Does pitcher stamina need to be changed?

[Garlon]

I really do not use the allow starters in relief option when I play but I suspect that comes in to play when basically all of the other relievers are unavailable. It may be difficult to code the game to allow starters in relief for some high leverage situations. If not done properly you will have your staff ace making 35 relief appearances and missing many starts. There would need to be some very specific criteria met like the game needs to be in extra innings and it cannot be a SP who just pitched recently, and the game needs to be tied or something. I think working out the strategy for how this could work in the game may be difficult.

[joefromchicago]

Originally Posted by Garlon

I really do not use the allow starters in relief option when I play but I suspect that comes in to play when basically all of the other relievers are unavailable. It may be difficult to code the game to allow starters in relief for some high leverage situations. If not done properly you will have your staff ace making 35 relief appearances and missing many starts. There would need to be some very specific criteria met like the game needs to be in extra innings and it cannot be a SP who just pitched recently, and the game needs to be tied or something. I think working out the strategy for how this could work in the game may be difficult.

I don't think it would be that difficult at all (although I have very unrealistic ideas about what it takes to do computer coding - I suspect it's all magic somehow). There's an option to give every member of the starting rotation a secondary role, and that includes a "high leverage" role. In my experience, though, the AI never assigns a secondary role to any starter. This is a huge wasted opportunity, in my opinion, and again it's a feature that's already in the game that could be utilized to improve historical accuracy without overhauling the pitching model.

As for how that might work in practice, I don't think a lot of additional "if-then" coding would be needed. If a starter with a secondary role of "high leverage" is too tired to pitch, then the game won't select him to pitch. That seems pretty straightforward to me. But, as I've pointed out before, a slightly tired starter is still usually a better choice to pitch in a save situation than a washed-up pitcher who is only in the bullpen because he's not good enough to start - which was usually the situation with "relievers" in those days.

And yes, top starters did lose starts because they were picking up relief opportunities. But that's historically accurate. It was not uncommon for the pitcher with the most IP on a team's staff to have the second-most starts. HOFers like Lefty Grove, Jesse Haines, Waite Hoyt, Ed Walsh, and Three-Finger Brown appeared 20% or more of the time out of the bullpen. And if the gamer doesn't want to see his top starter lose starts because he pitches in relief, there's a menu option already that will prevent starters from relieving. No need to make any changes to the AI.

So I don't think an overhaul of the AI would be needed. Just make the AI choose secondary roles for members of the starting rotation. That option is already there. For some reason, though, the game never uses it.

[kriscolic]

Originally Posted by Garlon

There really is no issue with how OOTP is handling pitcher stamina.

1. A pitcher is going to be given SP stamina if at least 40% of their Games Pitched were as Games Started. The formula is based upon IPouts per GS. This works in tandem with the League Totals Modifiers. Remember that you can choose to have pitcher Stamina determined based upon 1 season, 3 seasons, 5 seasons, or career.

2. If every pitcher were given SP potential stamina this would be extremely detrimental to the real starting pitchers historically as it would put relievers into the rotation and you will have guys like Mariano Rivera taking home the Cy Young Award as a starting pitcher.

3. Everything is currently in the game to create proper outcomes. Some settings need to be set to better default. For example, Use Relievers should be set to Normal for every season. Remember that the League Totals Modifiers will take care of IPouts per GS, so what remains is getting proper bullpen usage in terms of IPouts per relief appearance. If you set your pitching rotation and pitching staff limits correctly you can get great results for any season.

This guide will show you settings as well as historical statistics for various seasons. There are columns that show the expected average innings for relief pitchers for each team as well as the expected number of relief appearances. If you use the suggested strategy settings you will end up with really good results. Note, when I originally created this guide I put Use Closers as Very Often for every season, but I will update this. The reason I set it to that is for competitiveness because doing so has the computer identify the bullpen ace and by setting use closers to very often the computer will use their best reliever more appropriately in my open. The only issue with this is that you get higher save totals. I may have a revised version to provide appropriate closer settings.

Well that's just not necessarily true. All that would be needed would be two distinct sets of ratings for every pitcher, one for starting and one for relieving. To use your example, Mariano was a mediocre starter and a brilliant reliever. Just give him (and everyone else) two sets of ratings to reflect this fact. The game already alters Stuff for pitchers depending on their role. It's the right thought, but it doesn't go far enough.

[Lukas Berger]

Interesting conversation and with some qualifications, I do tend to agree with your conclusions.

I mentioned this to Markus and he's open to the idea that we can make some tweaks to the stamina calculations for those borderline 25-35% type guys.

Garlon, Joe, please remind me about this at some point not too long after we start beta testing.

I'm not as sure about the assigning secondary roles to SP's just because I'm not sure how much coding effort would be involved in getting this to work well and not just have the ai pitching the top SP's in relief every day if they do have a secondary role assigned. I did point this out to Matt though, and we'll see on that.

