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.
The Reliever Reinforcement Effect Hypothesis
danr
Veteran
Posts: 1,871 
|
Post by danr on Apr 4, 2016 9:18:00 GMT -5
What no one has mentioned is that starting pitchers generally no longer are expected to go more than 6 innings. Last season I looked at the average number of innings pitched by starting pitchers and it was just shy of six. The pitch count limits probably are the main reason for this. Also, statistics show that batters hit much better after seeing a pitcher two or three times in a game.
It seems obvious that a team intent on winning wants to have dominant pitchers in almost every inning. That means the bullpen should have three or four dominant pitchers, maybe even five for a team to win a championship percentage of its games.
Thus, what Eric described, along with the increasing number of innings pitched by RPs, are racheting up the value of reliably dominant RPs.
|
|
|
|
Post by mgoetze on Apr 4, 2016 9:27:06 GMT -5
What no one has mentioned is that starting pitchers generally no longer are expected to go more than 6 innings. Yes, mentioning things everyone already knows does not advance the discussion, generally speaking. This seems like a complete non-sequitur to me. If I'm leading 4-0 after 6 innings, the other team can have all the dominant relievers it wants. It goes without saying that, all other things being equal, a good bullpen is preferable to a bad bullpen. But you have not provided an argument that, given limited resources, you should invest them here rather than in other areas. |
|
|
|
Post by brianthetaoist on Apr 4, 2016 9:52:05 GMT -5
Interesting post ... there's really two theories here. First, that hitters chase more pitches in the 9th because of the pressure of an imminent loss, leading to greater success as a closer for someone like Koji with an elite chase pitch. Second, that the specter of an elite closer instills such dread it creates a similar pressure in earlier innings, causing a similar dynamic in the hitters.
I'm somewhat sympathetic to the first theory, but I'm highly doubtful of the second. I'm pretty sure you could never prove it anyway, but, just on a logical level, I don't think the tangible, immediate pressure of the game ending is at all analogous to the indirect, intellectualized pressure of "Mariano Rivera is coming in next inning." If lineup protection is mostly mythological, why would this less-immediate and wholly implicit version of it carry significant weight?
|
|
danr
Veteran
Posts: 1,871
|
Post by danr on Apr 4, 2016 10:20:41 GMT -5
This seems like a complete non-sequitur to me. If I'm leading 4-0 after 6 innings, the other team can have all the dominant relievers it wants. It goes without saying that, all other things being equal, a good bullpen is preferable to a bad bullpen. But you have not provided an argument that, given limited resources, you should invest them here rather than in other areas. I think you are being a little snarky, which definitely doesn't advance any argument. I didn't argue that a team should invest resources in RPs at the expense of other needs. A first class bullpen isn't of prime importance to a non-contending team that has many other needs. And whether a contending team needs a stronger bullpen really depends on where its strengths and weaknesses are. It may be that a contender, like KC, because it already has strong starters and RPs, could better spend its resources improving its hitting. But many contenders do not have dominant relief pitching. Many realize that because of various reasons - including what Eric points out and the fact that starters are pitching fewer innings - having dominant relief pitching is vital to success. If they already are contenders with the team they have, then they can afford to go big for dominant RPs, of whom there are not many. The laws of supply and demand kick in and the cost of good RPs goes up. It was obvious last year that the Sox needed improvement in both starters and relievers. Price is a significant improvement in the starting pitching but I am not sure he is enough. The Sox did significantly improve the bullpen and if Smith is OK it will be a dominant one. |
|
|
|
Post by mgoetze on Apr 4, 2016 12:14:34 GMT -5
Many realize that because of various reasons - including what Eric points out and the fact that starters are pitching fewer innings - having dominant relief pitching is vital to success. And us few idiots who don't "realize" this "fact" are so hopeless that you're not going to bother explaining those reasons to us. Got it. |
|
|
|
Post by jmei on Apr 4, 2016 12:15:27 GMT -5
Many realize that because of various reasons - including what Eric points out and the fact that starters are pitching fewer innings - having dominant relief pitching is vital to success. And us few idiots who don't "realize" this "fact" are so hopeless that you're not going to bother explaining those reasons to us. Got it. Please tone down the hostility. You have been warned about this before. Thanks. |
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Apr 4, 2016 12:50:26 GMT -5
