Good Performances By Bad Pitchers

holes573 · 10-30-2024, 11:06 AM

Have you ever been frustrated by the frequency of stellar performance of low rated pitchers against your team? Starters that ought to be in AAA somehow pitch like Cy Young and just shut down your team.

Yes, I know that happens in real life. But, I'm questioning the frequency with which it happens in OOTP. My impression is that the developers have included some random element for pitchers that essentially gives them "good days," "regular days" or "bad days." And, if a pitcher is having a "good day," that affects their whole performance for the day - they not only shut your team down for an inning or two, but for their whole performance that day, until they tire.

I have now done a detailed analysis of all the starters that faced my team for a season (I was playing the Giants in 1964 with OOTP 25). Using a 20-80 system, I broke down starters that faced my team into three quality groups: 43 and lower, 44-48 and 49 and higher.

Each group of starters had good days and bad days. I defined a "good day" as having an ERA for that start of less than 2.00 (essentially allowing 0 or 1 ER for the outing). Against my team (which had the #1 offense in the NL), the percentage of "good days" for each group were:

43 and lower: 33% good days
44-48: 28% good days
49+: 23% good days

(Against all teams, not just against the Giants, the 43 and lower group had a rate of 28% good days.)

As a statistician, I know the differences between these groups do not reach statistical significance, but at the least we can conclude that the three groups were essentially equivalent in their percentages of good days.

By why should that be? Bad pitchers shouldn't have just as many days with 0 or 1 ER as Medium and Good pitchers. In addition, should as many as one-third of Bad pitchers' outings be dominant? That's not only frustrating to face, but I don't believe it's realistic. Perhaps the OOTP algorithm needs to be adjusted.

Rain King · 10-30-2024, 11:22 AM

How many starts from each of these pitchers are in your sample sizes?

Are you using 100% accuracy or are these ratings potentially skewered by a Scout's opinion?

What are the innings pitched per start for each group (i.e. is the manager possibly pulling the lower rated starters early while sticking with higher rated starters longer giving the opposition more chances to score against them)?

holes573 · 10-30-2024, 11:38 AM

The number of starts per group were:

43 or lower: 45 starts
44-48: 61 starts
49+: 56 starts

I was using my scout's ratings, but two things about that. First, I pump a lot of money into my scouting, so it's likely fairly accurate. But, more importantly, each group includes a wide range of ratings. So, it's likely that 100% accurate ratings of pitchers won't change the categorization much.

The average innings per start for each group was a bit longer for better pitchers:

43 or lower: 5.8 innings
44-48: 6.2 innings
49+: 6.5 innings

So, not huge differences (about a half inning or so). I hypothesize that this difference is due to differences in stamina (with higher rated pitchers have slightly better stamina that poorer rated pitchers).

Sweed · 10-30-2024, 12:50 PM

First good on you for doing a a detailed analysis. So many times threads are started with "it feels like" and nothing else.

I'm not a trained statistician so can only throw out questions that may, or may not, be something to consider.

My first would be BABIP for these pitchers?
Are K rates higher on the "good days"?
Walk rates lower on "good days"?

Lot's of variables and as you say you are a statistician maybe you have already taken these into account and more?

As for the game deciding player x is going to have a good, regular, or bad day? I don't want to know how the under the hood goings on actually work. I've played under the assumption that players can get hot or cold within a game (which can carry over to players getting the "hot or cold" icon and start a streak), but haven't thought that good/bad/regular was determined under the hood pregame. To me, my assumption seems a better (for lack of better word) way of doing it than pregame. In any case that gets me back to not wanting to know the inner workings of the game.

So I'll throw out my questions and check back in to see where the discussion goes. If it looks like it's getting to under the hood things I don't want to know I'll probably stay away from that point on.

jksander · 10-30-2024, 05:31 PM

Hard to analyse this without corresponding batter data relating to your batters and their hotness / coldness ...

kq76 · 10-30-2024, 07:03 PM

I'm not bothered by the idea of pitchers getting good, bad, or meh days. I think that's how it should work. You often hear pitchers saying they just didn't have their good stuff that day or they did.

