Game not starting best 100% rested SP

[ColumbusJets]

1962 in OOTP 26. First time I've played 26. Didn't see this in 25.

Game defaults to 5 man rotations as in 25. In 26 my SPs are rested after 3 days. Game insists on starting #5 SP even when better ones are 100% rested.

I could change the game to 4 man rotations but fear that when some double headers are scheduled it would start 65% rested SPs in the rotation rather than the rested emergency SP in the bullpen. I don't want to disadvantage computer controlled teams.

[uruguru]

Originally Posted by ColumbusJets

1962 in OOTP 26. First time I've played 26. Didn't see this in 25.

Game defaults to 5 man rotations as in 25. In 26 my SPs are rested after 3 days. Game insists on starting #5 SP even when better ones are 100% rested.

I could change the game to 4 man rotations but fear that when some double headers are scheduled it would start 65% rested SPs in the rotation rather than the rested emergency SP in the bullpen. I don't want to disadvantage computer controlled teams.

There was a change in 26 that allowed pitchers in 5-man rotations to pitch on 3 days rest without penalty. However, when deciding "most rested", the game doesn't just look at the %, but at the pitch counts over the past 5 days. This ensures that 5-man rotations pitch properly (5-man) despite the rest change. But if you want to manually run a 4-man rotation in a 5-man league, you can do this. Which is historically accurate.

If you change to 4-man rotations, starters will reach 100% rest a day quicker, so this effect won't go away.

[ColumbusJets]

Originally Posted by uruguru

There was a change in 26 that allowed pitchers in 5-man rotations to pitch on 3 days rest without penalty. However, when deciding "most rested", the game doesn't just look at the %, but at the pitch counts over the past 5 days. This ensures that 5-man rotations pitch properly (5-man) despite the rest change. But if you want to manually run a 4-man rotation in a 5-man league, you can do this. Which is historically accurate.

If you change to 4-man rotations, starters will reach 100% rest a day quicker, so this effect won't go away.

That explains why the game would choose Joe Gibbon 100% rested 97 pitches 5 days ago over Sandy Koufax 100% rested 106 pitches 4 days ago. The game doesn't consider all 100%s to be equal.

I don't want to adopt a strategy for myself that the computer teams aren't capable of using. Checking further.

To clarify, before 26, when two pitchers were both 100% rested, the game would start the better one, right?

[kq76]

It'd be nice if there were more options in that dropdown, and it was clarified what each selection meant.

I decide who's starting every game and I've noticed the game definitely doesn't prepopulate the highest ranked rested pitcher in the rotation when "always start highest rested" is selected. For example, for my next game it wants to start SP Horie, who is quite a good pitcher with a 70 overall rating, #2 in my rotation, and he last pitched 95 pitches 5 days ago so he's 100% fully rested, but SP Ogawa is also 100% fully rested, having also pitched 95 pitches, albeit only 4 days ago (note that I'm playing a fictional league so you might not get similar rest), and he's #1 in my rotation with a 74 overall. So if Ogawa is #1 in my rotation and he's also fully rested, why doesn't OOTP want to pitch him? Well, the "Next Starter" is set as Horie. If I change next starter to someone else, then OOTP will want to start whoever that is, even if they started just yesterday.

So how is "Next Starter" being determined? I don't know. I compared them over maybe 20 stats and the only one that Horie is better at than Ogawa is wins and I refuse to believe the game is basing its decision on that. Maybe it's looking for the highest rested SP who started 5 or more days ago? If that's it, I don't have a problem with that reasoning as the more rest they get probably the more likely they'll perform better (and it may also lessen the risk of injury), but I just don't think that's what the user probably assumes when they see, "always start highest rested".

And "highest rested" definitely doesn't mean "most rested" as I have another fully capable SP, Kageyama, #4 in my rotation, who last pitched more than 5 days ago and he's not selected either.

It doesn't really bother me anymore because I choose who's starting every game now. But for people who are relying on that pull down, I'm sure it'd be nice if there were more options and more clarity to them.

So what do I currently go with? I simply look for the SP who hasn't pitched in the longest time. And if there's a tie for that then I choose the best SP (EDIT: And by best I just go off of SIERA, or overall rating if it's early). But I also might skip the SP with the most rest if a much better SP also last pitched 5 days ago or more. Or I might stick with the lesser SP if the opponent is a weak team. I will sometimes start a SP on 4 days rest, but I try not to and I usually have a capable SP in my bullpen. So really, I play it by ear, but I definitely try to ensure that whoever starts has had plenty of rest.

It'd be interesting to hear the logic others use. OOTP might even consider adding them to the dropdown one day.

[uruguru]

Originally Posted by ColumbusJets

I don't want to adopt a strategy for myself that the computer teams aren't capable of using. Checking further.

I only use it for my knuckleballers. If he's 100% and the AI wants to start a pitcher below him in the rotation, I switch it to knuckleballer.

Doesn't make much difference (a little) unless the knuckler is good enough to be your #1 starter. Then suddenly he's getting about 8 additional starts a year, like Wilbur Wood did when he was the #1 starter for the White Sox.

[ColumbusJets]

Relates to 1961 and similar seasons. It looks like the way to more starts from the best SPs and fewer lesser ones without a lot of manual changes is to not use the default settings of 5 man, highest rested, SPs in relief NO. That will get about 35 starts for the top of the rotation.

First screen shot is historic 1961 Dodgers

Second screen shot is 5 man, highest rested, in relief NO.

Third screen shot is 4 man, highest rested, in relief YES. That's a full YES not "in critical situations."