[joefromchicago]

Originally Posted by Lukas Berger

Interesting conversation and with some qualifications, I do tend to agree with your conclusions.

I mentioned this to Markus and he's open to the idea that we can make some tweaks to the stamina calculations for those borderline 25-35% type guys.

Garlon, Joe, please remind me about this at some point not too long after we start beta testing.

That's fantastic! I'll be sure to remind you guys.

Originally Posted by Lukas Berger

I'm not as sure about the assigning secondary roles to SP's just because I'm not sure how much coding effort would be involved in getting this to work well and not just have the ai pitching the top SP's in relief every day if they do have a secondary role assigned. I did point this out to Matt though, and we'll see on that.

In my replays, pitchers with 75-80 stamina (which is just about every starter who pitched more than 150 innings) are fully recovered four days after starting. On days 1 and 2, I would expect that a starter wouldn't be available to relieve because he's too tired. On day 3, he would be "slightly tired," which should mean that he's available to relieve if he's still a better choice than one of the pitchers in the bullpen. And then, on day 4, if he hasn't pitched in relief in between his starts, he should be available to start (that's another problem that I've identified in my replays - starters taking an extra day of rest between starts).

Some starting pitchers were used a lot in relief - I mentioned a few before, like Three-Finger Brown and Lefty Grove. Others, like Christy Mathewson and Eddie Plank, weren't. So not every starter should get assigned a secondary role as a reliever. I'd suggest that any starter who had 15% or more of his appearances in relief should get some kind of secondary reliever role. For the top starters who fit into that category, I'd suggest "high leverage." That should also work to limit their relief appearances.

I'm not sure how fast starters recover from relief appearances, but it should be faster than a normal "reliever." As I mentioned in my 1912 Senators replay, Walter Johnson started four times the day after he appeared in relief. Lefty Grove did the same in 1930. So it should be possible for a high-stamina starter to recover from a short relief stint immediately.

Likewise, pitchers that OOTP consigns to the bullpen should get "emergency starter" as a secondary role if they had more than 15% or 20% of their appearances as a starter. So adding secondary roles, combined with the stamina ratings and strategy settings, should work together to yield more realistic results for pitchers in the pre-reliever era.

[ALB123]

Wow. What a conversation. I can't believe I'm just discovering the thread.

I can appreciate both major views in this thread, both Garlon's and Joe's. I personally don't know enough about baseball history regarding Pitcher usage roles to intelligently form my own opinion.

So far, I have only been playing historical, but I don't do replays. My preference has been to pick a starting year and then let my own unique baseball universe play out, so anything I've observed is much different because the OOTP development engine and Talent Change Randomness is increased to 175, again, to enjoy a totally unique experience.

Still, I've had major concerns about relief pitcher stamina. First off, I must admit that I don't fully understand the Pitching Fatigue Level %. 100% is perfect, 0% is almost dead arm...that's easy to grasp, but I don't really know how much pitchers are affected with each single percentage difference. Is it a perfectly straight line regarding performance drop? Or, will a pitcher at 90% pitch just as well as he would at 100%? Obviously, he's going to hit that point when he's totally pooped much quicker if he starts pitching at only 90% compared to 100%, but that still doesn't answer my question about performance at 100% compared to performance at 90%

In real-life, let's say Mariano Rivera comes in to pitch the 9th inning and he faces 4 hitters, walking one of the batters...game over...Mariano threw 21 pitches and gets a save. If the Yankees are in a situation the very next night and bring in Mariano again to pitch the 9th inning and save the game, he is going to be just as effective, on point, in the zone, strong, etc. as he was the night previously. Hell, he might look and perform better! In OOTP, that's just not possible...as far as I know.

If Mariano comes into the 9th inning in my OOTP world and throws 21 pitches (he walks one guy like in the "real world" example above). He's going to be at 77% the next day. He's not in Yellow, but 77% feels to me like he couldn't give me another 21 pitch 9th inning as he could in the real world. Am I wrong?

I was very surprised when I was faced with this recently in my saved game:



112 pitches 5 days ago and he's not 100% rested? That was the first time I noticed it, but I have since seen it twice more like that. And based on prior experience, I don't believe Mariano will be back to 100% until 3 more days.

Am I making too big of a deal about this? Again, maybe I just don't understand that a 77% Mariano Rivera (or any other decent reliever) is still pretty awesome? If pitcher fatigue is a perfectly linear drop in performance from 100% to 0% then a 77% Mariano can't be nearly as good as a 100% Mariano.

I know this is much different than what you guys have been discussing, but it's still Pitcher Stamina related, so I figured I would give it a shot and post my questions here. I hope I have articulated my point well enough to understand what I'm trying to say. Sometimes the fingertips don't do what the squishy thing in my head attic wants.