What I wonder is, is the % pretty much the same for each pitcher, as you seem to be implying, or is it different. And I'm not sure we should assume it should be different. I mean sure, at first thought one would think better pitchers should have more good days than bad pitchers. But what if what it really is is when good pitchers have bad days they're not as bad as bad pitchers' bad days and when good pitchers have good days they have better days than bad pitchers, RA9 and IP-wise.

I'd be curious to see an analysis of this with real life pitchers.

mytreds · 10-30-2024, 07:22 PM

Quote:

Originally Posted by jksander

Hard to analyse this without corresponding batter data relating to your batters and their hotness / coldness ...

Did they change how hot/cold works? Because last I knew, it was just a visual, not an actual modifier.

OP, I’ve seen this often as well, mediocre pitchers somehow throwing 2 hit shutouts against my team. It’s easy to call foul on the game engine, but I chalk it up to variance over 162 games.

OutS|der · 10-30-2024, 10:31 PM

Pitcher overall rating is not the end all of stats. If you take a look into the editor you'll find that the projected WAR for pitchers can vary wildly between pitchers rated the same.

Displayed overall ratings have always been a guide, they have never truly meant one was "better", the formula changes every year.

holes573 · 10-31-2024, 02:10 PM

Lots of good discussion.

@Mytreds; Yes, getting blown away by a bap pitcher is just variance over 162 games. But, in this case, I measured that variance, and my interpretation of that data is that this happens too frequently.

@OutSider (sorry, I don't know how to type that line you have instead of "i"): Yes, there are other measures of pitcher quality that can be used (none of which are perfect), but the quality rating is the one that is most accessible in OOTP, and the likely the one most relied upon. My point is that this heavily relied upon quality rating implies that those with poor quality ratings have fewer good performances - but that wasn't true in my data.

OutS|der · 10-31-2024, 06:34 PM

Quote:

Originally Posted by holes573

@OutSider (sorry, I don't know how to type that line you have instead of "i"): Yes, there are other measures of pitcher quality that can be used (none of which are perfect), but the quality rating is the one that is most accessible in OOTP, and the likely the one most relied upon. My point is that this heavily relied upon quality rating implies that those with poor quality ratings have fewer good performances - but that wasn't true in my data.

Just because people don't understand how the ratings work doesn't mean we should keep inflating them. As mentioned the formula get changed every year.
You're also not using 100% scouting so there could be some fog of war you're missing.
In a 20-80 rating system, 5 points is worth around 0.5 WAR. That's nothing.

Change over to increments of 5 and you'll see how similar those players are. Try 100 accuracy as well, don't have to keep it that way . Can always force close the game so nothing is saved

darkcloud4579 · 11-01-2024, 12:26 AM

Quote:

Originally Posted by holes573

Have you ever been frustrated by the frequency of stellar performance of low rated pitchers against your team? Starters that ought to be in AAA somehow pitch like Cy Young and just shut down your team.

Yes, I know that happens in real life. But, I'm questioning the frequency with which it happens in OOTP. My impression is that the developers have included some random element for pitchers that essentially gives them "good days," "regular days" or "bad days." And, if a pitcher is having a "good day," that affects their whole performance for the day - they not only shut your team down for an inning or two, but for their whole performance that day, until they tire.

I have now done a detailed analysis of all the starters that faced my team for a season (I was playing the Giants in 1964 with OOTP 25). Using a 20-80 system, I broke down starters that faced my team into three quality groups: 43 and lower, 44-48 and 49 and higher.

Each group of starters had good days and bad days. I defined a "good day" as having an ERA for that start of less than 2.00 (essentially allowing 0 or 1 ER for the outing). Against my team (which had the #1 offense in the NL), the percentage of "good days" for each group were:

43 and lower: 33% good days
44-48: 28% good days
49+: 23% good days

(Against all teams, not just against the Giants, the 43 and lower group had a rate of 28% good days.)

As a statistician, I know the differences between these groups do not reach statistical significance, but at the least we can conclude that the three groups were essentially equivalent in their percentages of good days.

