"Times through the order" penalty for pitchers?

[OldFatGuy]

Originally Posted by ThePretender

Pretty easy to separate what is fatigue and what is TTO penalty. Fatigue would result in pitches being less effective and having less movement, possibly less velocity, and control not being as sharp. That's not the case.

Really? So you're saying that the fourth time through the order "benefit" that hitters get is 100% TTOP and 0% fatigue?

So... pitchers don't fatigue? Or if they do, there is no resulting added offense as a result?

I would recommend a reading of this: http://www.baseballprospectus.com/ar...rticleid=28506

[ThePretender]

I've read it before, and there's more evidence suggesting the issue is TTO. In some cases a pitcher may have fatigue issues, but generally speaking it's due to batters being more familiar with the pitches and sequencing than simply fatigue. It's a lazy answer, IMO,

[OldFatGuy]

nm.... evidence is all right there in the link. But whatever.

[slugger922cubs]

Originally Posted by ThePretender

I've read it before, and there's more evidence suggesting the issue is TTO. In some cases a pitcher may have fatigue issues, but generally speaking it's due to batters being more familiar with the pitches and sequencing than simply fatigue. It's a lazy answer, IMO,

Its incredibly hard, nearly impossible, to separate fatigue from times through the order. The 2 are intrinsically related. You can't get to 3rd and 4th time through the order without throwing lots of pitches. Also of note that the live pitches which calculate the pitch count are not the only pitches the pitcher throws. It doesn't show 30+ pre game warm up pitches and 8 pitches an inning to warm up. The difference between a number of pitches thrown for a quote low-pitch count pitcher and high pitch count pitcher ends up actually being pretty close when all of those pitches are included. It also doesnt take time into the equation. We know long innings both pitching and sitting do have negative effects.
Complicating it is TTOP is not a hard rule. It is no more than one factor of many when removing a pitcher. It does not affect all pitchers equally or much at all. Some pitchers have shown they get stronger later. TTOP mostly seems to affect pitchers who lack stuff and go figure "fatigue early" and run out of gas.
Saying it is "generally speaking" a fact that TTOP is more prominent than fatigueis hardly the case. Its pretty easy to look at the facts and see for most reliable SP fatigue is the driving factor. Most of the type of pitchers affected by lacking stuff and getting a TTOP are back of the rotation pitchers who just cant get it done more than 18 batters and get tired early.

[RonCo]

It makes sense that (and seems to be borne out in the data that) as human beings:

1) A pitcher will use his max effort at all times when on in relief, and might not at all times when starting. This idea should be in the game, and is.
2) A batter will hit a pitcher better after seeing more of his pitches. This should be in the game, and is.
3) A pitcher will lose some effectiveness at certain points as he fatigues. This should be in the game, and is.

So, what's the argument? The best that I can come up with is some folks arguing certain set points should be mushed one way or the other.

[slugger922cubs]

Does OOTP model it to some form. Yes it seems incredibly likey that is the truth. OOTP seems to do a decent job modeling pitching fatigue and when to pull starters in that sense.
The disagreement seems to be that some people seem to think that TTOP is a law of pitching management and we should have more information pertaining to it to make decisions. Whereas more information is good and being able to break down stats based on time through the order would be great. TTOP is just not an underlying law of pitching management. Its one factor in many for that has more value with certain types of pitchers. It is no more so than managing fatigue, rest, spot in the order and hot vs cold games.
Now I agree lets put an end to this nonsense. If you think I'm dumb for letting my ace go over 3 times through i think its a bad move just to pull a good pitcher because hes magically over batter 18.To each his own. That is why baseball is great. Numbers can give us insight and change how the game is played but theyre never fully complete the picture. It is a human game after all.

[chazzycat]

Originally Posted by slugger922cubs

If you think I'm dumb for letting my ace go over 3 times through i think its a bad move just to pull a good pitcher because hes magically over batter 18.To each his own.

I don't think anyone is saying you should pull your ace like that. The takeaway here should be to pull your bad/borderline starters earlier. An ace would be starting from a much higher point, so they could withstand a small penalty and still be the best option.

