Bundy and the third time through

[chpettit19]

50 minutes ago, TwinsDr2021 said:

Got it everybody doing this means one playoff game three years ago, going forward I will be sure to remember that everybody doing somethings could mean a unique scenario three seasons ago.

And yes it is lazy to make a broad statement that everybody is doing something, but then not being able to back it up or refusing to back it up. And again I am talking about Bundy's latest game.

If this conversation was about Winder being pulled after 5 and 77 pitches and somebody complained it was too soon and you or me responded with too bad that is way the everybody is doing it, you know what that would be OK and not lazy, why? because we could point out numerous games every damn day by just about every team that it was happening.

I have looked at hundreds of box scores in the last 24 hours and I found 2 similiar games to this one. Bundy's 6 inning 60 game (which he did give up the lead and his only run in the 6th, so maybe that is what the Twins were looking at, but that is vastly different than everybody is doing it), Archer's 5 inning 57 pitch game which the twins lost when Pagen blew a save to Detroit. Maybe I missed some, but if everybody is doing it, it shouldn't take hundreds of box scores to find 3?

The disconnect here is that you think the other teams are taking guys out after 5 and 80 pitches because of the 80 pitches. I'm saying that's what you're wrong about. Do you read the idea of that ESPN article as a pitch count thing? Because that's not what it's about. Take the Wainwright quote for example. "If you're OK with throwing five and you came out of the game and you thought you did your job, then the standard is set." 5-6 is the standard because that's the area where you're getting through the lineup twice. That's what the teams are targeting. 

They then also want relievers to come into clean innings instead of trying to clean up messes since that's the best way to reduce runs against. Guys are throwing 80 pitches in 5 or 6 and being taken out because that's their new role. It's trained into them in the minors. They're not being taken out because of the 80 pitches, they're being taken out because they've done their job for the day. It's about "out getters" now. Starters are bulk out getters and the team wants 15-18+ outs out of them. As they approach the 15 out mark managers and pitching coaches start looking at a crazy amount of factors and make a decision on when to pull them. Sometimes it's after 15 outs. Sometimes it's after 21. Sometimes it's in between. If they start getting up to 100 pitches in 4 then pitch count becomes a factor. But the 80-90 pitches isn't what's triggering them being removed from the game.

Now there are guys that are on pitch counts due to arm health concerns. Like Archer. And teams typically try to back those guys up with another 9-12 out guy. The Twins have gotten away from that after Winder was in that role to start the year. I don't get why, but it's the choice they made. Relievers aren't all just 3 out guys anymore. Some are 6 out guys. They all have different roles and they adapt to them. If you know you're only expected to get 18 outs that day you're not conserving pitches like guys used to do to be able to get 21-24 outs.

[chpettit19]

10 minutes ago, big dog said:

So you want people to find examples where a pitcher has through 58 pitches through 5 innings and was taken out? How about if instead of expecting other people to prove your point, you slog through a whole bunch of game logs and prove it yourself?

[edit- ninja'd by chpettit19 but at least I was succinct ]

I just have so much to say!

[jorgenswest]

1 hour ago, Riverbrian said:

I think the original point is a very interesting consideration and probably has a high level of correctness and I'll admit it was something that I hadn't thought of. 

However, that top of the order that skews the data will be the same top of the order that the starting pitcher must face after going through the order twice. So while the data may be skewed, those top of the order guys who skew the data are still holding a bat in their hands ready to skew the data some more. 

 

 

They will also be the same guys the reliever faces. Over time any reliever coming in to face Arraez, Buxton and Correa is going to a have less success than they would have against Urshela, Jeffers and Celestino.

I think broadcasters and writers incorrectly suggest that the main reason starters numbers are so much worse the third time through is because hitters get to see them a third time. I think the reality of the data is that it includes a whole bunch of Arraez level batters and hardly any of the Jeffers. 

[wsnydes]

10 minutes ago, jorgenswest said:

They will also be the same guys the reliever faces. Over time any reliever coming in to face Arraez, Buxton and Correa is going to a have less success than they would have against Urshela, Jeffers and Celestino.

I think broadcasters and writers incorrectly suggest that the main reason starters numbers are so much worse the third time through is because hitters get to see them a third time. I think the reality of the data is that it includes a whole bunch of Arraez level batters and hardly any of the Jeffers. 

