BA's Carolina League Top Twenty

[Lucky Jim]

I knew I saw the kid throwing and benefitting from a slow curve.

But you also realize that no one's saying he didn't benefit from a slow curve, too, right?

[crawdad]

And then compare it to the progression in Matusz's IP per start.

I mean, looking at average IP per start gives you a lot of noise-polluted information. For instance, he only went into the seventh in one of his last 5 starts. And his big IP games came mostly in a stretch of six starts. Here you go:

His IP by start:

5.0

4.0

6.0

6.0

5.2

6.1

5.2

8.0 (May 29/CB start).

7.0

8.0

9.0

6.1

6.0

8.0

6.0

6.2

6.0

6.0

7.0

Absent six starts (his poor first two, and his excellent four mid-season), Bergesen appears to be a 6 IP for the most part, even well after the introduction of the curveball. So, what are the odds that he just got hot?

This is the precise trap that folks fall into when they let their biases factor into their analysis. This sample is tiny, and therefore easily polluted. Further, it takes something virtually unmeasurable, creates a non-falsifiable hypothesis out of it, and uses it to validate a pre-conception/pre-determination of Bergesen's worth.

That sends up all kinds of red flags for folks. It plays out from here, like this: neither of you thought BB would be a legit MLB starter (I was skeptical too, mind you). When he started getting successful, you identified a difference. And that had to be it. But to what extent is this new theory lashed to your pre-determined notions? Because both of you well know that absent any real ability to quantify that difference (or even prove it was the difference) the entire line of analysis is problematic.

I'm glad I started this conversation because I think it's an interesting one. And I'm really only playing devil's advocate here, so please don't be offended by my combative tone.

That said, if Tillman was promoted before his change-up was as developed, and then his change-up came around, I wouldn't say that he went from a "bad" pitcher to a "good" pitcher. And the use of "world changing" and "good" and "bad" implies a significant bias (probably one inherent in scouting circles, and as we all agree, probably rightly so) against GB pitchers.

In the end, the point most of us are trying to make is only that we can't really know what role the curve is playing in BB's success. We all acknowledge it's playing some role. But most of us disagree that Bergesen's prior success was an aberration and that the curve somehow reinvented him. And that's the very rhetoric that Craw was using.

It would be one thing if he'd never been successful before. But that's not the case. So, while he clearly needed something else to work in the big leagues, the "reason" for his success continues to be the fastball that moves an inordinate amount. Now, that "something else" very much appears to be a curve. But it's a curve that's somewhere between a show-me curve and an MLB average curve that he throws 10%. So, while it's an important factor, how much weight we give it - on both sides - would seem to be a matter of how good we thought Bergesen was before.

Except for me. Who didn't think Bergesen was all that great before. And who doesn't think that Bergesen is going to be all that great going forward.

Trying to pin his success to that curve, rather than the things that had made him successful over the past two years, really does seem the epitome of ex-post rationalization. And ignoring the curve would be the same, for those who claimed he would come up and succeed as is. Absent any good data - and IP per start for a rookie pitcher clearly isn't good data - it appears the weight given to the curve is merely a matter of bias.

Well, the IP difference between those two populations has a p<0.01 using an alpha of 0.05, so that seems discernible.

I think you are focusing way too much on words like good, bad, and world changing. How about this . . . Bergesen looked and racked up with his peripherals everything that made him look like a spot starter and middle reliever. Then he came up to the majors and pretty much performed as such. Most likely realizing that his stuff that worked in the minors was not working in the majors, he dipped back into developing a pitch he gave up on. Somehow he figured out how to use that pitch and he became a different pitcher. I really think that curveball changed his standing immensely. There is a considerably big difference between being a middle reliever and being a 3/4 starter.

[Hank Scorpio]

But you also realize that no one's saying he didn't benefit from a slow curve, too, right?

Not in this thread, they didn't.

[Stotle]

And then compare it to the progression in Matusz's IP per start.