Interesting post ... there's really two theories here. First, that hitters chase more pitches in the 9th because of the pressure of an imminent loss, leading to greater success as a closer for someone like Koji with an elite chase pitch. Second, that the specter of an elite closer instills such dread it creates a similar pressure in earlier innings, causing a similar dynamic in the hitters. I'm somewhat sympathetic to the first theory, but I'm highly doubtful of the second. I'm pretty sure you could never prove it anyway, but, just on a logical level, I don't think the tangible, immediate pressure of the game ending is at all analogous to the indirect, intellectualized pressure of "Mariano Rivera is coming in next inning." If lineup protection is mostly mythological, why would this less-immediate and wholly implicit version of it carry significant weight? I think there's an argument that the effect only works for guys who are viewed as unhittable, elite of the elite. I know damn well that as a fan I used to think that the 8th inning when Mo was warming up was like the 9th anywhere else. It would be very easy to measure the effect that any closer, in a given season, had on the chase rates of his set-up guys, provided they pitched enough innings without him up next. Failing that, you still have cross-season comparisons, whose only drawback would be that you'd need a larger effect size to be sure it was real. Given the success the Royals had, and the fact that Farrell not only had the concept that 9th inning pressure enhanced chase rates, but explained it in public, I would bet that these numbers were run by a majority of analytically-oriented teams over the last couple of years. |
|
|
|
Post by jimed14 on Apr 4, 2016 14:02:40 GMT -5
So, do you think that Koji will alter his pitch mix now that he's the setup guy or will this be a bit of a test of this theory?
|
|
|
|
Post by umassgrad2005 on Apr 4, 2016 14:07:22 GMT -5
Why don't you just look at all closers for last 3 years and what the stats were for pitching in save situation and non save situation? It's not an end all, but it would give you a good idea if what you claim has any merit. If your right you should see a massive jump in strikeouts, and massive decrease in walks, hits and era when pitching for a save compared to non save situation.
|
|
|
|
Post by jimed14 on Apr 4, 2016 14:23:54 GMT -5
Unless I'm missing something, can't you just compare O-Swing percentages to see if players chase more pitches in the 9th?
I just don't see evidence that players chased Koji's pitches outside the zone more as a closer than in other seasons. His highest O-swing% season on his splitter was in 2012 when he wasn't a closer.
Obviously that's just one pitcher in a extremely small sample, but it was the example used to start the thread.
|
|
|
|
Post by jmei on Apr 4, 2016 14:36:23 GMT -5
More generally, I think the inability of analysts to find evidence that players play better or worse in clutch situations generally cuts against this theory. For instance, I'm not aware of any analysis which shows that starting pitchers can "pitch to the score" (in other words, that their performance varies depending on the leverage of the situation) or that closers pitch better or worse in save situations than they do in non-save situations, and there is plenty of analysis which cuts the other way. Similar analysis has been performed on hitters, with similar results. Considering the above, unless you can find real "smoking gun"-type data, I'm inclined to think that this is an interesting theory but no more than that. This is fundamentally asymmetrical. No one has really studied un-clutch performance, except as an adjunct to clutch, in very broad studies. Indeed, it makes little psychological sense that the average player could be better in high-leverage situations than in ordinary ones -- you immediately run into the "why isn't he that good all of the time" question. The extra adrenaline tends to make the nervous system noisier. It's far more credible that there are certain types of high-leverage situations where a subset of players perform sub-optimally, and largely because of a change in approach rather than any "clutch" or "choke" factor.. Elsewhere, I posted evidence that Pedroia is too aggressive in high-leverage situations when anyone other than Ortiz is the next hitter. The splits in hitting rates were small but collectively significant, but the split in WPA relative to OPS was massive, suggesting that all of the difference happened when the game was on the line. And that fits with what I remember about his performance perfectly. That's the sort of effect I'm hypothesizing here -- one too narrow to be discovered without a lot of work. My guess is that it's very clear if you look at every pitch with pitch/fx, and quite possibly you need hit/fx data as well. I don't have either databse, but MLB teams do, and they seem to be acting as if this hypothesis is true. You're not making any sort of claim about a subset of players, though. As far as I can tell, your claim is that a chain of elite relievers induces in all hitters a kind of anxiety that causes such hitters to perform worse than usual. That's an incredibly broad claim that there has been essentially no statistical support for. The difference between clutch and un-clutch is purely semantic-- clutch could just be the absence of un-clutch-- and even if it wasn't, there's actually a fair amount of analysis of un-clutch (for instance, of players like Alex Rodriguez who purportedly perform worse