Which is why we're also seeing more power in the top third of the lineup as well.  It's not table setters in the 1/2 slot and then power, it's power at 1-5.  Your best all around hitter may still be in the 3 hole, but the power has shifted towards the top of the lineup rather than the middle.

[TwinsDr2021]

52 minutes ago, wsnydes said:

Between @chpettit19and the article @Riverbrianlinked last week here (linked once again), plenty of back up has been provided to you against your point.  You keep dwelling on pitch count when pitch count isn't even the determining factor of when a guy is pulled.  It is a factor, but it's not THE factor anymore.  

If anyone is lazy here, it's me for simply referring you to someone else's post from last week linking to an article written by someone else!

My argument was never about pitcher's pitching less innings, less pitches, favorable matchups, or any of the other things, my argument was simplify do teams take a starting pitcher out in 5 innings and around 58 , or really any other inning where a pitcher is averaging 10/11 pitches an inning?

You say pitch count isn't the major factor, but with all the examples I provided it actually does seems to be a big factor? No?

[USAFChief]

24 minutes ago, jorgenswest said:

They will also be the same guys the reliever faces. Over time any reliever coming in to face Arraez, Buxton and Correa is going to a have less success than they would have against Urshela, Jeffers and Celestino.

I think broadcasters and writers incorrectly suggest that the main reason starters numbers are so much worse the third time through is because hitters get to see them a third time. I think the reality of the data is that it includes a whole bunch of Arraez level batters and hardly any of the Jeffers. 

I don't think that's true, but in any case it's irrelevant. The question is, will a fresh reliever have better success against the top 5 in a lineup than the third time from the starter?

As RB noted, the runs given up count the same no matter the actual reason. I don't care if Bundy's ERA might be slightly lower if he got to face 27 hitters instead of 22. I care about not giving up those runs in the first place.

 

[TwinsDr2021]

56 minutes ago, big dog said:

So you want people to find examples where a pitcher has through 58 pitches through 5 innings and was taken out? How about if instead of expecting other people to prove your point, you slog through a whole bunch of game logs and prove it yourself?

[edit- ninja'd by chpettit19 but at least I was succinct ]

That is what I did Big Dog, I went though hundreds of box scorse, looking for something close to that, every start by the Yanks, Braves, Rays, Mets and daily box scores for the last few weeks, and I couldn't find a pitcher taken out in the 5th averaging less (besides the ones I mentioned above) than 12 pitches (60 pitchers) or 6 innings and less than 7 pitches. unless that pitcher was coming back from a injury, a rain delay, or the pitcher was terrible. So I guess your point was?

[wsnydes]

5 minutes ago, TwinsDr2021 said:

You say pitch count isn't the major factor, but with all the examples I provided it actually does seems to be a big factor? No?

The article I linked, the ESPN article @chpettit19referenced and everything he's stated says that it is not.  I'll even grant you that the Bundy case is an extreme one, but they're not looking at pitch count.  And that's the point we're trying to make to you.  And we've demonstrated that it's not just the Twins doing this.

[TwinsDr2021]

52 minutes ago, chpettit19 said:

The disconnect here is that you think the other teams are taking guys out after 5 and 80 pitches because of the 80 pitches. I'm saying that's what you're wrong about. Do you read the idea of that ESPN article as a pitch count thing? Because that's not what it's about. Take the Wainwright quote for example. "If you're OK with throwing five and you came out of the game and you thought you did your job, then the standard is set." 5-6 is the standard because that's the area where you're getting through the lineup twice. That's what the teams are targeting. 

 

That makes sense and I don't disagree with that, but if what you are saying is true, shouldn't we see more guys being taken out in the 5th with less than say 65 pitches? Or 6 innings with lets say 75 pitches?

But that isn't what the box scores are showing, they are showing if a pitcher is efficient they are usually given more outs/innings and if they aren't horribly efficient they are taken out (Example Winder last night 5/77, that seems very consistent across the league).

That is my whole point, I feel like you are trying to say something I already agree with and have said absolutely true.

[TwinsDr2021]

2 minutes ago, wsnydes said:

The article I linked, the ESPN article @chpettit19referenced and everything he's stated says that it is not.  I'll even grant you that the Bundy case is an extreme one, but they're not looking at pitch count.  And that's the point we're trying to make to you.  And we've demonstrated that it's not just the Twins doing this.