[Garlon]

Keep in mind that there is also a game setting for pitcher stamina and that can affect the rate at which pitchers recover as well.

[joefromchicago]

Originally Posted by ALB123

112 pitches 5 days ago and he's not 100% rested? That was the first time I noticed it, but I have since seen it twice more like that. And based on prior experience, I don't believe Mariano will be back to 100% until 3 more days.

Am I making too big of a deal about this? Again, maybe I just don't understand that a 77% Mariano Rivera (or any other decent reliever) is still pretty awesome? If pitcher fatigue is a perfectly linear drop in performance from 100% to 0% then a 77% Mariano can't be nearly as good as a 100% Mariano.

Garlon's point about the league stamina setting is, I think, correct. As I mentioned earlier in this thread, there are two stamina ratings and together they measure two different things. There's the league stamina setting (found in the league's Stats & AI page) and each pitcher's individual stamina rating. Those, in turn, determine both in-game stamina (how many pitches a pitcher can throw before getting tired) and between-game stamina (how long a pitcher takes before recovering fatigue).

I'm not entirely sure how the stamina ratings work together, but I think the league setting establishes the "scale" and the individual pitcher ratings are a percentage of that scale. So, to use some made-up numbers, if the league setting is "high," and that setting establishes a scale of 1-100, then a pitcher with a 75/100 stamina rating would have a 75 "adjusted" stamina. If, on the other hand, the scale was 1-50, that pitcher would have a 37.5 "adjusted" stamina.

In your example, I don't know what the league stamina setting is, but my guess is that it's either "low" or "very low." In the replays that I've done in conjunction with my exploration of pitcher usage in the pre-reliever era, league settings have been set at "high." At a high setting, a pitcher like Martinez with a 65/80 stamina who threw 112 pitches in a game would probably be back to 100% within four days.

With regard to relievers, I don't know how quickly they recover or what their effectiveness would be if they pitch when they're less than 100% rested. Not surprisingly, relievers haven't been my focus when I've been looking at the pre-reliever era. I can say, however, that, in my experience, the AI is very reluctant to pitch anybody who is less than 100% rested - that's true for both starters and relievers.

At times, that leads to some very strange results. For instance, in my replays I try to keep rosters at more-or-less historical levels. That means that sometimes there may be only one or two pitchers in the bullpen. If those guys aren't 100% rested, the AI will simply stay with the starter, even though it might be better to pinch-hit for him. The AI sees the bullpen pitchers as unavailable because they're not fully rested, and it sees the other starters in the rotation as unavailable because they're either fatigued or else because the AI just doesn't like using starters in relief. As a result, the starter will stay in the game and throw an ungodly number of pitches. In my 1912 Senators replay, for instance, it wasn't unusual for Walter Johnson to throw over 200 pitches in a game.

That's not to say that starters in that era didn't throw a lot of pitches - they did. But the AI logic is premised on a strict split between starters and relievers, and it's very difficult for a pitcher to straddle the line between the two roles. Pitchers, however, did that all the time in the pre-reliever era. In that sense, the real-life managers had more options available to them than the AI allows them. If a starter like Walter Johnson got tired, his manager, Clark Griffith, would view every pitcher on the bench as a possible reliever. In OOTP, on the other hand, the AI looks first at the pitchers who aren't in the rotation. If they're unavailable, then the AI far too often is stumped. It won't use a top-end starter who is even minimally fatigued and it won't use a bottom-end starter because they're usually not very good. Griffith, in contrast, would have taken the game situation into context and said: "if it's close, I'll use my best pitcher, unless he's completely exhausted. If it's not close, any of those knuckleheads at the end of the bench will do." Having the AI assign secondary roles to pitchers might, I think, be one important way to address this problem.

[scurvycure]

This discussion is much-appreciated. I have started to check on the rest status of every team's pitching staff in my replay before each game. Anecdotally, it seems like bringing in a late inning reliever who is not completely rested is, more often than not, like throwing gas on a fire. Walk, single, wild pitch, walk, single, single. Man that is painful. As double headers, injuries, rain delays and extra inning games wear down a staff, I find too many games where a team is starting a tired starter, or has only one or fewer rested relievers. I decided to increase my league's roster size to ensure that every team had 12 man pitching staffs. That isn't realistic for 1972 but I would rather have that than unrealistic run production caused in part by pitcher fatigue.

[Garlon]

If you want to see more problems go to your single season leaders and look at batter strikeouts. You will see some ridiculous totals there. This is caused by the computer not assigning a substitute at some positions, particularly C and CF, but it could be any position. This will occur even if you give set rosters to allow for 16 position players. You end up with players on a team at certain positions who end up playing every inning of every game and they end up batting about .140 with 275 strikeouts regardless of the season.