in the playoffs than in the regular season) that has found no durable un-clutch/choker trait. More importantly, if your only real evidence for this theory is that teams seem to be acquiring multiple elite relievers, there are plenty of alternative explanations for that trend. For instance, maybe the overemphasis of the SABR orthodoxy that relievers are overrated has made elite relievers a market inefficiency, maybe changing aging curves have done the same thing (by making starting pitchers/position players a more inefficient way to spend), maybe teams are recognizing the added value of elite relievers in the playoffs (where they can pitch a much, much higher proportion of a team's innings), maybe there are more elite relievers to go around and so teams have more of them than we'd expect, maybe it's even just random chance! |
|
|
|
Post by brianthetaoist on Apr 4, 2016 20:21:30 GMT -5
Interesting post ... there's really two theories here. First, that hitters chase more pitches in the 9th because of the pressure of an imminent loss, leading to greater success as a closer for someone like Koji with an elite chase pitch. Second, that the specter of an elite closer instills such dread it creates a similar pressure in earlier innings, causing a similar dynamic in the hitters. I'm somewhat sympathetic to the first theory, but I'm highly doubtful of the second. I'm pretty sure you could never prove it anyway, but, just on a logical level, I don't think the tangible, immediate pressure of the game ending is at all analogous to the indirect, intellectualized pressure of "Mariano Rivera is coming in next inning." If lineup protection is mostly mythological, why would this less-immediate and wholly implicit version of it carry significant weight? I think there's an argument that the effect only works for guys who are viewed as unhittable, elite of the elite. I know damn well that as a fan I used to think that the 8th inning when Mo was warming up was like the 9th anywhere else. It would be very easy to measure the effect that any closer, in a given season, had on the chase rates of his set-up guys, provided they pitched enough innings without him up next. Failing that, you still have cross-season comparisons, whose only drawback would be that you'd need a larger effect size to be sure it was real. Given the success the Royals had, and the fact that Farrell not only had the concept that 9th inning pressure enhanced chase rates, but explained it in public, I would bet that these numbers were run by a majority of analytically-oriented teams over the last couple of years. Yeah, that's why I used Rivera for the example (and you used Kimbrel, Chapman, etc). Again, it's a cool thought, but it's only remotely possible if you (the hitter) believe deeply that trying to go against the closer is a lost cause ... and it still doesn't feel like the same psychological dynamic to me as the end of the game. It's true but more theoretical/intellectual than direct threat, global warming vs tiger in the front yard. In my various career paths, I've worked a lot in the psychological interactions between message and action, and you really need a direct, almost visceral stimulus to go against the kind of ingrained responses a hitter has to certain pitch/location events. And I don't think it'd be easy at all to measure. Hell, relief pitchers are hard enough just in general, but finding enough situations to control against seems like it'd be impossible. The variables are almost perfectly designed *against* a meaningful control ... because if these set up guys are pitching in a similar situation (ahead by a relatively small amount in the 8th inning), by definition the closer's going to be following them the heavy preponderance of the instances. And if they are pitching in a blowout or in a losing situation, well that's a huge confounding variable right there. I guess you could posit a ramping up of pressure (hitters naturally more likely to chase in the 8th than in the 4th, for instance, with a big step up in the 9th) made marginally steeper by the presence of an elite closer, and I could maybe buy it as a natural response to heightened pressure. And it could be possible to test that first, global hypothesis. |
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Apr 4, 2016 21:46:25 GMT -5
This is fundamentally asymmetrical. No one has really studied un-clutch performance, except as an adjunct to clutch, in very broad studies. Indeed, it makes little psychological sense that the average player could be better in high-leverage situations than in ordinary ones -- you immediately run into the "why isn't he that good all of the time" question. The extra adrenaline tends to make the nervous system noisier. It's far more credible that there are certain types of high-leverage situations where a subset of players perform sub-optimally, and largely because of a change in approach rather than any "clutch" or "choke" factor.. Elsewhere, I posted evidence that Pedroia is too aggressive in high-leverage situations when anyone other than Ortiz is the next hitter. The splits in hitting rates were small but collectively significant, but the split in WPA relative to OPS was massive, suggesting that all of the difference happened when the game was on the line. And that fits with what I remember about his