I mean, looking at average IP per start gives you a lot of noise-polluted information. For instance, he only went into the seventh in one of his last 5 starts. And his big IP games came mostly in a stretch of six starts. Here you go:

His IP by start:

5.0

4.0

6.0

6.0

5.2

6.1

5.2

8.0 (May 29/CB start).

7.0

8.0

9.0

6.1

6.0

8.0

6.0

6.2

6.0

6.0

7.0

Absent six starts (his poor first two, and his excellent four mid-season), Bergesen appears to be a 6 IP for the most part, even well after the introduction of the curveball. So, what are the odds that he just got hot? What are the odds that this is, in some way, equally correlated to comfort level.

Just as much as I watched the curveball and thought it appeared to help, when BB first came up I saw a nervous pitcher who was hurrying through his motion and speeding the game up when he got into trouble. What role does his reduction in panic play?

This is the precise trap that folks fall into when they let their biases factor into their analysis. This sample is tiny, and therefore easily polluted. Further, it takes something virtually unmeasurable, creates a non-falsifiable hypothesis out of it, and uses it to validate a pre-conception/pre-determination of Bergesen's worth.

That sends up all kinds of red flags for folks. It plays out from here, like this: neither of you thought BB would be a legit MLB starter (I was skeptical too, mind you). When he started getting successful, you identified a difference. And that had to be it. But to what extent is this new theory lashed to your pre-determined notions? Because both of you well know that absent any real ability to quantify that difference (or even prove it was the difference) the entire line of analysis is problematic.

I'm glad I started this conversation because I think it's an interesting one. And I'm really only playing devil's advocate here, so please don't be offended by my combative tone.

That said, if Tillman was promoted before his change-up was as developed, and then his change-up came around, I wouldn't say that he went from a "bad" pitcher to a "good" pitcher. And the use of "world changing" and "good" and "bad" implies a significant bias (probably one inherent in scouting circles, and as we all agree, probably rightly so) against GB pitchers.

In the end, the point most of us are trying to make is only that we can't really know what role the curve is playing in BB's success. We all acknowledge it's playing some role. But most of us disagree that Bergesen's prior success was an aberration and that the curve somehow reinvented him. And that's the very rhetoric that Craw was using.

It would be one thing if he'd never been successful before. But that's not the case. So, while he clearly needed something else to work in the big leagues, the "reason" for his success continues to be the fastball that moves an inordinate amount. Now, that "something else" very much appears to be a curve. But it's a curve that's somewhere between a show-me curve and an MLB average curve that he throws 10%. So, while it's an important factor, how much weight we give it - on both sides - would seem to be a matter of how good we thought Bergesen was before.

Except for me. Who didn't think Bergesen was all that great before. And who doesn't think that Bergesen is going to be all that great going forward.

Trying to pin his success to that curve, rather than the things that had made him successful over the past two years, really does seem the epitome of ex-post rationalization. And ignoring the curve would be the same, for those who claimed he would come up and succeed as is. Absent any good data - and IP per start for a rookie pitcher clearly isn't good data - it appears the weight given to the curve is merely a matter of bias.

Shrug. If you wat to say it's correlation rather than causation, fine. I know watching him through his first four or five starts it was apparent that around the fifth inning teams were absolutely locked in on his stuff. This isn't something you are going to chart out in a statistic (unless you go to the hit type of each hitter he faced). Teams were hitting him harder. The first time around he'd get more strikes looking, and he'd get more players making soft contact down in the zone. I don't expect to convince you of anything -- there isn't a magic number, metric or stat pool I can show you that conclusively his CB has been a major factor in his ability to keep hitters off balance and work through the lineup an extra time.

This is what I wrote after pitch charting him in his second start:

Impressions from Start #2: Again, no real surprises. When he's spotting his pitches, he can be difficult to square-up on, but without that changeup and without an ability to change the hitter's eye-level, I just don't see him making it through Major League lineups two and three times. It certainly isn't enough to say he isn't ready to be up with the big club, but he needs to get that changeup working, and he needs to have more confidence up in the zone with his four-seamer. Hitters are just way too comfortable by their second at bat.