[joefromchicago]

Originally Posted by scurvycure

I decided to increase my league's roster size to ensure that every team had 12 man pitching staffs. That isn't realistic for 1972 but I would rather have that than unrealistic run production caused in part by pitcher fatigue.

What settings are you using for rotation size and pitcher usage (strict order, start highest rested)? What are your settings for reliever usage?

[Garlon]

I am currently testing some pitcher settings. I do not think that Use Relievers should ever be changed from Normal, only the Pitcher Stamina, Use Closers and PH for Pitchers should be adjusted for different eras.

1871-1935
Pitcher Stamina: Very High
Use Relievers: Normal
Use Closers: Very Rarely
PH for Pitchers: Very Rarely

1936-1960
Pitcher Stamina: High
Use Relievers: Normal
Use Closers: Rarely
PH for Pitchers: Rarely

1961-1976
Pitcher Stamina: High
Use Relievers: Normal
Use Closers: Sometimes
PH for Pitchers: Normal

1977-1992
Pitcher Stamina: High
Use Relievers: Normal
Use Closers: Often
PH for Pitchers: Normal

1993-2019
Pitcher Stamina: High
Use Relievers: Normal
Use Closers: Very Often
PH for Pitchers: Normal

[Garlon]

This is Mariano Rivera from an 1871-2019 game from today.

[ALB123]

Originally Posted by joefromchicago

Garlon's point about the league stamina setting is, I think, correct. As I mentioned earlier in this thread, there are two stamina ratings and together they measure two different things. There's the league stamina setting (found in the league's Stats & AI page) and each pitcher's individual stamina rating. Those, in turn, determine both in-game stamina (how many pitches a pitcher can throw before getting tired) and between-game stamina (how long a pitcher takes before recovering fatigue).

I'm not entirely sure how the stamina ratings work together, but I think the league setting establishes the "scale" and the individual pitcher ratings are a percentage of that scale. So, to use some made-up numbers, if the league setting is "high," and that setting establishes a scale of 1-100, then a pitcher with a 75/100 stamina rating would have a 75 "adjusted" stamina. If, on the other hand, the scale was 1-50, that pitcher would have a 37.5 "adjusted" stamina.

In your example, I don't know what the league stamina setting is, but my guess is that it's either "low" or "very low." In the replays that I've done in conjunction with my exploration of pitcher usage in the pre-reliever era, league settings have been set at "high." At a high setting, a pitcher like Martinez with a 65/80 stamina who threw 112 pitches in a game would probably be back to 100% within four days.

With regard to relievers, I don't know how quickly they recover or what their effectiveness would be if they pitch when they're less than 100% rested. Not surprisingly, relievers haven't been my focus when I've been looking at the pre-reliever era. I can say, however, that, in my experience, the AI is very reluctant to pitch anybody who is less than 100% rested - that's true for both starters and relievers.

At times, that leads to some very strange results. For instance, in my replays I try to keep rosters at more-or-less historical levels. That means that sometimes there may be only one or two pitchers in the bullpen. If those guys aren't 100% rested, the AI will simply stay with the starter, even though it might be better to pinch-hit for him. The AI sees the bullpen pitchers as unavailable because they're not fully rested, and it sees the other starters in the rotation as unavailable because they're either fatigued or else because the AI just doesn't like using starters in relief. As a result, the starter will stay in the game and throw an ungodly number of pitches. In my 1912 Senators replay, for instance, it wasn't unusual for Walter Johnson to throw over 200 pitches in a game.

That's not to say that starters in that era didn't throw a lot of pitches - they did. But the AI logic is premised on a strict split between starters and relievers, and it's very difficult for a pitcher to straddle the line between the two roles. Pitchers, however, did that all the time in the pre-reliever era. In that sense, the real-life managers had more options available to them than the AI allows them. If a starter like Walter Johnson got tired, his manager, Clark Griffith, would view every pitcher on the bench as a possible reliever. In OOTP, on the other hand, the AI looks first at the pitchers who aren't in the rotation. If they're unavailable, then the AI far too often is stumped. It won't use a top-end starter who is even minimally fatigued and it won't use a bottom-end starter because they're usually not very good. Griffith, in contrast, would have taken the game situation into context and said: "if it's close, I'll use my best pitcher, unless he's completely exhausted. If it's not close, any of those knuckleheads at the end of the bench will do." Having the AI assign secondary roles to pitchers might, I think, be one important way to address this problem.

Joe, I believe there are more than two settings for pitcher stamina. The two you mentioned, a pitcher's personal rating (ex. 25/80 or 50/80), the general strategic tendencies found on the left-side of the Stats & AI page (ex. Low, Normal, High) and then on the right-side of Stats & AI you have the Starting Pitcher Stamina and Relief Pitcher Stamina modifiers which change every season. For 1995, my current season, SP Stamina is set to 1.151 while RP Stamina is set to 1.000. I've only paid attention for a few seasons, but the RP Stamina modifier has stayed at 1.000 for multiple seasons.