By why should that be? Bad pitchers shouldn't have just as many days with 0 or 1 ER as Medium and Good pitchers. In addition, should as many as one-third of Bad pitchers' outings be dominant? That's not only frustrating to face, but I don't believe it's realistic. Perhaps the OOTP algorithm needs to be adjusted.

This is realistic and wonderful and shows why the game is so good. You all are so hellbent on the game being a ratings = good dice roller and that's just not real life. Anybody in the bigs can have a hot streak, hot week or hot day. It's not that weird. Bad is an aggregate thing.

CubsFan1967 · 11-02-2024, 11:52 AM

Quote:

Originally Posted by holes573

Have you ever been frustrated by the frequency of stellar performance of low rated pitchers against your team? Starters that ought to be in AAA somehow pitch like Cy Young and just shut down your team.

Yes, I know that happens in real life. But, I'm questioning the frequency with which it happens in OOTP. My impression is that the developers have included some random element for pitchers that essentially gives them "good days," "regular days" or "bad days." And, if a pitcher is having a "good day," that affects their whole performance for the day - they not only shut your team down for an inning or two, but for their whole performance that day, until they tire.

I have now done a detailed analysis of all the starters that faced my team for a season (I was playing the Giants in 1964 with OOTP 25). Using a 20-80 system, I broke down starters that faced my team into three quality groups: 43 and lower, 44-48 and 49 and higher.

Each group of starters had good days and bad days. I defined a "good day" as having an ERA for that start of less than 2.00 (essentially allowing 0 or 1 ER for the outing). Against my team (which had the #1 offense in the NL), the percentage of "good days" for each group were:

43 and lower: 33% good days
44-48: 28% good days
49+: 23% good days

(Against all teams, not just against the Giants, the 43 and lower group had a rate of 28% good days.)

As a statistician, I know the differences between these groups do not reach statistical significance, but at the least we can conclude that the three groups were essentially equivalent in their percentages of good days.

By why should that be? Bad pitchers shouldn't have just as many days with 0 or 1 ER as Medium and Good pitchers. In addition, should as many as one-third of Bad pitchers' outings be dominant? That's not only frustrating to face, but I don't believe it's realistic. Perhaps the OOTP algorithm needs to be adjusted.

I play out all my games and consider facing a starting pitcher with a 6+ ERA an automatic loss. Also, the guy hitting .142 against the rest of the league will go 8 for 10 in the series against you.

Pdubya64 · 11-02-2024, 02:28 PM

Yeah, while there are times I get frustrated with OOTP, if I wanted a static Baseball game I would still be playing Stratomatic.

I think it's a bit unfortunate that OOTP's "nature" is such that you can't really ever be sure of statistical logic. There are so many variables like each head scout, coach and player... where the hell would you even start?

I say unfortunate in the sense of not being able to appreciate outcomes when you don't know what is happening inside the black box. Best not to think too much about it, at least that is what I decided.

dannibalcorpse · 11-02-2024, 11:32 PM

Quote:

Have you ever been frustrated by the frequency of stellar performance of low rated pitchers against your team?

i've been a Mets fan for 30+ years, i just figured this was how it was for everyone

fredbeene · 11-03-2024, 06:05 PM

Quote:

Originally Posted by Sweed

First good on you for doing a a detailed analysis. So many times threads are started with "it feels like" and nothing else.

I'm not a trained statistician so can only throw out questions that may, or may not, be something to consider.

My first would be BABIP for these pitchers?
Are K rates higher on the "good days"?
Walk rates lower on "good days"?

Lot's of variables and as you say you are a statistician maybe you have already taken these into account and more?

As for the game deciding player x is going to have a good, regular, or bad day? I don't want to know how the under the hood goings on actually work. I've played under the assumption that players can get hot or cold within a game (which can carry over to players getting the "hot or cold" icon and start a streak), but haven't thought that good/bad/regular was determined under the hood pregame. To me, my assumption seems a better (for lack of better word) way of doing it than pregame. In any case that gets me back to not wanting to know the inner workings of the game.