I agree with all of that. except nobody has demonstrated that what the Twins did with Bundy any other team is doing or really ever does at all, there is no argument from me on anything that is being said, stated, proved by fact except that. I believe I can say based on looking at hundreds of box scores that 9/10 times other teams would have let Bundy start the 6th (right or wrong decision is not what I am arguing .) FYI I found many examples where a pitcher was brought out for the 6th with about the same amount of pitches, and sometimes it worked and sometimes it didn't, and there many that the pitcher stayed good and finished 7 or 8 plus innings.

And that is what I am basing my argument on, if you state that "everybody is doing" that specific decision it should be reality easy to confirm that conclusion, all of the facts that numbers, even the one example of the Rays and everything else that has been given is all great information and wonderful work on digging it up, and I truly do appreciate all that work didn't IMO. 

[wsnydes]

3 minutes ago, TwinsDr2021 said:

I agree with all of that. except nobody has demonstrated that what the Twins did with Bundy any other team is doing or really ever does at all, there is no argument from me on anything that is being said, stated, proved by fact except that. I believe I can say based on looking at hundreds of box scores that 9/10 times other teams would have let Bundy start the 6th (right or wrong decision is not what I am arguing .) FYI I found many examples where a pitcher was brought out for the 6th with about the same amount of pitches, and sometimes it worked and sometimes it didn't, and there many that the pitcher stayed good and finished 7 or 8 plus innings.

And that is what I am basing my argument on, if you state that "everybody is doing" that specific decision it should be reality easy to confirm that conclusion, all of the facts that numbers, even the one example of the Rays and everything else that has been given is all great information and wonderful work on digging it up, and I truly do appreciate all that work didn't IMO. 

You're missing the point.  It's not about the pitch count.  Even your own findings back that up.  I've even given you the Bundy case as an extreme example.  And that's because the pitch count wasn't relevant to the decision.  You may not find an better example of the irrelevance of the pitch count than this one Bundy start.  But every team is trending away from pitch counts to make the decision, which is what our argument has been.  

FYI I found many examples where a pitcher was brought out for the 6th with about the same amount of pitches, and sometimes it worked and sometimes it didn't, and there many that the pitcher stayed good and finished 7 or 8 plus innings.

Based on this sentence alone, I would imagine that if you kept digging into this, that the number of times that it doesn't work out is higher than the times that it did.  And that's why Bundy didn't go out.  Especially with a backend starter like Bundy, the times where it doesn't work out is going to be greater than the times that it does.  Teams are simply leveraging the odds, not paying attention to the pitch count.

[Riverbrian]

40 minutes ago, jorgenswest said:

They will also be the same guys the reliever faces. Over time any reliever coming in to face Arraez, Buxton and Correa is going to a have less success than they would have against Urshela, Jeffers and Celestino.

I think broadcasters and writers incorrectly suggest that the main reason starters numbers are so much worse the third time through is because hitters get to see them a third time. I think the reality of the data is that it includes a whole bunch of Arraez level batters and hardly any of the Jeffers. 

2nd Paragraph... I am in agreement with you. What you are saying is something that would be easily overlooked by broadcasters, writers and fans and it probably creates a half baked narrative presented as fully cooked and swallowed whole by many. There are a lot of things like that in the world of baseball. 

1st Paragraph... In the case of the Twins,

I haven't logged the specific examples, I have nothing conclusive but I do have a general impression from watching almost every night that Rocco tends to utilize Duran and the perceived better bullpen arms available when facing the top the order and lesser available bullpen arms when facing the bottom of the order. So, I think you can take what you are talking about and add it to the many things that Rocco has to consider when making the bullpen decisions that he does. 

On Monday... It was Jax replacing Bundy to face the top of the lineup. It was Duffey replacing Jax to face the bottom of the lineup. 

The only place I question Rocco on Monday was in the 9th when he turned to Pagan to face hitters 2, 3 and 4 with the game tied in the 9th but that is because I'm a little Pagan spooked. Clearly the Twins still consider Pagan to be high leverage. 

I would have thrown Duran in the 9th while making phone calls to other teams to bolster the future pen. ?    

 

 

[chpettit19]

15 minutes ago, wsnydes said:

You're missing the point.  It's not about the pitch count.  Even your own findings back that up.  I've even given you the Bundy case as an extreme example.  And that's because the pitch count wasn't relevant to the decision.  You may not find an better example of the irrelevance of the pitch count than this one Bundy start.  But every team is trending away from pitch counts to make the decision, which is what our argument has been.  