Originally Posted by scurvycure

This discussion is much-appreciated. I have started to check on the rest status of every team's pitching staff in my replay before each game. Anecdotally, it seems like bringing in a late inning reliever who is not completely rested is, more often than not, like throwing gas on a fire. Walk, single, wild pitch, walk, single, single. Man that is painful. As double headers, injuries, rain delays and extra inning games wear down a staff, I find too many games where a team is starting a tired starter, or has only one or fewer rested relievers. I decided to increase my league's roster size to ensure that every team had 12 man pitching staffs. That isn't realistic for 1972 but I would rather have that than unrealistic run production caused in part by pitcher fatigue.

I started peeking at other pitching staffs belonging to the last-place teams. Imagine my surprise when I see the next scheduled SP at 68% rested and/or exhausted bullpens. I couldn't believe it! About 5 seasons ago, I too increased the MLB Active Roster size limit to 27 and forced the AI teams to make at least one them a relief pitcher. Adding just one additional reliever often made tremendous differences in the rest status of their bullpen. My team, of course, was forbidden to exceed 25 men on the Active Roster.

Originally Posted by joefromchicago

What settings are you using for rotation size and pitcher usage (strict order, start highest rested)? What are your settings for reliever usage?

I know this question was meant for scurvycure, but I'm going to butt in anyway...

I set the league to use a 5-man rotation and strict order.

Hook for SP is +4. Hook for RP is -5 (Very Quick). Use of Relievers is Very Often. Use of Closers is Very Often. Pitcher Stamina is Low. These settings are all defaults for 1995 in ootp.

Originally Posted by Garlon

I am currently testing some pitcher settings. I do not think that Use Relievers should ever be changed from Normal, only the Pitcher Stamina, Use Closers and PH for Pitchers should be adjusted for different eras.

1871-1935
Pitcher Stamina: Very High
Use Relievers: Normal
Use Closers: Very Rarely
PH for Pitchers: Very Rarely

1936-1960
Pitcher Stamina: High
Use Relievers: Normal
Use Closers: Rarely
PH for Pitchers: Rarely

1961-1976
Pitcher Stamina: High
Use Relievers: Normal
Use Closers: Sometimes
PH for Pitchers: Normal

1977-1992
Pitcher Stamina: High
Use Relievers: Normal
Use Closers: Often
PH for Pitchers: Normal

1993-2019
Pitcher Stamina: High
Use Relievers: Normal
Use Closers: Very Often
PH for Pitchers: Normal

Wow! If I try these settings, which I am going to do, I'll be using the last set, 1993 to 2019. Would you mind explaining how you came to the decision to suggest Pitcher Stamina is set to High? The default is Low, so I am surprised a bit. Yes, I became involved with this thread because I have a personal gripe with the issue of pitcher stamina in ootp, but I don't like to change things just to make things different. If possible, I like to know the theory or reasoning behind the suggestion. I'm sure you know what I mean.

I'm just as curious about your suggestion to always have Use Relievers set to Normal.

Originally Posted by Garlon

This is Mariano Rivera from an 1871-2019 game from today.

Are you saying that you just had ootp simulate every year from 1871 to 2019 in one uninterrupted run? I don't even know how to do that...Is it possible to quickly explain how to do that so there aren't any interruptions? My initial guess would be to create a new baseball universe, but begin without a job so every team is controlled by the AI. Or is it something completely different?

[joefromchicago]

Originally Posted by ALB123

Joe, I believe there are more than two settings for pitcher stamina. The two you mentioned, a pitcher's personal rating (ex. 25/80 or 50/80), the general strategic tendencies found on the left-side of the Stats & AI page (ex. Low, Normal, High) and then on the right-side of Stats & AI you have the Starting Pitcher Stamina and Relief Pitcher Stamina modifiers which change every season. For 1995, my current season, SP Stamina is set to 1.151 while RP Stamina is set to 1.000. I've only paid attention for a few seasons, but the RP Stamina modifier has stayed at 1.000 for multiple seasons.

Yep, forgot about that. That modifier, I believe, is applied to the "scale" that is established by the league stamina setting. But I could be wrong.

Originally Posted by ALB123

I know this question was meant for scurvycure, but I'm going to butt in anyway...

I set the league to use a 5-man rotation and strict order.

Hook for SP is +4. Hook for RP is -5 (Very Quick). Use of Relievers is Very Often. Use of Closers is Very Often. Pitcher Stamina is Low. These settings are all defaults for 1995 in ootp.

