knuckleball pitcher. A good blocker, with a lot of chances is going to generate a pretty high positive total using any system that ranks a players ability to block a ball in the dirt based upon compared to average.

With other fielding stats, there are so many chances that it naturally progress's to a zero sum, but with catchers you are dealing some what with selective choices in both pitch selection and on who catches the more wild pitchers.

   26. epoc Posted: March 04, 2014 at 02:48 AM (#4665878)

Last year, per 100 pitches thrown, Clayton Kershaw's fastball was 1.81 runs better than average, his slider 0.63 runs, and his curveball 2.69. If we take these numbers literally, that means that Clayton Kershaw throwing a frameable fastball to Jose Molina would be 30% more effective, 81% more effective with the slider, 20% more effective with his curveball.

I just don't think an effect that large could go unnoticed. You would expect to see lots of pitchers --even star pitchers -- going to pieces on days when they have to pitch to the backup catcher.


It seems reasonable to me that such an effect would go unnoticed. There are something like fifty frameable pitches per game, and Molina is adding about .26 runs (vs. average) in that amount of pitches (according to this study). A 3.6 pitcher is nearly indistinguishable from a 3.86 pitcher, especially over the course of just one game. If a 3.6 pitcher threw a 3.86 game to a backup, no one would notice. And then of course random variation is at play, so the results are going to be impossible to notice in a one-game sample.


Put it this way: for an average pitcher on a frameable pitch, throwing to Jose Molina is as big a change as having Clayton Kershaw's slider. Do you buy it?


Yeah, I buy that, at least in theory. Being able to get a borderline pitch reliably called for a strike seems like it should be about as valuable as a good pitch thrown.

Like I say, I think these guys are doing this the right way. I'm just not sure that the thing they're measuring corresponds to show-up-on-the-scoreboard runs in the same way that a play made or not made by a shortstop. I'm willing to say that Jose Molina generates 36 "Molinas" per 7000 chances, I'm just not convinced that 1 Molina equals 1 run.


I think this, along with vi's 12, is really interesting. And you don't have to stop with the pitch-framing runs. You can say the same about the whole principle behind WAr - converting base hits and outs into runs, and then converting runs into wins. Sure, there's a theoretical relationship, but how closely does that theoretical relationship track with how games are actually being won and lost on the field? It is kind of cool to break the game down into smaller and smaller discrete bits, but each subdivision makes the margin of error that much larger.

I do think the work we've seen on pitch-framing shows that it's a repeatable skill and a very valuable one at that. And intuitively I find it easy to understand why it would be so valuable. But I side with some of the other posters here in being hesitant to attribute extreme win values to the theoretical runs saved through framing. My question now is whether I should be similarly hesitant to attribute win values to the theoretical runs earned or saved through hitting, defense, and baserunning.

   27. cardsfanboy Posted: March 04, 2014 at 03:08 AM (#4665880)
I think this, along with vi's 12, is really interesting. And you don't have to stop with the pitch-framing runs. You can say the same about the whole principle behind WAr - converting base hits and outs into runs, and then converting runs into wins. Sure, there's a theoretical relationship, but how closely does that theoretical relationship track with how games are actually being won and lost on the field? It is kind of cool to break the game down into smaller and smaller discrete bits, but each subdivision makes the margin of error that much larger.


Part of the point that they were talking about is that it really doesn't matter what scale we rate War or any of these stats on, they don't have to actually represent actual wins above average, just that the system is fairly accurate and consistent regardless of the wording used. When you start to mix and match different "scales" into one system, then there is very likely a problem. Catchers defense is rightfully measured on it's impact on individual pitches, while war and other run value stats are measured on individual at bats, the result of the at bat. When you start to combine the two, you are going to run into problems.

I argued on the other thread that the using similar terms (wins or runs saved) that we have used in other systems is unintentionally misleading. They are measuring on different scales.
   28. Der-K and the statistical werewolves. Posted: March 04, 2014 at 11:40 AM (#4666016)
When you start to combine the two, you are going to run into problems.

Yeah, but I don't think that's insurmountable. We see this already in other contexts and it's rare that it messes us up in a significant way wrt position players (defensive sub / pr types being the big one in my mind).

I agree with the context, but I was just pointing out that it doesn't have to be a sum zero stat. Let's say you have a very good catcher at blocking balls, and he happens to be the catcher for a knuckleball pitcher. A good blocker, with a lot of chances is going to generate a pretty high positive total using any system that ranks a players ability to block a ball in the dirt based upon compared to average.

With other fielding stats, there are so many chances that it naturally progress's to a zero sum, but with catchers you are dealing some what with selective choices in both pitch selection and on who catches the more wild pitchers.

I got your point here and agree with you about opportunities and potential impact. That still could (and, in my mind, should) be represented as zero-sum. Compare savings v. expected savings, subject to pitch location, type, and movement. Show two ways: total impact (indexed to zero) and as a rate. Other breakdowns could be useful as well, but that's a good start for 20K feet discussions. As of now, I have no idea what a given number means... just that this stat, still in its infancy, shouldn't be added to the framing numbers (which are zero sum) on the player cards.