FYI I found many examples where a pitcher was brought out for the 6th with about the same amount of pitches, and sometimes it worked and sometimes it didn't, and there many that the pitcher stayed good and finished 7 or 8 plus innings.

Based on this sentence alone, I would imagine that if you kept digging into this, that the number of times that it doesn't work out is higher than the times that it did.  And that's why Bundy didn't go out.  Especially with a backend starter like Bundy, the times where it doesn't work out is going to be greater than the times that it does.  Teams are simply leveraging the odds, not paying attention to the pitch count.

The part I bolded about a backend starter like Bundy is a huge part of this. Comparing how backend starters are treated to how frontend starters are treated gives people bad info. If Max Scherzer is at 58 pitches after 5 the Twins would let him go longer. In fact when Ryan or Gray are at 58 after 5 they go back for the 6th. But they're better than Bundy and their "standard" (to steal the word Wainwright used) is higher than Bundy's. That's why this is such a difficult thing to discuss. There are just so many variables. Trying to make overarching statements about 1 stat in 1 start is a losing battle.

[bean5302]

There remains this TTO myth which is regurgitated as some sort of absolute fact.

The question is whether or not a middle reliever is better than your starter on the third time through the order, if you're really trying to figure out the odds of winning a game. If the starter is MLB average, it looks like this:
 

Median averages
First time through the order ERA = 2.90, xFIP = 3.61
Second time through the order ERA = 3.41, xFIP = 4.17
Third time through the order ERA = 4.44, xFIP = 4.08

Looks like that third time is really brutal until you look at the xFIP. The source of the TTO myth is home runs. So do pitchers give up more home runs because they're tiring or because batters have figured them out or because managers have adpoted the TTO approach with a quick hook and reach the cane out the moment a pitcher gives up a homer?

First time FB% = 34.9%, Hard Hit% = 30.0%
Second time FB% = 36.1%, Hard Hit% = 29.9%
Third time FB% = 37.1%, Hard Hit% = 30.9%

Basically, to me, there isn't a significant difference in hard hit rates or fly ball rates on the first, second and third time through the order. The xFIPs for pitchers I've reviewed (more than just the ones below) do not show a huge jump on the third time through the order. The quick hook factor seems to be at play as much as anything. A pitcher gives up a home run, Quick Hook decides the pitcher is done and yanks them before they have the chance to work through trouble or Quick Hook calls on a struggling reliever to come in and get through the trouble... and the middle reliever can't, thus an inflated ERA for the starter which the starter sorta didn't earn. That said, the manager is always wrong. If the manager pulls the pitcher early, the manager is wrong. If they leave the pitcher in and the pitcher gives up more runs, the manager was wrong. If they leave the pitcher to work through the issue and the pitcher succeeds, the manager was wrong because it was still risky. Either that or people forget there was a stressful spot to begin with.

In any case...

The median pitcher in the example above was Luis Severino this year
First time ERA = 2.90, xFIP = 3.40
Second time ERA = 2.10, xFIP = 3.46
Third time ERA = 6.67, xFIP = 2.99

Luis Severino career
First time ERA = 2.90, xFIP = x3.16
Second time ERA = 3.26, xFIP = 2.99
Third time ERA = 5.18, xFIP = 3.34

Dylan Bundy career
First time ERA = 4.08, xFIP = 4.19
Second time ERA = 4.96, xFIP = 4.62
Third time ERA = 5.81, xFIP = 4.75

Chris Archer career
First time ERA = 3.09, xFIP = 3.37
Second time ERA = 3.91, xFIP = 3.65
Third time ERA = 4.32, xFIP = 3.94

[jorgenswest]

1 hour ago, Riverbrian said:

2nd Paragraph... I am in agreement with you. What you are saying is something that would be easily overlooked by broadcasters, writers and fans and it probably creates a half baked narrative presented as fully cooked and swallowed whole by many. There are a lot of things like that in the world of baseball. 

1st Paragraph... In the case of the Twins,

I haven't logged the specific examples, I have nothing conclusive but I do have a general impression from watching almost every night that Rocco tends to utilize Duran and the perceived better bullpen arms available when facing the top the order and lesser available bullpen arms when facing the bottom of the order. So, I think you can take what you are talking about and add it to the many things that Rocco has to consider when making the bullpen decisions that he does. 

On Monday... It was Jax replacing Bundy to face the top of the lineup. It was Duffey replacing Jax to face the bottom of the lineup. 

