SoxProspects News
|
|
|
|
Legal
Forum Ground Rules
The views expressed by the members of this Forum do not necessarily reflect the views of SoxProspects, LLC.
© 2003-2024 SoxProspects, LLC
|
|
|
|
|
Forum Home | Search | My Profile | Messages | Members | Help |
Welcome Guest. Please Login or Register.
Vazquez Pitch Framing Log
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 12, 2014 12:44:35 GMT -5
I thought I'd edit this top post each game so that the data would always be handy.
All data is from BP. The columns are Pitch Framing Chances, Expected and Actual Strikes, the difference, and then Runs Saved in a count-dependent and then count-neutral manner. I'll post what I find out about these metrics later in this thread.
Date SP Chances Exp Str Act Str Extra Act Runs Raw Runs WAR 9-Jul DLR 68 22.6 23 0.4 0.0 0.1 1.1 11-Jul Lac 103 29.3 31 1.7 0.3 0.2 2.3 13-Jul Buc 60 24.3 27 2.7 0.6 0.4 7.9 19-Jul DLR 74 19.3 22 2.7 0.6 0.4 6.4 21-Jul Lac 46 20.3 23 2.7 0.2 0.4 7.7 23-Jul Buc 68 21.8 25 3.2 0.5 0.5 7.6 24-Jul DLR 83 27.6 28 0.4 -0.2 0.0 -0.7 26-Jul Lac 70 24.2 26 1.8 0.0 0.3 3.2 27-Jul Web 90 25.4 27 1.6 0.7 0.3 4.8 29-Jul DLR 88 26.3 31 4.7 0.8 0.7 8.6 30-Jul Wor 90 31.1 32 0.9 0.7 0.1 3.1 8/1-3 Misc 216 63.8 70 6.2 0.4 0.9 3.6 8/5-6 Misc 146 37.9 41 3.1 0.6 0.5 3.8 8/7-8 Misc 140 42.7 48 5.3 0.7 0.8 5.7 9-Aug Buc+ 152 51.2 54 2.8 0.7 0.4 3.3 12-Aug Kel 87 22.1 28 5.9 0.0 0.9 7.7 14-Aug Web 72 18.2 19 0.8 0.7 0.1 3.9 8/15-16 Misc 175 57.4 62 4.6 0.5 0.6 3.4 18-Aug Wor 74 24.5 30 5.5 0.8 0.9 12.2 19-Aug [Web] 61 14.5 12 -2.5 -1.5 -0.4 -12.1 8/21-5 Misc 355 109.0 111 2.0 0.2 0.2 0.8 26-Aug DLR 106 29.8 32 2.2 0.1 0.3 2.4 29-Aug Ran 85 22.4 28 5.6 0.7 0.9 10.3 31-Aug Buc 53 19.6 24 4.4 0.5 0.6 11.2 1-Sep DLR 77 22.4 22 -0.4 -0.2 0.0 -0.8 3-Sep Ran 83 23.0 25 2.0 0.1 0.2 2.1 4-Sep Wor 76 29.3 29 -0.3 0.1 0.0 0.4 9/5-6 We,Bu 168 54.5 64 9.5 1.2 1.4 8.3 7-Sep DLR 77 25.4 27 1.6 0.2 0.2 2.7 9-Sep Ran 82 22.2 24 1.8 0.3 0.3 3.8 11-Sep Buc 74 29.4 28 -1.4 -0.3 -0.2 -3.2 9/12-13 142 49.3 53 3.7 0.3 0.6 3.8 9/14-18 314 99.8 110 10.2 1.7 1.5 5.1 Tot 3655 1140.6 1236 95.4 12.0 14.2 3.8*
* May not be accurate; I need to re-derive a possible correction algorithm
7/11 comment: So far in these two games, meaningless as they may be in the long run, he's framed at about a +2 WAR rate. As we discussed in the CVaz recalled thread, he had little chance for framing on the 9th because the ump called such a good game.
7/13: I don't think he can keep doubling his raw (count-neutral) runs saved every game. In these three games (regressing the clutch framing), he's framed at +30 runs per year; if that were a career figure it would rank him 5th out of 127 catchers, after J. Molina, Ross, McCann, and Chris Stewart. The last two game have been Molina-level. So far, so good, and of course observers (including his pitchers) have been raving.
7/19. Nope, didn't double his numbers, just matched them. He's up to 36.4 R / 120 G, which is 4.0 WAR, and would place him 3rd behind Molina and Ross.
7/21. I've added a column, WAR, which is the seasonal WAR from pitch framing if the entire season of 1050 innings were played at that quality. I'm now adjusting that for number of framing chances -- the average catcher would have had 400 in 45 innings behind the plate.