So your claim is that because I said he was missing X (changing eye-level and speed), then he added X in the form of a curveball and started succeeding a lot more, I'm letting my bias cloud the picture. If that's your thought, fine. I'm okay with it. Just want to be clear that I'm not looking back and making this analysis -- it was something I spotted, thought and wrote about before he started throwing a curve. Anyway, I'm content in my analysis from a quasi-scouting perspective, but understand it falls short of what you require from an emperical perspective.

But I'm right....

[Lucky Jim]

Well, the IP difference between those two populations has a p<0.01 using an alpha of 0.05, so that seems discernible.

I think you are focusing way too much on words like good, bad, and world changing. How about this . . . Bergesen looked and racked up with his peripherals everything that made him look like a spot starter and middle reliever. Then he came up to the majors and pretty much performed as such. Most likely realizing that his stuff that worked in the minors was not working in the majors, he dipped back into developing a pitch he gave up on. Somehow he figured out how to use that pitch and he became a different pitcher. I really think that curveball changed his standing immensely. There is a considerably big difference between being a middle reliever and being a 3/4 starter.

I know you think this. Everyone knows you think this. Your reliance on tiny, and very noisy, samples to support your point remains problematic. Whether you consider the difference discernible or not.

Your analysis of the difference between a middle reliever and a mid-rotation starter is relevant only if 1.) you think he was a middle reliever before; and 2.) you think the curveball is what makes him a mid-rotation starter.

Neither have been proved, even remotely.

You've crafted a narrative that fits you pre-determination, and you've propped it up w/ what's available. The problem is, that what's available isn't really able to prop it up.

Common sense says that it's not going to be uncommon to see a difference in a rookie pitcher's first MLB starts and his 15-20th, for instance.

I'd also like to note that, on several occasions, I've seen you chastise folks for lack of precision in what they've written on this board (including me). I don't think I'm making too much of phrases like "world changing" or "bad groundball pitcher." You clearly see a pardigm shift. But there aren't really any paradigms available to shift. There's only a pre-season appreciation of Bergesen, and a tiny sample of starts.

[LookinUp]

I mean, looking at average IP per start gives you a lot of noise-polluted information.

...

This is the precise trap that folks fall into when they let their biases factor into their analysis. This sample is tiny, and therefore easily polluted. Further, it takes something virtually unmeasurable, creates a non-falsifiable hypothesis out of it, and uses it to validate a pre-conception/pre-determination of Bergesen's worth.

...

Trying to pin his success to that curve, rather than the things that had made him successful over the past two years, really does seem the epitome of ex-post rationalization.

Do you kiss your wife with that mouth? :laughlol:

[crawdad]

I know you think this. Everyone knows you think this. Your reliance on tiny, and very noisy, samples to support your point remains problematic. Whether you consider the difference discernible or not. You've crafted a narrative that fits you pre-determination, and you've propped it up w/ what's available. The problem is, that what's available isn't really able to prop it up.

Common sense says that it's not going to be uncommon to see a difference in a rookie pitcher's first MLB starts and his 15-20th, for instance.

I'd also like to note that, on several occasions, I've seen you chastise folks for lack of precision in what they've written on this board (including me). I don't think I'm making too much of phrases like "world changing" or "bad groundball pitcher." You clearly see a pardigm shift. But there aren't really any paradigms available to shift. There's only a pre-season appreciation of Bergesen, and a tiny sample of starts.

That is incredibly disrespectful. Apologize.

[Lucky Jim]

Shrug. If you wat to say it's correlation rather than causation, fine. I know watching him through his first four or five starts it was apparent that around the fifth inning teams were absolutely locked in on his stuff. This isn't something you are going to chart out in a statistic (unless you go to the hit type of each hitter he faced). Teams were hitting him harder. The first time around he'd get more strikes looking, and he'd get more players making soft contact down in the zone. I don't expect to convince you of anything -- there isn't a magic number, metric or stat pool I can show you that conclusively his CB has been a major factor in his ability to keep hitters off balance and work through the lineup an extra time.