I would be very hesitant to use the "strict order" setting for any league in any era, in large part because it is extremely rare to find any team that actually did use a strict rotation for an entire season. In 1995, the Braves came very close to having a strict five-man rotation (Maddux-Glavine-Avery-Mercker-Smoltz), but that was the result of having some exceptionally durable pitchers and a schedule that had no double-headers and no rainouts.

No other team that year could boast such a stable rotation, which is one reason why I'd probably set the rotations in 1995 to "strict order, occasionally highest rested." It is, I think, more historically accurate. The other reason is that adhering to a strict rotation can lead to some pretty bad results if you have any injuries or double-headers. Taking away a rest day will lead to a cascading series of exhausted starters who never have a chance to get fully rested until there's an off-day.

I don't know what the "low" setting for league stamina means in terms of "between-game" stamina, but if your starters are constantly tired, then you either need to raise the league stamina setting, go to six-man rotations, or change the rotation setting (or a combination thereof). By 1995, teams were routinely using five-man rotations, so they should work in OOTP. Another solution would be to assign some of your relievers a secondary role of "emergency starter." That would help solve the problem for your team, but, as I've mentioned elsewhere in this thread, the AI teams will never give any pitcher a secondary role.

[Garlon]

Recently I have been doing more testing with the game. I was noticing that top SP strikeout totals were not where they should be with nearly no SP making it to 300 k in a season. This setting for Stamina, though it seems odd, seems to allow the better strikeout pitchers to strikeout more batters, apparently keeping their stuff for longer in games. The league totals modifier for starting pitcher stamina will still bring things into place for the entire league in terms on innings pitched for starting pitchers.

As for the 1871-2019 game, I start in 1871 with 4 teams, then I add 4 more teams in 1876, 1882, and 1892. So after 1892 I have the same 16 teams as there were in 1901. When doing this I assign the proper team and franchise abbreviation to the teams as they were in 1901. At the beginning of the 1900 season I then turn on the option to automatically expand the league according to history, and in 1901 the game starts using the real schedules and logos and will carry out the expansion properly in subsequent seasons.

Setting Relievers to Normal seems to be the best thing to do for any season. The league totals modifier for starting pitcher stamina distributes the IP correctly between the SP and RP, so if you set Use Relievers to anything above Normal apparently you are just telling the game to make more pitching changes among relievers. If you set this to anything below Normal the game will be reluctant to switch relievers when it should. I have found that keeping that at Normal seems to produce better results. In fact if you increase Use Relievers and Use Closers together you actually get fewer saves from your closers.

These are the settings I am currently using for different seasons. Joe will disagree with my use of 4-man strict rotation from 1903-1935. I was trying 5-man highest rested for 1920-1935 today, but I am not convinced with the results yet. I want to use 5-man rotation but doing so has an impact on Ruth HR totals unfortunately. This is because there are now a weaker set of pitchers starting, making it easier for the rest of the league to steal some HRs away from Ruth. I may try again though.

Even though I do not want to use a roster of more than 25 players, right now due to problems with the game the computer is tending to leave some substitution slots empty at some positions if you use lower than 16 position players. This still can happen even with 16 position players but not as often. When this happens the starting position player will play every game of the season and stay fatigued and end up batting about .140 with 275 strikeouts for the season.

Beginning in 1993 you see that the average team was getting more than 400 relief appearances from their relievers. So you need to consider how many appearances you are expecting from each reliever. If you try to play 1997 with 5 relievers, they each need to make over 90 appearances and they will not perform well because of that.

[joefromchicago]

Originally Posted by Garlon

Joe will disagree with my use of 4-man strict rotation from 1903-1935.

Well, yeah. I can't see using strict order as a rotation setting before around 1970, and even then I'd probably go with "strict order, occasionally highest rested."

But I think our divergent views on this point result from our different objectives. I think your focus is to make sure that the stats are distributed in a historically accurate way. I'm more focused on trying to have the players play in a historically accurate way. I think both perspectives are equally valid and equally valuable. Ideally, OOTP would satisfy both of us, with players getting historically accurate stats in a historically accurate playing environment.

I'll add that I'm probably less concerned about insuring that the stats shake out in an accurate manner because my ultimate goal isn't so much to improve historical leagues as it is to improve fictional leagues. But, as I've explained in the past, the only way to make sure that fictional leagues are historically accurate is to make sure that historical leagues are historically accurate.

Originally Posted by Garlon

Beginning in 1993 you see that the average team was getting more than 400 relief appearances from their relievers. So you need to consider how many appearances you are expecting from each reliever. If you try to play 1997 with 5 relievers, they each need to make over 90 appearances and they will not perform well because of that.

In 1997 the average team used 19 pitchers. So it would be highly unlikely that a team would go through a season using only five relievers. The Dodgers used the fewest pitchers that season (16), and they still had 11 relievers who made a total of 407 relief appearances. I'd say that a staff of five starters and six relievers would be about average for a team in 1997.