The only place I question Rocco on Monday was in the 9th when he turned to Pagan to face hitters 2, 3 and 4 with the game tied in the 9th but that is because I'm a little Pagan spooked. Clearly the Twins still consider Pagan to be high leverage. 

I would have thrown Duran in the 9th while making phone calls to other teams to bolster the future pen. ?    

 

 

If this is real we should also see hitters batting better the third time through. I went through the 2021 AL Central and looked at the starting players for each team. I ended up with a group of 48. What I found isn't much different from what I found in a larger leaguewide study about 5 years ago. 

There were 17 of 48 hitters that had their highest OPS in the third time they saw the pitcher. That is about 1/3. This time they were most successful upon seeing the pitcher the second time and least successful the first time. I also found the 22 of the 48 hitters had the lowest OPS of the three seeing a pitcher the third time. Hitters were more likely to have their worst numbers when seeing a pitcher the third time. There is a skew here also. You probably only get to see a pitcher a third time if they are pitching well. 

This isn't argument that the Twins should have kept Bundy in or taken him out. I would have bet on Jax. 

I hear the narrative about the third time through routinely. The inference is that somehow a batter has a much better chance after seeing a pitcher a few times and hence the alarming splits from the pitchers. I hear it from analysts I respect like Aaron Gleeman who made this statement yesterday on KFAN. My argument is that the difference in the split is almost entirely based on the quality of hitters in that third time faced pitcher split. If it were about learning we would see it in hitters also. 

I would hope that this narrative would change. The Twins brought in Jax at the point because for any 1 or 2 innings he is simply a more effective pitcher than Bundy. They would want the better pitcher to face the top of the line up. I really don't believe that the top of the line up is more likely to to damage simply because it is the third time through.

[chpettit19]

1 hour ago, jorgenswest said:

If this is real we should also see hitters batting better the third time through. I went through the 2021 AL Central and looked at the starting players for each team. I ended up with a group of 48. What I found isn't much different from what I found in a larger leaguewide study about 5 years ago. 

There were 17 of 48 hitters that had their highest OPS in the third time they saw the pitcher. That is about 1/3. This time they were most successful upon seeing the pitcher the second time and least successful the first time. I also found the 22 of the 48 hitters had the lowest OPS of the three seeing a pitcher the third time. Hitters were more likely to have their worst numbers when seeing a pitcher the third time. There is a skew here also. You probably only get to see a pitcher a third time if they are pitching well. 

This isn't argument that the Twins should have kept Bundy in or taken him out. I would have bet on Jax. 

I hear the narrative about the third time through routinely. The inference is that somehow a batter has a much better chance after seeing a pitcher a few times and hence the alarming splits from the pitchers. I hear it from analysts I respect like Aaron Gleeman who made this statement yesterday on KFAN. My argument is that the difference in the split is almost entirely based on the quality of hitters in that third time faced pitcher split. If it were about learning we would see it in hitters also. 

I would hope that this narrative would change. The Twins brought in Jax at the point because for any 1 or 2 innings he is simply a more effective pitcher than Bundy. They would want the better pitcher to face the top of the line up. I really don't believe that the top of the line up is more likely to to damage simply because it is the third time through.

The problem with the narrative is that broadcasters and writers need to play to the general public, not the super fans that find themselves on sites like this. They also need to be somewhat concise to keep listeners/readers engaged. And those things lead to grand generalizations. "Starters" and "relievers" are what the general public baseball fans know. It's what they're used to. "Out getters" is a deeper conversation that is written and talked about by baseball nerds on baseball nerd sites for baseball nerd fans. There have been links provided to such content in this thread and others. The guys on Twins broadcasts are also not as informed as the baseball nerds on baseball sites, but the general fan assumes they are. If you sat with a current baseball FO exec or coach and listened to a game with them they'd tell you half a dozen times that what the broadcasters are saying isn't how it's done anymore. It's just the nature of the beast.

So the problem is that people want to discuss things as grand generalizations when the truth is the plans the teams execute are far, far, far more player specific. The "Bundy rules" aren't the same as the "Ryan rules" which aren't the same as the "Archer rules" which differ from the "Gray rules." And each other player's rules/roles change depending on game situation, pen arms available, and dozens of other factors. But people want general rules. So we discuss the third time through the order situation as a general rule for pitchers. And it's definitely a large part of the equation for a large part of the league's starters. But when a team has a huge lead the rules change in that game and Bundy ends up with 8 innings pitched. 