7/23. Well, we're now waiting for him to have a merely good game, let alone an off night.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 12, 2014 14:12:55 GMT -5
Some basic pitch-framing facts, based on 2008-2013 data:
There are on average 8.9 pitch framing chances per defensive inning.
There is actually a stronger year-to-year correlation of framing runs per inning than of framing runs per normalized inning (8.9 chances). This is helpful, since it indicates that for certain purposes, the noise introduced by variations in framing chances per inning can be ignored. Specifically, we can use an innings cutoff for sample sizes rather than a framing chances cutoff, making the discussion of sample sizes much easier to follow.
You can get a very robust year-to-year correlation of .79 for actual framing runs saved per inning, all the way down to a minimum of 340 innings (about 39 games) in the less active year. Basically, the metric works for regular and backup catchers, just not scrubs.
The correlation for count-neutral runs saved is even stronger (.85). However, the difference between the two, which is to say clutch framing by count, has a statistically significant correlation of .23. Exactly 20% of clutch framing is predictive year-to-year, versus 85% of theoretical framing. I would use .77 * Theoretical + .23 * Actual for an annual skill-measuring metric. I'll have to think about whether you'd use the same formula for a career.
|
|
|
Post by philsbosoxfan on Jul 14, 2014 5:28:42 GMT -5
I'm curious. What is the range for season long WAR across the major leagues when you isolate on pitch framing alone? If sabermetrics have enough data, how much impact does it have in the grand scheme of things ?
|
|
|
Post by jmei on Jul 14, 2014 7:50:04 GMT -5
That'll depend on the methodology you use, but by the Statcorner framing leaderboards (which, unlike the BP ones, don't require a subscription), the best framers in a given year are usually 2-3 wins better than average, while the worst framers in a given year are similarly 2-3 wins worse than average. You can play with the leaderboards yourself here (remember that 10 runs = one win). I believe the BP leaderboards have roughly a similar spread. ADD: RAA = runs above average; that's generally the column you want to look at.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 14, 2014 11:42:00 GMT -5
That'll depend on the methodology you use, but by the Statcorner framing leaderboards (which, unlike the BP ones, don't require a subscription), the best framers in a given year are usually 2-3 wins better than average, while the worst framers in a given year are similarly 2-3 wins worse than average. You can play with the leaderboards yourself here (remember that 10 runs = one win). I believe the BP leaderboards have roughly a similar spread. ADD: RAA = runs above average; that's generally the column you want to look at. BP's range is at least twice as large. The standard deviation of runs per year (1050 defensive innings, roughly 120 games) for (count-sensitive) career framing is 19.1. Jose Molina is +51, David Ross is +45, Ryan Doumit is -43 and Rob Johnson is -37. Check out the top post for yesterday's data and, within a few minutes, a comment.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 14, 2014 12:00:40 GMT -5
(remember that 10 runs = one win). ADD: RAA = runs above average; that's generally the column you want to look at. That's a function of offensive level, and this year, as of right now, it's down to 9.1 runs per win.
|
|
|
Post by ramireja on Jul 14, 2014 16:48:09 GMT -5
Given that Ross is a great pitch framer himself....what do people think of bringing him back for another year if he's agreeable to further mentor Vazquez (provide insight to our pitching staff) until Swihart is 'ready?'
|
|
radiohix
Veteran
'At the end of the day, we bang. We bang. We're going to swing.' Alex Verdugo
Posts: 6,207
|
Post by radiohix on Jul 14, 2014 21:04:41 GMT -5
He's a catching virtuso compared to those guys! I love him.
|
|
|
Post by philsbosoxfan on Jul 14, 2014 22:45:10 GMT -5
That'll depend on the methodology you use, but by the Statcorner framing leaderboards (which, unlike the BP ones, don't require a subscription), the best framers in a given year are usually 2-3 wins better than average, while the worst framers in a given year are similarly 2-3 wins worse than average. You can play with the leaderboards yourself here (remember that 10 runs = one win). I believe the BP leaderboards have roughly a similar spread. ADD: RAA = runs above average; that's generally the column you want to look at. BP's range is at least twice as large. The standard deviation of runs per year (1050 defensive innings, roughly 120 games) for (count-sensitive) career framing is 19.1. Jose Molina is +51, David Ross is +45, Ryan Doumit is -43 and Rob Johnson is -37. Check out the top post for yesterday's data and, within a few minutes, a comment. Now I'm curious. Are BP's numbers subjective, ie, video review, or are they computer generated based on pitch fx loctions ? After seeing a post by somebody in one of the game threads on the ratings of JBJ's catches, I really don't have much any faith in BP's ability to analyze subjectively.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 14, 2014 22:56:44 GMT -5
BP's range is at least twice as large. The standard deviation of runs per year (1050 defensive innings, roughly 120 games) for (count-sensitive) career framing is 19.1. Jose Molina is +51, David Ross is +45, Ryan Doumit is -43 and Rob Johnson is -37. Check out the top post for yesterday's data and, within a few minutes, a comment. Now I'm curious. Are BP's numbers subjective, ie, video review, or are they computer generated based on pitch fx loctions ? After seeing a post by somebody in one of the game threads on the ratings of JBJ's catches, I really don't have much faith in BP's ability to analyze subjectively. Total pitch/fx data orgy. I think they adjust for pitcher and umpire, too. And that wasn't BP that had credited Bradley with no <10% probability catches, that was BIS.