So I'll throw out my questions and check back in to see where the discussion goes. If it looks like it's getting to under the hood things I don't want to know I'll probably stay away from that point on.

Yes THANK YOU for included data and analysis! this helps clarify and allows users to brainstorm for possible additional analysis

It may not be as fun, but to test i always have to use 100% accuracy ratings to remove the bad scouting variable.
Also important to know the full set up. particularly dev settings and TC.

Is it possible next season these players are improving. Perhaps this is the first sign of future brilliance? : )

Guthrien · 11-03-2024, 07:09 PM

While I am extremely doubtful about a programming function that tries to level out 'bad pitchers' vs your team, I'd be lying if I said I never thought it. Especially in games I play out. I also try to keep in mind pitcher's who I've seen throw no or low hit games in reality the last few years who I would have trouble rating as "high quality".

I think, like so so many occurrences in this game, we are fooled by noise in what is after all statistically a thimbleful of events (I think you would agree as a statistician?) but might be happening because this is after all a programmed game.

No other game has convinced me to be less amenable to the suggestion of small data bias, and yet no other game has made me think it so much when I'm emotionally invested in success.

10-30-2024, 11:22 AM	#2
Rain King Hall Of Famer Join Date: Nov 2005 Posts: 2,876	How many starts from each of these pitchers are in your sample sizes? Are you using 100% accuracy or are these ratings potentially skewered by a Scout's opinion? What are the innings pitched per start for each group (i.e. is the manager possibly pulling the lower rated starters early while sticking with higher rated starters longer giving the opposition more chances to score against them)? Last edited by Rain King; 10-30-2024 at 01:43 PM. Reason: grammar

10-30-2024, 05:31 PM	#5
jksander All Star Starter Join Date: Oct 2002 Location: Indianapolis IN Posts: 1,380	Hard to analyse this without corresponding batter data relating to your batters and their hotness / coldness ... __________________ "Oh No! We Suck Again!" -- Reviving the White Sox in 2025 -- An OOTP 26 Dynasty Jochen "The Joker" Fontaine: The Road to Glory -- An OOTP 26 "First Person In-Character" Historical Dynasty "Ain't Gonna Work As Topping's Farm No More" -- A's Baseball in a Reimagined Fifties -- An OOTP 25 Dynasty

10-30-2024, 07:03 PM	#6
kq76 Global Moderator Join Date: Nov 2002 Location: Vancouver, Canada Posts: 11,485	I'm not bothered by the idea of pitchers getting good, bad, or meh days. I think that's how it should work. You often hear pitchers saying they just didn't have their good stuff that day or they did. What I wonder is, is the % pretty much the same for each pitcher, as you seem to be implying, or is it different. And I'm not sure we should assume it should be different. I mean sure, at first thought one would think better pitchers should have more good days than bad pitchers. But what if what it really is is when good pitchers have bad days they're not as bad as bad pitchers' bad days and when good pitchers have good days they have better days than bad pitchers, RA9 and IP-wise. I'd be curious to see an analysis of this with real life pitchers. __________________ My OOTP Wishlist \| My FAQ List OOTP Wiki \| Your Recommended Team Nicknames, By City (A Crowdsourced Project)

11-02-2024, 02:28 PM	#13
Pdubya64 Major Leagues Join Date: Jul 2008 Location: Staunton, VA Posts: 480	Yeah, while there are times I get frustrated with OOTP, if I wanted a static Baseball game I would still be playing Stratomatic. I think it's a bit unfortunate that OOTP's "nature" is such that you can't really ever be sure of statistical logic. There are so many variables like each head scout, coach and player... where the hell would you even start? I say unfortunate in the sense of not being able to appreciate outcomes when you don't know what is happening inside the black box. Best not to think too much about it, at least that is what I decided. __________________ "Chew, if only you could see what I've seen with your eyes." - Roy Batty Blade Runner