None of that even gets into using data from the last 5 years or so as teams have adjusted to the new role pitchers fill and how that data is far too little to go off. Teams started pulling their pitchers earlier because of historical data. Then the teams started putting Acuna and Buxton types in the leadoff spot to give them a better chance of seeing the low end starters a 3rd time or forcing the manager to go to the pen earlier than they'd like. Now you have Aaron Judge types hitting in the 2 hole instead of Nick Punto types. That brings the numbers back up in general terms of 3rd time through the order since you're getting the big boppers as soon as the lineup turns over. That's how it all works. And pitcher usage and lineup construction will continue to adapt and adjust and change until we all die. 

Third time through isn't a hard and fast cutoff line. There's far too many factors to discuss on boards like this. But it is 100% a real decision driving data point based on crazy amounts of data. Like millions and millions of lines worth of data. More data than the general broadcaster and writer can explain to the general baseball fan.

[jorgenswest]

48 minutes ago, chpettit19 said:

The problem with the narrative is that broadcasters and writers need to play to the general public, not the super fans that find themselves on sites like this. They also need to be somewhat concise to keep listeners/readers engaged. And those things lead to grand generalizations. "Starters" and "relievers" are what the general public baseball fans know. It's what they're used to. "Out getters" is a deeper conversation that is written and talked about by baseball nerds on baseball nerd sites for baseball nerd fans. There have been links provided to such content in this thread and others. The guys on Twins broadcasts are also not as informed as the baseball nerds on baseball sites, but the general fan assumes they are. If you sat with a current baseball FO exec or coach and listened to a game with them they'd tell you half a dozen times that what the broadcasters are saying isn't how it's done anymore. It's just the nature of the beast.

So the problem is that people want to discuss things as grand generalizations when the truth is the plans the teams execute are far, far, far more player specific. The "Bundy rules" aren't the same as the "Ryan rules" which aren't the same as the "Archer rules" which differ from the "Gray rules." And each other player's rules/roles change depending on game situation, pen arms available, and dozens of other factors. But people want general rules. So we discuss the third time through the order situation as a general rule for pitchers. And it's definitely a large part of the equation for a large part of the league's starters. But when a team has a huge lead the rules change in that game and Bundy ends up with 8 innings pitched. 

None of that even gets into using data from the last 5 years or so as teams have adjusted to the new role pitchers fill and how that data is far too little to go off. Teams started pulling their pitchers earlier because of historical data. Then the teams started putting Acuna and Buxton types in the leadoff spot to give them a better chance of seeing the low end starters a 3rd time or forcing the manager to go to the pen earlier than they'd like. Now you have Aaron Judge types hitting in the 2 hole instead of Nick Punto types. That brings the numbers back up in general terms of 3rd time through the order since you're getting the big boppers as soon as the lineup turns over. That's how it all works. And pitcher usage and lineup construction will continue to adapt and adjust and change until we all die. 

Third time through isn't a hard and fast cutoff line. There's far too many factors to discuss on boards like this. But it is 100% a real decision driving data point based on crazy amounts of data. Like millions and millions of lines worth of data. More data than the general broadcaster and writer can explain to the general baseball fan.

It really doesn’t feel like you have read my thoughts. This thread has taken many interesting and unintended directions about the Twins move to take out Bundy,

It isn’t about millions of data points. It isn’t about whether or not Bundy should have remained in the game. It is about whether the data cited truly support the motion that batters have an advantage when they see a pitcher the third time in the game.

My contention is that this data is misunderstood and misused. The main factor driving a poorer OPS a pitcher has the third time through is simply the quality of opponent in the population of that sample. It is heavily skewed towards the top of the line up.

It should not be surprising that Bundy’s OPS against for hitters 1-2 in the line up is .957 and the OPs against of 7-9 is .601. No one says it because it is obvious. The batters at the top of the line up are are far superior. That sample of the third time through is made up of a bunch of 1-2 hitters and hardly any of the 7-9. That skew in the sample gives the false impression that the third time a pitcher faces the top of order is somehow different than the first two times.

[chpettit19]

18 minutes ago, jorgenswest said:

It really doesn’t feel like you have read my thoughts. This thread has taken many interesting and unintended directions about the Twins move to take out Bundy,

It isn’t about millions of data points. It isn’t about whether or not Bundy should have remained in the game. It is about whether the data cited truly support the motion that batters have an advantage when they see a pitcher the third time in the game.