performance perfectly. That's the sort of effect I'm hypothesizing here -- one too narrow to be discovered without a lot of work. My guess is that it's very clear if you look at every pitch with pitch/fx, and quite possibly you need hit/fx data as well. I don't have either databse, but MLB teams do, and they seem to be acting as if this hypothesis is true. You're not making any sort of claim about a subset of players, though. As far as I can tell, your claim is that a chain of elite relievers induces in all hitters a kind of anxiety that causes such hitters to perform worse than usual. That's an incredibly broad claim that there has been essentially no statistical support for. The difference between clutch and un-clutch is purely semantic-- clutch could just be the absence of un-clutch-- and even if it wasn't, there's actually a fair amount of analysis of un-clutch (for instance, of players like Alex Rodriguez who purportedly perform worse in the playoffs than in the regular season) that has found no durable un-clutch/choker trait. More importantly, if your only real evidence for this theory is that teams seem to be acquiring multiple elite relievers, there are plenty of alternative explanations for that trend. For instance, maybe the overemphasis of the SABR orthodoxy that relievers are overrated has made elite relievers a market inefficiency, maybe changing aging curves have done the same thing (by making starting pitchers/position players a more inefficient way to spend), maybe teams are recognizing the added value of elite relievers in the playoffs (where they can pitch a much, much higher proportion of a team's innings), maybe there are more elite relievers to go around and so teams have more of them than we'd expect, maybe it's even just random chance! Sorry if I didn't make it clear, but I thought I'd said at least once and implied originally that it's only pitchers who thrive on pitches chased outside the zone who benefit from the effect. In terms of un-clutch, I've already mentioned one data point -- Bobby Valentine's glorious 2012 Red Sox. I just dug up the data file. In the first 4 games of the season, they went 4/5, 2 2B, 3B, 2 SF, 1 SO in 7 PA in the 9th, opponent save situation. (And 5/9, 2 SO in non-save situations.) From then until June 15th, they went 10/63, 2 2B, 2 HR, 9 BB (1 IBB), 17 SO. That's .159 / .264 / .286. And then they started pressing. Subsequently they went 13/100, 1 2B, 3 HR, 2 BB, 30 SO. 2 BB in 102 PA over 3 1/2 months. That's .130 / .147 / .230. They had 20 SO and 0 BB from June 22 to August 28th, then rebounded with 10 SO and 2 BB the rest of the way. Right now, b-ref's team splits are down, but as soon as they're up I'll run the save vs. non-save situation comparison. (Yeah, I know what you're thinking ... I excluded extra inning opponent save situations. I figured the split on that sample size wouldn't tell us much. But they were even worse in terms of OPS : 3/24, 1 BB, 6 SO, .125 / .160 / .125. And, oh yeah, they lost all 7 of those extra-inning home games, getting shut out over 20 innings, while hitting .184 / .245 / .204 in 53 extra-inning PA before the other team took the lead. I'm looking forward to your explanation as to why this was random and/or why it's the only time something like this ever happened to a group of players over a period of time.) |
|
|
|
Post by jmei on Apr 4, 2016 22:15:08 GMT -5
I'm looking forward to your explanation as to why this was random and/or why it's the only time something like this ever happened to a group of players over a period of time.) But seriously, you're looking at a bad offensive team's performance against the best relievers in the league in a small sample and then slapping on some arbitrary endpoints (why separate the first four games of the season? why draw a line on June 15?). Flip enough coins and you'll get a long string of tails in a row and all that. Even if that performance was non-random, I'm not sure what one team's second-half offensive performance in save situations in a lost season has to do with the benefits of chaining elite relievers. ADD: this isn't the first time, nor will it be the last time we have this exact same argument. I'm fine dropping it if you are. |
|
|
|
Post by telson13 on Apr 4, 2016 22:51:58 GMT -5
Interesting post ... there's really two theories here. First, that hitters chase more pitches in the 9th because of the pressure of an imminent loss, leading to greater success as a closer for someone like Koji with an elite chase pitch. Second, that the specter of an elite closer instills such dread it creates a similar pressure in earlier innings, causing a similar dynamic in the hitters. I'm somewhat sympathetic to the first theory, but I'm highly doubtful of the second. I'm pretty sure you could never prove it anyway, but, just on a logical level, I don't think the tangible, immediate pressure of the game ending is at all analogous to the indirect, intellectualized pressure of "Mariano Rivera is coming in next inning." If lineup protection is mostly mythological, why would this less-immediate and wholly implicit version of it carry significant weight? I think there's an argument that the effect only works for guys who are viewed as unhittable, elite of the elite. I know damn well that as a fan I used to think that the 8th inning when Mo was