|
|
|
Post by mgoetze on Jul 14, 2014 23:05:29 GMT -5
And that wasn't BP that had credited Bradley with no <10% probability catches, that was BIS. The people whose fielding stats I've been railing against are Inside Edge, not BP or BIS.
|
|
|
Post by philsbosoxfan on Jul 14, 2014 23:08:45 GMT -5
Now I'm curious. Are BP's numbers subjective, ie, video review, or are they computer generated based on pitch fx loctions ? After seeing a post by somebody in one of the game threads on the ratings of JBJ's catches, I really don't have much faith in BP's ability to analyze subjectively. Total pitch/fx data orgy. I think they adjust for pitcher and umpire, too. And that wasn't BP that had credited Bradley with no <10% probability catches, that was BIS. Cool glad to hear that. I was thinking the post I saw pointed to BP not BIS but I could be wrong, wouldn't be the first time. Either way, I have more confidence in computer generated numbers. (Can't wait for field fx, hopefully I'll see those results sometimes before I'm dead).
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 14, 2014 23:18:49 GMT -5
And that wasn't BP that had credited Bradley with no <10% probability catches, that was BIS. The people whose fielding stats I've been railing against are Inside Edge, not BP or BIS. I didn't realize they were different! For a company that's been around as long as they have (1993), they've stayed under the radar. I'd heard their name before, but not for a long while ... I think I thought that they were a BIS trademark, maybe after being acquired. Or something. To say something on-topic, how long before we start seeing this data included in catcher's WAR? That will really alter some opinions (and make WAR less credible to its skeptics, incidentally).
|
|
|
Post by mgoetze on Jul 14, 2014 23:31:19 GMT -5
To say something on-topic, how long before we start seeing this data included in catcher's WAR? That will really alter some opinions (and make WAR less credible to its skeptics, incidentally). I think most people who care about WAR would easily accept pitch framing being included in catcher WAR ... but to be consistent you would also have to adjust pitcher WAR accordingly, which I suspect would be more controversial.
|
|
|
Post by philsbosoxfan on Jul 14, 2014 23:31:20 GMT -5
Which was really somewhat the reason for the original question. Since WAR is based on replacement level, not average, good defensive catchers should end up with an unexpected boost to their value ranking. Catching has always been the exception because nobody really knows the actual effect of the 'intangibles'.
The future trade value of CVaz and Swihart could increase substantially. Maybe the Sox knew exactly what they were doing when they drafted 6 catchers.
On the other hand, embedding a chip in a baseball to determine balls and strikes would completely nullify any pitch framing value.
ADD: The more you (I) think about it the bigger the potential effect on how baseball is viewed. For one thing, if good defensive catchers get a WAR boost, somebody has to have an off-setting WAR decline. Also, just looking at things like pitcher K's/9. How much is that affected by throwing to a good pitch framer vs throwing to a bad framer. Pretty much all the pitching stats are affected because pitchers are pitching with different counts.
Put on a one to one basis we're all familiar with, how much benefit did Lester get with Ross catching him vs. Peavy with AJP catching him. There's obviously a difference but maybe the difference isn't as big as we think it is.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 20, 2014 8:02:44 GMT -5
Check out the updated log.
The thing is, he looks as good as his numbers.
|
|
|
Post by chavopepe2 on Jul 20, 2014 8:58:41 GMT -5
Check out the updated log. The thing is, he looks as good as his numbers. Hey Eric, I just wanted to thank you for maintaining this. It is something I find very interesting to follow.
|
|
|
Post by philsbosoxfan on Jul 20, 2014 10:41:54 GMT -5
As do I but I have a question.
Is there any way to determine what portion is on the catcher and what portion is on the pitcher ?