My contention is that this data is misunderstood and misused. The main factor driving a poorer OPS a pitcher has the third time through is simply the quality of opponent in the population of that sample. It is heavily skewed towards the top of the line up.

It should not be surprising that Bundy’s OPS against for hitters 1-2 in the line up is .957 and the OPs against of 7-9 is .601. No one says it because it is obvious. The batters at the top of the line up are are far superior. That sample of the third time through is made up of a bunch of 1-2 hitters and hardly any of the 7-9. That skew in the sample gives the false impression that the third time a pitcher faces the top of order is somehow different than the first two times.

I fully understand the point you're trying to make, and I'm directly refuting your stance on that because you're going off of data from "about 5 years ago." The millions of lines of data go back to when 1-2 hitters in lineups were slap hitter, table setters. And every starter, Bundy types and Kershaw types, were all asked to go deep into the game. So, yes, it is 100% about millions of data points. Teams used that data to adapt their approach to not let their mediocre starters face the heart of the order a 3rd time because that data, which isn't skewed by the things you're concerned recent data is skewed by, showed that letting mediocre pitchers face a lineup (especially the big boppers that used to hit 3-5) a 3rd time was worse than letting a middle reliever face them the first time.

Now you're looking at the data from the adjustment that was made to shorten starts. Offenses adjusted by taking the Punto and Casilla type slap hitters out of the 1-2 holes and putting Buxton and Judge types there. The question is how teams adjust their pitching next. The league is trying to get them to go back to longer starts by implementing roster restriction rules. We'll see what happens. But, yes, the recent data has some bias on both sides that lead it closer to the middle. If a bad pitcher gets into the lineup a 3rd time they're mostly facing the big time hitters. But most of the pitchers who pitch into a lineup a 3rd time frequently are the best pitchers in the game so they're lowering the stats. So you'd expect there to be a reduced difference in numbers when it's the best hitters and best pitchers making up the majority of the stats on both offense and defense. My stance is you're going off inadequate data.

[Shaitan]

Maybe Bundy had to go to the bathroom and Rocco didn't want to bring that up in the presser.

[heresthething]

Interesting point to me about this debate is what the hitters are doing by the third time around.  Am recalling a KFAN interview Berriro did with retired ump Tschida.  His idea was to limit the amount of in game electronics, he was saying these hitters are watching video of the pitcher and it's giving them an advantage they never used to have.  I don't think hitters are better then they used to be, maybe overall the quality is better?  So if even #1 pitchers are getting pulled early there's more to this debate than just the pitcher.

[Derojo]

Ultimately, an inspiring game of baseball is more about competitive excellence than winning. Pitching five innings of shutout baseball is less excellence than 6, 7, 8, 9, 10. Frankly, in today's game we would have never known Jack Morris's excellence in the 91 WS, 7th game. My opinion, but I remember when the pursuit of excellence carried more weight than winning, especially as compared to today. A No-hitter or perfect game is far more impressive by one pitcher because mentally and physically it is harder to attain. Excellence does matter because it inspires excellence. As for winning, someone always wins, even in the crappiest game played in MLB history.

[bean5302]

8 hours ago, jorgenswest said:

If this is real we should also see hitters batting better the third time through. I went through the 2021 AL Central and looked at the starting players for each team. I ended up with a group of 48. What I found isn't much different from what I found in a larger leaguewide study about 5 years ago. 

There were 17 of 48 hitters that had their highest OPS in the third time they saw the pitcher. That is about 1/3. This time they were most successful upon seeing the pitcher the second time and least successful the first time. I also found the 22 of the 48 hitters had the lowest OPS of the three seeing a pitcher the third time. Hitters were more likely to have their worst numbers when seeing a pitcher the third time. There is a skew here also. You probably only get to see a pitcher a third time if they are pitching well...

 

It seems like a sound idea on the surface, but unless you adjusted for how batters performed against only the starter on the 3rd time through the order, it's not apples to apples. By the TTO, in a large percentage of cases, the starter is no longer pitching.

This year there are 130 starters who've pitched 25+ innings 1st TTO, 123 who've pitched 2nd TTO, but even at just 10 innings, only 105 pitchers have gone 3rd TTO. The sample size changes because of fangraphs locked defaults.

[Coach Wheels]

I think when your bullpen is bad as the Twins bullpen, you let your starters get as close to 18 outs every night.