[NoOne]

if the pitches behave in the same way as they did previously, it's most definitely not fatigue. if a cutter doesn't cut as much, then it's clearly part of the phenomenom of study. this would be the simplest way to rule this out or show that it is defintely a possible cause due to it's change in behaviour.

there must be a cause for change! if behaviour remains the same, it is not the impetus of that change in result!

the phenomenom is an end result... it's no doubt a culmination of multiple things working together. however, the weights of those things are what people are arguing about. don't get me wrong, i like arguing too

you simply map out all of the forces invovled and see what the data says... you don't pick cherry pick 1 thing and say that is a good answer, like that baseball prospectus link did.

this guy is a psycologist.. .his stats class was some dumbed-down "stats for the life sciences" BS... it's like "calculus for the life sciences" that M.D.'s take... it's stupid-calc and stupid-stats. i know, i took theh calc one and a real calc class, lol. you learn the same concepts but in a less sophisticated way/less depth. just enough to be able to understand the results of various white papers/research.

he picked a path instead of letting the path present itself... you will easily find false-positives doing things this way. he found an answer he wanted to find. you can't go into it "wanting" somethign ... most people cannot be impartial in that light.

if you want to take a mass amount of data and make it say somethign specifically, you can do so... even when that statment is likely wrong. and all the math can be perfectly done and logical... but the overall approach is the fault.

[RonCo]

And, just for historical relevance...http://www.ootpdevelopments.com/boar...5&postcount=40

[slugger922cubs]

Originally Posted by NoOne

If the pitches behave in the same way as they did previously, it's most definitely not fatigue. if a cutter doesn't cut as much, then it's clearly part of the phenomenom of study. this would be the simplest way to rule this out or show that it is defintely a possible cause due to it's change in behaviour.

But velocity, spin rates, and movement do decline. Especially on fastballs and especially on less effective pitchers. The exact ones that are allegedly prone to this TTOP effect. Now we can debate at what range is it negligible or not but there are certain changes.

data presented in:
http://www.sabermagician.com/visitin...order-penalty/

[RonCo]

Originally Posted by slugger922cubs

The disagreement seems to be that some people seem to think that TTOP is a law of pitching management and we should have more information pertaining to it to make decisions.

I don't read much of the questions to be of that nature, but to each his own. I think the conversation about whether a pitcher's loss of effectiveness is because of only fatigue or because of hitter familiarity is interesting. Part of the issue is that every pitcher is different in his ability to throw a pitch-count. So the equations work fine, and knowing they are there is not much of an advantage or disadvantage, for that matter...because the "right time" to pull pitcher A will be different from the "right time" to pull pitcher B.

In this way, OOTP seems to me to be very much directionally correct, leading to the only real argument being as to whether the performance adjustments are placed in the right place pitcher-by-pitcher.

At the end of the day, since I'm positive Markus and crew wouldn't release where those are and how big they are, an OOTP GM/manager will always need to get a feel for their guys to judge this. Just like real life, eh?

[slugger922cubs]

My impression was that ta couple people were trying to break it down into that a pitcher gets x-penalty each time through the order or the third+ times through the order and that's the biggest disagreement because if a penalty exist it's certainly not a constant and it is something that at best can only be isolated on a pitcher by pitcher basis and not in the aggregate. As acknowledged pulling an ace on TTOP. Pulling a fifth starter not so much. Very well the difference between an ace and a three starter and a fifth starter seems to be the ability to maintain stuff deeper in games. That much the numbers does tell us.

[RonCo]

There will always be some who simplify things. Bottom line, I'm 100% sure that OOTP, as sophisticated as it is today, gets this "wrong" to some degree, because the real life model is extremely complicated (and really just now being deeply studied), and because the model by nature has to modularize the basic nature of things in ways to make the game work while at the same time giving it a deep truthiness that appeals.

Regardless, the game is pretty sophisticated here, and as I noted above, has been for the last decade.

[OldFatGuy]

Originally Posted by slugger922cubs

My impression was that ta couple people were trying to break it down into that a pitcher gets x-penalty each time through the order or the third+ times through the order and that's the biggest disagreement because if a penalty exist it's certainly not a constant and it is something that at best can only be isolated on a pitcher by pitcher basis and not in the aggregate. As acknowledged pulling an ace on TTOP. Pulling a fifth starter not so much. Very well the difference between an ace and a three starter and a fifth starter seems to be the ability to maintain stuff deeper in games. That much the numbers does tell us.