It occurs to me that the old saying that a pitcher can get strikes called if he's off the plate but is consistently there. We used to hear that Maddox, for example, could get away with throwing off the plate because he was so consistently good at throwing the ball where he wanted. Using Ross' steller stats as an example, how much is influenced by the fact that he's Lester's caddy ?
|
|
|
Post by soxfan06 on Jul 20, 2014 11:14:02 GMT -5
Another question, how good of a bat would we need to see out of Vazquez in order to consider him an elite catcher in this league? I know that it helps a lot that the position is weak throughout the league, but everything I've read paints Vazquez as one of the best defensive backstops in the league with ease. So what would he realistically have to do at the plate to make us say, wow this kid is one of the 5 best catchers in the Majors (obviously that may not happen, this is more of a what if question)?
|
|
|
Post by WindyCityRedSox169 on Jul 20, 2014 11:31:48 GMT -5
As do I but I have a question. Is there any way to determine what portion is on the catcher and what portion is on the pitcher ? It occurs to me that the old saying that a pitcher can get strikes called if he's off the plate but is consistently there. We used to hear that Maddox, for example, could get away with throwing off the plate because he was so consistently good at throwing the ball where he wanted. Using Ross' steller stats as an example, how much is influenced by the fact that he's Lester's caddy ? That seems like an impossible feat on an extremely subjective matter that there is no metric used to define. Pitchers that get calls when they have command of the strike zone/an ace and pitchers that don't because of youth/lack of command are made by the human element of the game. There is no metric that evaluates the calls of umpires accurately I am guessing.
|
|
|
Post by jmei on Jul 20, 2014 11:37:41 GMT -5
Ross' career marks (i.e., his pre-Boston career) suggests that his elite framing skill is mostly independent of the pitchers who've been throwing to him.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 22, 2014 8:41:23 GMT -5
As do I but I have a question. Is there any way to determine what portion is on the catcher and what portion is on the pitcher ? It occurs to me that the old saying that a pitcher can get strikes called if he's off the plate but is consistently there. We used to hear that Maddox, for example, could get away with throwing off the plate because he was so consistently good at throwing the ball where he wanted. Using Ross' steller stats as an example, how much is influenced by the fact that he's Lester's caddy ? That seems like an impossible feat on an extremely subjective matter that there is no metric used to define. Pitchers that get calls when they have command of the strike zone/an ace and pitchers that don't because of youth/lack of command are made by the human element of the game. There is no metric that evaluates the calls of umpires accurately I am guessing. BP has a database with the pitch/fx data for every pitch thrown, with the batter handedness, count, and the identity of the catcher, pitcher, and umpire. (I'm pretty sure that the data is Brooks Baseball's park-adjusted data.) They have also derived, for each location, the average percentage of called strikes given batter handedness (and, I'm pretty sure, count). This is the only part of their methodology with wiggle room -- deciding on the size of the location bucket (e.g., 1" squares versus baseball-sized squares) and the smoothing algorithm between buckets. Given that, it's very straightforward (with a stat technique called ANOVA) to isolate the catcher while controlling for the pitcher and umpire. You have to assume that pitchers who have only thrown to one catcher (or to one umpire) in their career have no pitcher effect, and that, I think, is a little tricky. Now, for today's update -- I've started normalizing for framing chances per game, which reveals last night's effort to be his second best yet. And I've used that to rate each game (in terms of WAR over 120 games if they were all framed that well). I'm going to ignore the fact that his last three games have been much better than his first two, because no one has ever framed pitches, over the course of a whole season, as well as Vazquez has over his last three starts.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,912
Member is Online
|
Post by ericmvan on Jul 22, 2014 8:50:46 GMT -5
Another question, how good of a bat would we need to see out of Vazquez in order to consider him an elite catcher in this league? I know that it helps a lot that the position is weak throughout the league, but everything I've read paints Vazquez as one of the best defensive backstops in the league with ease. So what would he realistically have to do at the plate to make us say, wow this kid is one of the 5 best catchers in the Majors (obviously that may not happen, this is more of a what if question)? Right now, it looks like anything above replacement level hitting. He's been a 4.5 WAR framer and they've yet to attempt a SB. That's a lot of defense. "Right now" of course is the catch. We have no idea if his performance so far is the equivalent of a really good hitter starting his career by hitting .550 / .650 / 1.000 over his first 5 games, or .350 / .450 / .600, or whatever. IOW, we don't know if he'll keep this up or come back to earth, and if so, how much. All we know is that he's been tremendous so far. (If we had game-by-game logs for other catchers, we'd have an idea how much game-to-game variance there is in pitch-framing, which would help answer the question. The less variance there is, the less air we have to take out of his current numbers) Given the subjective assessments of his framing, it's safe to say he's really good at it. Good enough, I would say, that if Swihart is ordinary at it, there's a real question as to who will be the more valuable.
|
|
|
Post by mgoetze on Jul 22, 2014 10:57:09 GMT -5
Hurry up and sign this kid to a 9-year contract before someone tells him what he's worth.
|
|
|
Post by ethanbein on Jul 22, 2014 11:45:03 GMT -5
Per StatCorner, Vazquez has now been the best pitch framer this year on a per-game basis of any catcher who has caught at least 4 pitches. He already has stolen more strikes this year than Ryan Hanigan who the Rays target for his framing abilities.
|
|
|