10-30-2024, 11:06 AM	#1
holes573 Bat Boy Join Date: Mar 2017 Posts: 14	Good Performances By Bad Pitchers Have you ever been frustrated by the frequency of stellar performance of low rated pitchers against your team? Starters that ought to be in AAA somehow pitch like Cy Young and just shut down your team. Yes, I know that happens in real life. But, I'm questioning the frequency with which it happens in OOTP. My impression is that the developers have included some random element for pitchers that essentially gives them "good days," "regular days" or "bad days." And, if a pitcher is having a "good day," that affects their whole performance for the day - they not only shut your team down for an inning or two, but for their whole performance that day, until they tire. I have now done a detailed analysis of all the starters that faced my team for a season (I was playing the Giants in 1964 with OOTP 25). Using a 20-80 system, I broke down starters that faced my team into three quality groups: 43 and lower, 44-48 and 49 and higher. Each group of starters had good days and bad days. I defined a "good day" as having an ERA for that start of less than 2.00 (essentially allowing 0 or 1 ER for the outing). Against my team (which had the #1 offense in the NL), the percentage of "good days" for each group were: 43 and lower: 33% good days 44-48: 28% good days 49+: 23% good days (Against all teams, not just against the Giants, the 43 and lower group had a rate of 28% good days.) As a statistician, I know the differences between these groups do not reach statistical significance, but at the least we can conclude that the three groups were essentially equivalent in their percentages of good days. By why should that be? Bad pitchers shouldn't have just as many days with 0 or 1 ER as Medium and Good pitchers. In addition, should as many as one-third of Bad pitchers' outings be dominant? That's not only frustrating to face, but I don't believe it's realistic. Perhaps the OOTP algorithm needs to be adjusted.

10-30-2024, 11:38 AM	#3
holes573 Bat Boy Join Date: Mar 2017 Posts: 14	The number of starts per group were: 43 or lower: 45 starts 44-48: 61 starts 49+: 56 starts I was using my scout's ratings, but two things about that. First, I pump a lot of money into my scouting, so it's likely fairly accurate. But, more importantly, each group includes a wide range of ratings. So, it's likely that 100% accurate ratings of pitchers won't change the categorization much. The average innings per start for each group was a bit longer for better pitchers: 43 or lower: 5.8 innings 44-48: 6.2 innings 49+: 6.5 innings So, not huge differences (about a half inning or so). I hypothesize that this difference is due to differences in stamina (with higher rated pitchers have slightly better stamina that poorer rated pitchers).

10-30-2024, 10:31 PM	#8
OutS\|der Hall Of Famer Join Date: Apr 2010 Location: In A Van Down By The River Posts: 2,571 Infractions: 0/1 (1)	Pitcher overall rating is not the end all of stats. If you take a look into the editor you'll find that the projected WAR for pitchers can vary wildly between pitchers rated the same. Displayed overall ratings have always been a guide, they have never truly meant one was "better", the formula changes every year.

10-31-2024, 02:10 PM	#9
holes573 Bat Boy Join Date: Mar 2017 Posts: 14	Lots of good discussion. @Mytreds; Yes, getting blown away by a bap pitcher is just variance over 162 games. But, in this case, I measured that variance, and my interpretation of that data is that this happens too frequently. @OutSider (sorry, I don't know how to type that line you have instead of "i"): Yes, there are other measures of pitcher quality that can be used (none of which are perfect), but the quality rating is the one that is most accessible in OOTP, and the likely the one most relied upon. My point is that this heavily relied upon quality rating implies that those with poor quality ratings have fewer good performances - but that wasn't true in my data.

11-03-2024, 07:09 PM	#16
Guthrien Minors (Double A) Join Date: Mar 2011 Posts: 154	While I am extremely doubtful about a programming function that tries to level out 'bad pitchers' vs your team, I'd be lying if I said I never thought it. Especially in games I play out. I also try to keep in mind pitcher's who I've seen throw no or low hit games in reality the last few years who I would have trouble rating as "high quality". I think, like so so many occurrences in this game, we are fooled by noise in what is after all statistically a thimbleful of events (I think you would agree as a statistician?) but might be happening because this is after all a programmed game. No other game has convinced me to be less amenable to the suggestion of small data bias, and yet no other game has made me think it so much when I'm emotionally invested in success.