And because the evidence shows it to be so different on a pitcher by pitcher basis, that alone is evidence that it's more fatigue than TTOP. TTOP is a BATTER thing... it's a benefit the batter supposedly gets by seeing a pitcher multiple times on the same day and thus the BATTER is more effective. If there is such a benefit, then it shouldn't matter who the pitcher is... because it's a batter thing.

Fatigue, on the other hand, is a PITCHER thing... it is a change in his effectiveness based on the changes he endures as his body tires.

As I've said before, there is probably something to the idea that batters gain some advantage the more they see a pitcher overall and the more they see him in a given day.... but most of the decline in effectiveness over the course of a game is almost certainly due to pitcher fatigue and not this "benefit."

[Rain King]

If it is a "pitcher" thing, it could also be just as much about the diversity of their arsenal as it is fatigue.

[OldFatGuy]

Originally Posted by Rain King

If it is a "pitcher" thing, it could also be just as much about the diversity of their arsenal as it is fatigue.

True that, but if they have a limited pitch arsenal, they probably aren't starting pitchers, and therefore haven't been included in any of the studies regarding the TTOP effect.

But even if we do consider that, it is still "a pitcher thing".... and therefore it's a mistake to just program in a TTOP effect for everyone.

Perhaps the best way to handle it is to program in a significant, across the board TTOP penalty for all pitchers with only 2 pitches that kicks in after the first time through (thus forcing you almost to make them relievers) and programming fatigue to occur over the course of the game for starters as a way to simulate both a real TTOP effect and the fact that with starters it's probably really more of a fatigue thing. And for all I know, that's exactly the way OOTP does it. *shrug*

[RonCo]

Originally Posted by OldFatGuy

As I've said before, there is probably something to the idea that batters gain some advantage the more they see a pitcher overall and the more they see him in a given day.... but most of the decline in effectiveness over the course of a game is almost certainly due to pitcher fatigue and not this "benefit."

I think you can play logic games on this until you get tied into knots. When you use "probably" or "almost certainly" in a sentence, you need to keep in mind that you don't actually know. And, in fact, I doubt there are many folks on this board who can actually say with 100% certainty exactly how the proportions should be weighted.

1. There is considerable data to show that it's a complex mapping of effects (fatigue, batter familiarity, effect of arsenal depth and quality among others).
2. All of those effects change by pitcher.
3. Some of those effects change by batter.
4. To leave any of them out would probably be wrong. Note the word "probably" there.
5. To weight any of them massively over the other would also probably be wrong. Again with the "probably."
6. I think the game needs to acknowledge all these in some series of algorithms that let them play together naturally, and therefore creates an overall effect that makes sense and feels right.
7. I think it seems to do it quite well overall. If one sees something deeply wrong in the results, it's good to highlight it.

[RonCo]

Originally Posted by OldFatGuy

Perhaps the best way to handle it is to program in a significant, across the board TTOP penalty for all pitchers with only 2 pitches that kicks in after the first time through ...

I'll say in advance that I have zero knowledge of this algorithm, but to do this well I would expect something like:

1) A TTO value that is built off the pitcher's arsenal and that uses both pitch quantity and quality in some fashion. Given OOTP's pitching model, I might also modifiy this value by the pitcher's STU/CON/MOV base ratings. I might also modify it with the pitcher's intelligence and maybe the catcher's intelligence to a lesser degree.
2) A Fatigue value that is based on the pitcher's endurance rating and the number of pitches he's thrown.
3) A batter's Perception value that is built on his base ratings, his intelligence, and the number of times he's seen the pitcher today (a pinch hitter would start at 0).

I would then mush these things together in some standard way to influence the final result. Despite the "standard" fashion, if the three factors work together well, each pitcher/batter matchup would be different every time they matched up.

Hence, no "playbook" of guaranteed decisions could really be made unless a GM/Manager was so amazingly anal that it wouldn't matter anyway.