warming up was like the 9th anywhere else. It would be very easy to measure the effect that any closer, in a given season, had on the chase rates of his set-up guys, provided they pitched enough innings without him up next. Failing that, you still have cross-season comparisons, whose only drawback would be that you'd need a larger effect size to be sure it was real. Given the success the Royals had, and the fact that Farrell not only had the concept that 9th inning pressure enhanced chase rates, but explained it in public, I would bet that these numbers were run by a majority of analytically-oriented teams over the last couple of years. Errr...except there's the issue that the times the closer doesn't come after the setup man is often predicated on the setup man having an off day, which would skew the data with a severe selection bias. Maybe look at chase rates for pitchers when ahead vs. behind...and related to "closeness" of the game (beyond even just save situations). Kind of a Schrödinger's cat issue, I think. I can imagine there's a way around it, but I'm not sure how to tease it out analytically. I suppose cross-season works, but that seems like it would require a ton of data, from change-of-closer adjacent seasons (or within season). |
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Apr 4, 2016 23:24:55 GMT -5
I think there's an argument that the effect only works for guys who are viewed as unhittable, elite of the elite. I know damn well that as a fan I used to think that the 8th inning when Mo was warming up was like the 9th anywhere else. It would be very easy to measure the effect that any closer, in a given season, had on the chase rates of his set-up guys, provided they pitched enough innings without him up next. Failing that, you still have cross-season comparisons, whose only drawback would be that you'd need a larger effect size to be sure it was real. Given the success the Royals had, and the fact that Farrell not only had the concept that 9th inning pressure enhanced chase rates, but explained it in public, I would bet that these numbers were run by a majority of analytically-oriented teams over the last couple of years. Errr...except there's the issue that the times the closer doesn't come after the setup man is often predicated on the setup man having an off day, which would skew the data with a severe selection bias. Maybe look at chase rates for pitchers when ahead vs. behind...and related to "closeness" of the game (beyond even just save situations). Kind of a Schrödinger's cat issue, I think. I can imagine there's a way around it, but I'm not sure how to tease it out analytically. I suppose cross-season works, but that seems like it would require a ton of data, from change-of-closer adjacent seasons (or within season). Great point about the selection bias. You'd have to do it cross-season, which indeed adds a lot of noise. But there's an insane amount of data. You could build a model for every hitter's propensity to swing at a given pitch, based on pitch type, count, location, velocity, movement, and all of those variables for the preceding pitch. All of the sample sizes will be small, but there are lots of sophisticated techniques to smooth the data and hence deal with that (the most obvious being aggregating everyone's data to derive some universal tendenies as a baseline). So now you're controlling for every variable but inning and score and following pitcher, if any. And of course you can build the model for pitchers and look at the hitters. (In fact, you'd iterate between the two models so that they both ended up controlling for who they faced.) If you think about it, a model that predicts results for every hitter based on all the aforementioned variables would be an insanely valuable analytic tool. It's kind of an obvious Holy Grail of pitch/fx data analysis, and any team that decided it wasn't worth the effort because of the noise of the dependent variable (game result) would be all over it as soon as hit/fx data arrived. And if you had that model, you could answer this question pretty easily. And a million questions like it. Which is why you would try as hard as possible to build such a model. |
|
|
|
Post by brianthetaoist on Apr 6, 2016 9:09:53 GMT -5
More importantly, if your only real evidence for this theory is that teams seem to be acquiring multiple elite relievers, there are plenty of alternative explanations for that trend. For instance, maybe the overemphasis of the SABR orthodoxy that relievers are overrated has made elite relievers a market inefficiency, maybe changing aging curves have done the same thing (by making starting pitchers/position players a more inefficient way to spend), maybe teams are recognizing the added value of elite relievers in the playoffs (where they can pitch a much, much higher proportion of a team's innings), maybe there are more elite relievers to go around and so teams have more of them than we'd expect, maybe it's even just random chance! Huh, I didn't read much baseball over the offseason, but was this point discussed much? It seems like a good theory ... since the expanded playoff roster is lowering the value of the regular season relative to non-WC days and raising the value of small postseason advantages, it seems like this is one place where that shift would show up. |
|
|