Also, the point that any credits/demerits shown here ought to be accounted for in pitching/other metrics is an important one.
   29. epoc Posted: March 04, 2014 at 12:52 PM (#4666091)
Yeah, but I don't think that's insurmountable. We see this already in other contexts and it's rare that it messes us up in a significant way wrt position players (defensive sub / pr types being the big one in my mind).


I agree that it's conceptually no different than turning stolen bases into theoretical runs and then into theoretical wins. The ball/strike call isn't fundamentally different than any other measurable event on the baseball field. And WAr already includes replacement runs and positional adjustments which aren't baseball events at all, so we were already at a pretty extreme level of abstraction, and of mixing different "scales." I don't think the abstraction or the mixing makes these stats inaccurate, but there's necessarily some imprecision, and on a player level I wonder just how much imprecision there is. For instance, does basing roster decisions on these framing stats give teams any real advantage over basing such decisions on scouting reports about catcher defense? Is 6 WAr more precise than a 7 on a scouting report?
   30. Zach Posted: March 04, 2014 at 02:28 PM (#4666178)
It seems reasonable to me that such an effect would go unnoticed. There are something like fifty frameable pitches per game, and Molina is adding about .26 runs (vs. average) in that amount of pitches (according to this study). A 3.6 pitcher is nearly indistinguishable from a 3.86 pitcher, especially over the course of just one game. If a 3.6 pitcher threw a 3.86 game to a backup, no one would notice. And then of course random variation is at play, so the results are going to be impossible to notice in a one-game sample.

.26 runs is 4 wins over the course of a season. Andrelton Simmons last year was worth 5 wins on defense according to BBref DWAR. That was the best season ever recorded. Ozzie Smith's best season ever was 4.7. This is a big, big effect.

Thinking about the issue more carefully, I can be more articulate about what I don't like. Pitch framing is a parameter which measures the fluctuation of run expectancy within an at-bat. However, conventional linear weights measure the total change in run expectancy from the beginning of an at-bat to the end. It's similar to the way that a cool breeze can be moving at two miles an hour, but the individual air molecules can be moving at several hundred.

If you could accurately measure all the individual skills which affect an at-bat on a pitch by pitch basis, you could sum them over all pitches to get a fair measure of a player's value. And in fact, you would have to -- a single is worth about half a run, so the sum of all changes in run expectancy for an at bat which ends in a single must sum to that. But the summed magnitude in the changes to run expectancy due to each individual pitch can easily be much larger than that.

Making things worse is that we're only looking at one skill of one player involved in the at-bat, instead of all skills of all three players. So a catcher that, say, is good at framing but bad at pitch calling (so that he encourages pitchers to nibble) gets credit for framing but no debit for calling marginal pitches. And we're not debiting or crediting those runs to either the batter or the pitcher when we do it.

I think it's quite possible that the run value of a skill measured on a pitch-by-pitch basis will be larger in general than the net defensive value of a player measured on a play-by-play basis.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Rough Carrigan
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogNL WILD CARD 2014 OMNICHATTER
(2 - 8:10am, Oct 01)
Last: DKDC

NewsblogOT: NFL/NHL thread
(8174 - 8:01am, Oct 01)
Last: Norcan

NewsblogWSJ: Playoff Hateability Index
(16 - 8:00am, Oct 01)
Last: Swoboda is freedom

NewsblogOT: Politics, September, 2014: ESPN honors Daily Worker sports editor Lester Rodney
(4083 - 7:59am, Oct 01)
Last: Bitter Mouse

NewsblogAL WILD CARD GAME 2014 OMNICHATTER
(1128 - 7:39am, Oct 01)
Last: Joey B. has reignited his October #Natitude

NewsblogSpector: Stats incredible! Numbers from the 2014 MLB season will amaze you
(35 - 7:28am, Oct 01)
Last: RMc is a fine piece of cheese

NewsblogThe Calm-Before-The-Storm and Postseason Prediction OMNICHATTER, 2014
(111 - 7:14am, Oct 01)
Last: Jolly Old St. Nick Is A Jolly Old St. Crip

NewsblogThe Economist: The new market inefficiencies
(19 - 2:21am, Oct 01)
Last: David Concepcion de la Desviacion Estandar (Dan R)

Hall of MeritMost Meritorious Player: 2014 Discussion
(14 - 2:17am, Oct 01)
Last: bjhanke

Hall of MeritMost Meritorious Player: 1958 Ballot
(13 - 1:55am, Oct 01)
Last: neilsen

NewsblogBrown: Winners And Losers: MLB Attendance In 2014, Nearly 74 Million Through The Gate
(33 - 11:27pm, Sep 30)
Last: Bhaakon

NewsblogMLB’s Biggest Star Is 40 (And He Just Retired). That Could Be A Problem.
(76 - 11:27pm, Sep 30)
Last: Walt Davis

NewsblogESPN: Ron Gardenhire out after 13 Seasons with Twins
(42 - 10:49pm, Sep 30)
Last: The District Attorney

Hall of MeritMost Meritorious Player: 1959 Discussion
(6 - 10:35pm, Sep 30)
Last: MrC

NewsblogOT: The Soccer Thread, September 2014
(455 - 10:08pm, Sep 30)
Last: The John Wetland Memorial Death (CoB)

Page rendered in 0.4754 seconds
52 querie(s) executed