[scurvycure]

Originally Posted by joefromchicago

What settings are you using for rotation size and pitcher usage (strict order, start highest rested)? What are your settings for reliever usage?

I am using 5 man rotations with most teams set to Strict, occasionally highest rested. Use of relievers is set to normal.

Other settings: Hook starter -2 (quick); Hook reliever +1 (I'm not sure what those mean or what impact they have, but those were the settings the game implemented when I ran my initial setup.)

Pitcher stamina Normal, PH for pitchers Normal
Starting pitcher stamina = 1.013 and RP stamina is 1.000


I will add that I started paying close attention to this beginning at the late-June early July stage of the schedule. Teams are making their second swing of road trips to opposing teams and so any early season rainouts are showing up now as double-headers and/or fewer off days. I did notice that the couple teams who are set to use strict rotations are facing issues with tired starters, so I will have to change that setting.


I've also realized how much it pays to have at least one or two higher stamina guys in the bullpen for those occasions when you need someone to just eat up some innings or make a spot start. And I like ALB's suggestion that even though I increased roster limits to allow teams to maintain 12 man staffs, the team I am managing every day should have to play at the original 25 man roster limit.

[ALB123]

Originally Posted by joefromchicago

I would be very hesitant to use the "strict order" setting for any league in any era, in large part because it is extremely rare to find any team that actually did use a strict rotation for an entire season. In 1995, the Braves came very close to having a strict five-man rotation (Maddux-Glavine-Avery-Mercker-Smoltz), but that was the result of having some exceptionally durable pitchers and a schedule that had no double-headers and no rainouts.

No other team that year could boast such a stable rotation, which is one reason why I'd probably set the rotations in 1995 to "strict order, occasionally highest rested." It is, I think, more historically accurate. The other reason is that adhering to a strict rotation can lead to some pretty bad results if you have any injuries or double-headers. Taking away a rest day will lead to a cascading series of exhausted starters who never have a chance to get fully rested until there's an off-day.

One reason I was setting the rotation to Strict Order was that I have the luxury of playing a game and then looking over all of my rosters quickly to see if anyone got hurt. With a 5-man rotation, if pitcher 3 stubs his toe and needs 10 days off, I get that #3 spot filled by the end of the very next day, no waiting for his normal position to be due to pitch again. Everything keeps running as nothing happened.

The bigger reason why I was using Strict Order was that in my MLB world I have been using OOTP Default settings for Stamina, as I've already mentioned...and with a 5-man strict rotation, let's say after the first 5 games, Pitcher 1 is next to start and then P2, P3, etc... Pitcher 1 will be at 100% rested just in time to take the mound. P2 will always be somewhere between 75-80% rested. P3 will be lower, etc... So, in my mind the "Highest Rested" will always be the pitcher due to pitch because my 5-man rotation will always have 5 pitchers listed.

As you've pointed out, and I feel foolish for not thinking of this myself, is that the AI doesn't always work that way. I am certain that by choosing Strict Order I have screwed up the AI rotation several times. I feel terrible that I now know it's mainly my fault, probably in most of those cases. I'll be making that change immediately and hopefully, that will limit AI pitching staff disasters. I am going to also increase Pitcher Stamina in the General Strategic Tendencies to Normal (maybe even High) which should also help the AI quite a bit.

Originally Posted by Garlon

Setting Relievers to Normal seems to be the best thing to do for any season. The league totals modifier for starting pitcher stamina distributes the IP correctly between the SP and RP, so if you set Use Relievers to anything above Normal apparently you are just telling the game to make more pitching changes among relievers. If you set this to anything below Normal the game will be reluctant to switch relievers when it should. I have found that keeping that at Normal seems to produce better results. In fact if you increase Use Relievers and Use Closers together you actually get fewer saves from your closers.

These are the settings I am currently using for different seasons. Joe will disagree with my use of 4-man strict rotation from 1903-1935. I was trying 5-man highest rested for 1920-1935 today, but I am not convinced with the results yet. I want to use 5-man rotation but doing so has an impact on Ruth HR totals unfortunately. This is because there are now a weaker set of pitchers starting, making it easier for the rest of the league to steal some HRs away from Ruth. I may try again though.

Even though I do not want to use a roster of more than 25 players, right now due to problems with the game the computer is tending to leave some substitution slots empty at some positions if you use lower than 16 position players. This still can happen even with 16 position players but not as often. When this happens the starting position player will play every game of the season and stay fatigued and end up batting about .140 with 275 strikeouts for the season.

Thank you for sharing that chart. I downloaded the image for safekeeping.

Once again, I feel a bit foolish for not recognizing what you mentioned about setting Use Relievers above Normal and especially the combination of Use Relievers and Use Closers long ago. Well, I have noticed it, in a way, but it didn't register that it was a settings issue. For a few seasons now, I've noticed that Save totals were down throughout the MLB and just shrugged my shoulders.