This is what I wrote after pitch charting him in his second start:

So your claim is that because I said he was missing X (changing eye-level and speed), then he added X in the form of a curveball and started succeeding a lot more, I'm letting my bias cloud the picture. If that's your thought, fine. I'm okay with it. Just want to be clear that I'm not looking back and making this analysis -- it was something I spotted, thought and wrote about before he started throwing a curve. Anyway, I'm content in my analysis from a quasi-scouting perspective, but understand it falls short of what you require from an emperical perspective.

But I'm right....

I haven't once debated whether or not the curveball matters. I think it pretty clearly played a role in BB's success. I was only responding to what I thought was the over-statement in this post:

Lets not get ahead of ourselves, Bergesen's coming out party was a product of utilizing a brand new pitch once he hit the majors. It was not a product of what he was doing in the minors.

It seems many scouts have a hard time distinguishing between genuinely good groundball out guys and bad ones. Next season will be a good indication of which camp Britton falls into. Bergie was a bad groundball out guy . . . but, again, he figured out a curveball and the world changed for him.

[Hank Scorpio]

Someone's got a case of the Mondays!!!!!!!!

[Lucky Jim]

That is incredibly disrespectful. Apologize.

Are you going to challenge me to a duel?

And, no, to be clear, I'm not going to apologize.

[Lucky Jim]

Do you kiss your wife with that mouth? :laughlol:

You think any woman would stick around long enough to marry a mouth like that?

[crawdad]

I haven't once debated whether or not the curveball matters. I think it pretty clearly played a role in BB's success. I was only responding to what I thought was the over-statement in this post:

And you have overreacted on this point . . . what my hypothesis is that he became a different pitcher when he began utilizing a curveball. That is it. The world changed for him in that the use of the curveball appears to be the major change. There is some periphery that backs this us. But I have written again and again that there is not enough information to back this up 100%.

What is known is that the consensus was very doubtful of Bergesen. Now, it kind of is not.

What I did not need was someone misconstruing my statement after I already explained it by claiming that I was being disingenuous in my rationale. That was incredibly bush league. Incredibly bush league.

[Lucky Jim]

Obviously folks are riled up. I've pushed this conversation because I think it's an interesting look at what we can see, what we can't see, and what we do at the intersection of those two things.

Generally, playing along these lines would be impossible, but Craw and Stotle are both savvy, and highly informed, posters who have methodologies that can be pushed against.

If I knew it was going to result in pre-noon apoplexy, I wouldn't have engaged in it. Lesson learned.

I won't push back.

[crawdad]

Obviously folks are riled up. I've pushed this conversation because I think it's an interesting look at what we can see, what we can't see, and what we do at the intersection of those two things.

Generally, playing along these lines would be impossible, but Craw and Stotle are both savvy, and highly informed, posters who have methodologies that can be pushed against.

If I knew it was going to result in pre-noon apoplexy, I wouldn't have engaged in it. Lesson learned.

I won't push back.

Next time do not slander my scientific process. Claiming that I only consider information that supports me is an attack on my character. This is not about people being oversensitive. It is about being able to engage in a conversation by critiquing the idea and not the person who devised a process.

[Lucky Jim]

And you have overreacted on this point . . . what my hypothesis is that he became a different pitcher when he began utilizing a curveball. That is it. The world changed for him in that the use of the curveball appears to be the major change. There is some periphery that backs this us. But I have written again and again that there is not enough information to back this up 100%.

What is known is that the consensus was very doubtful of Bergesen. Now, it kind of is not.

What I did not need was someone misconstruing my statement after I already explained it by claiming that I was being disingenuous in my rationale. That was incredibly bush league. Incredibly bush league.

You're taking this far too seriously. If you think you don't bring "bias" to the table when you look at things, then you're bringing more to the table than I even imagined. The point wasn't to point out any disingenuousness. It's to point out how difficult it is to see things objectively, especially when dealing with imperfect information and small samples.

Needless to say, it's not a dialogue I'll wade into again. Thanks for everything so far, though. I thought it was pretty enlightening. I hope the board did too.