Originally Posted by joefromchicago

In 1997 the average team used 19 pitchers. So it would be highly unlikely that a team would go through a season using only five relievers. The Dodgers used the fewest pitchers that season (16), and they still had 11 relievers who made a total of 407 relief appearances. I'd say that a staff of five starters and six relievers would be about average for a team in 1997.

I would say 6 relievers in 1997 as well. That's what OOTP defaults to in 1995 and I don't think there were really any philosophical changes as to how to best build a 25-man roster in the next 2 years.

Speaking of...I might just have terrible Google Fu, but at some point in every ootp season I say to myself, "Hmmm...I'd like to see what the Yankees opening day 25-man roster was this season..." and I can almost never find that information. 40-man roster? Yes...All MLB Yankees that year? Yup. But the basic 25-man I can't usually find it.

Originally Posted by scurvycure

I am using 5 man rotations with most teams set to Strict, occasionally highest rested. Use of relievers is set to normal.

Other settings: Hook starter -2 (quick); Hook reliever +1 (I'm not sure what those mean or what impact they have, but those were the settings the game implemented when I ran my initial setup.)

Pitcher stamina Normal, PH for pitchers Normal
Starting pitcher stamina = 1.013 and RP stamina is 1.000


I will add that I started paying close attention to this beginning at the late-June early July stage of the schedule. Teams are making their second swing of road trips to opposing teams and so any early season rainouts are showing up now as double-headers and/or fewer off days. I did notice that the couple teams who are set to use strict rotations are facing issues with tired starters, so I will have to change that setting.


I've also realized how much it pays to have at least one or two higher stamina guys in the bullpen for those occasions when you need someone to just eat up some innings or make a spot start. And I like ALB's suggestion that even though I increased roster limits to allow teams to maintain 12 man staffs, the team I am managing every day should have to play at the original 25 man roster limit.

What season are you currently playing, scurvycure?

[scurvycure]

Originally Posted by ALB123

What season are you currently playing, scurvycure?

Sorry, this is 1972.



By the way, one team I am kind of fascinated by from that season is the White Sox. Their real life rotation starts are pretty unique relative to all other teams: Wood - 49 starts, Bahnsen - 41, Bradley - 40, Lemonds - 18, Fisher - 4, and 2 others with 1 start. That means Wood, Bahnsen and Bradley started 84% of the team's starts that year. I set them up to use a 4 man rotation, Always Start the Highest Rested.

Once I started playing, it didn't take long to realize that a 4 man rotation setting wasn't going to work because all of the starters were often tired when their turn came up. I changed them to use a 5 man rotation but kept the Always Start Highest Rested. I am currently 79 games into the season and Wood, Bahnsen and Bradley all have made 16 starts (61% of team's starts). Their 4th starter has made 9 starts and two others have made 1 each.

Small sample size I realize, but overall that setting is doing better than what I expected in replicating how their starters were used in real life. I am curious to hear what other ideas anyone has to recreate unusual usage like that?

And in general, for any team that had one or two pitchers who started at an every 4th day frequency, but all other starters have a lower start frequency, is the Always Start Highest Rested the correct setting to use to try to replicate that in OOTP?

Edit:
"I am currently 79 games into the season and Wood, Bahnsen and Bradley all have made 16 starts (61% of team's starts). Their 4th starter has made 9 starts and two others have made 1 each." Oops! Major error on my part as I was only looking at active roster stats. Here are the corrected start totals 79 games into the season:
Wood, Bahnsen, Bradley - 16 each
Lemonds -10
Baldwin -9
Geddes-6
Frailing-3
two others with 1 each
So on second thought I have to say that those settings have not really done a good job of replicating the CWS real life usage for that season.

Closer examination reveals that in real life Wood averaged 7-2/3 inn per start (376 inn, 49 starts); Bradley averaged 6.5 inn/start; Bahnsen averaged a little over 6 inn in his 41 starts (he also pitched 2 games in relief for a total of 3-2/3 inn). In my OOTP replay, Wood is averaging 8.125 inn/start; Bradley is averaging 7 inn/start; Bahnsen is averaging 7.4 inn/start.

The thought occurs to me that, if the manager was to pull those pitchers a little earlier in every game, perhaps it would increase the chances they would be rested a day earlier and thus, when using the Always Start Highest Rested setting, more likely to get closer to the starting frequency Billy Martin used in real life.

One further note, the real life White Sox as a team had a total of 394 innings pitched in relief total for the season. In my replay they are at 167 inn pitched in relief at roughly the halfway point of the season. Again, small sample size, but perhaps further support for the proposition that starters are pitching too far into games to be able to replicate their real life starts that year.