So I think the computer teams would get a better mix of SPs this way than with the default and the human player could manually give himself a better mix than the computer is giving him with the default.

[ColumbusJets]

Originally Posted by uruguru

I only use it for my knuckleballers. If he's 100% and the AI wants to start a pitcher below him in the rotation, I switch it to knuckleballer.

Doesn't make much difference (a little) unless the knuckler is good enough to be your #1 starter. Then suddenly he's getting about 8 additional starts a year, like Wilbur Wood did when he was the #1 starter for the White Sox.

It's clear that with the defaults the computer isn't managing pitching the way a reasonable human would. Bottom line is that the 5th starter gets too many starts. And I seriously doubt the game gives a pitcher any additional rating for being "super rested", that is, an amount greater than the 100% shown.

[uruguru]

Originally Posted by ColumbusJets

It's clear that with the defaults the computer isn't managing pitching the way a reasonable human would. Bottom line is that the 5th starter gets too many starts. And I seriously doubt the game gives a pitcher any additional rating for being "super rested", that is, an amount greater than the 100% shown.

I'm also confident the pitcher is not getting any additional rating for being "super rested"

So why have it in the game? It essentially guarantees that, barring injury, there should always been two starters that are 100% rested.

Why is that good? Because the game doesn't pull spot starters out of the pen when doubleheaders are coming. If a DH happens and no one in the bullpen is rested, then there is still a starter for each game.

[ColumbusJets]

Game is set to 5 man rotation. When the 1962 WS came around, the computer switched the rotations to 4 man and stuck with it even though the game 1 starter could have started games 4 and 7. There's a disconnect here. The game makes SPs 100% rested after 3 days but makes them wait 4 days to pitch.

[uruguru]

Originally Posted by ColumbusJets

Game is set to 5 man rotation. When the 1962 WS came around, the computer switched the rotations to 4 man and stuck with it even though the game 1 starter could have started games 4 and 7. There's a disconnect here. The game makes SPs 100% rested after 3 days but makes them wait 4 days to pitch.

Yes, starting pitchers rotate just like in 25. That hasn't changed. They just reach 100% a day earlier.

[micropterus58]

Many versions ago you used to be able to set a pitcher to start a percentage of games. I used to use this as a spot starter. I don't remember any complaints whether this was a good approach or not. Maybe some else has some data. Maybe this could be implemented again. Maybe it caused other issues.

[ColumbusJets]

Originally Posted by uruguru

Yes, starting pitchers rotate just like in 25. That hasn't changed. They just reach 100% a day earlier.

Since the computer teams can't benefit from faster return to 100% it seems unfair for me to do it.

[uruguru]

Originally Posted by ColumbusJets

Since the computer teams can't benefit from faster return to 100% it seems unfair for me to do it.

I agree. Like I said, I only use it when I have a knuckleball starter. If they are 100% rested, then I will start them before any pitcher below them in the rotation.

[ColumbusJets]

If the game selected the better of two 100% rested pitchers then it would mostly perform correctly with 5 man rotation when SPs are 100% after three days. The remaining problem - a lesser one - would be using the fifth SP as a RP when it was likely he wouldn't be needed as a SP before he was rested again.

[uruguru]

Originally Posted by ColumbusJets

If the game selected the better of two 100% rested pitchers then it would mostly perform correctly with 5 man rotation when SPs are 100% after three days. The remaining problem - a lesser one - would be using the fifth SP as a RP when it was likely he wouldn't be needed as a SP before he was rested again.

The AI doesn't really "look ahead" at the upcoming schedule to tune its rotation.

For example, if I see a double-header coming up, I will temporarily increase the rotation size to 6 three days in advance and toss the spot starter at #6 to ensure he's rested for his start in Game 2 of the DH.

Conversely, if there's a day off coming, I'll toss the #5 starter into the bullpen when appropriate, and then leave him in there until 3 days before he starts again. Sometimes he get innings and keeps the rest of the relievers from getting overworked (I play 5 starters, 5 relievers)

I do get enjoyment out of managing the pitching staff this way, including the part about being willing to start knucklers a day early sometimes if they are 100%. It makes me feel like I'm actually running the team.

But if the AI was smart enough to handle the rotation properly, I would be fine with that.

[David Watts]

I might be wrong, but I still say the biggest flaw in the "always start the highest rested" feature is that it operates based on rotation size and not on the basis of what it's meant to do, that being always starting the highest rested pitcher.

If you have rotation size set to 5 pitchers, always start the highest rested will mainly look to skip the 5th starter when possible. If you set the rotation size to 6, it will look to skip skip the 6th starter. I don't think rotation size should factor into how "always start the highest rested" works period. I think the game should add another feature that tells the game how many starters we want the AI to prioritize in terms of getting the highest rested starts. If I'm playing during the dead ball period this might be the starters listed 1 and 2 on my staff, or maybe the top 3 starters. I feel this would work along with the ability to have starters pitch in relief to eliminate the need for game to have bullpens during the periods in baseball when teams didn't have them.

When playing 1901 right now I can get decent results if I set the game to a 5 man rotation, always start the highest rested, starters allowed in relief and set bullpen size to only 1 pitcher. That one pitcher in the bullpen will still get way too many appearances, but it's not nearly as bad if I have 2 or 3 pitchers in the bullpen.

Set it to 6 starters with zero pitchers in the bullpen and things get ugly fast. If a starter makes 30 total starts in a season it's a miracle. Start the highest rested needs to eliminate the rotations from the equation.

[ColumbusJets]

Is there a plausible explanation for the rest difference?

[ColumbusJets]

Originally Posted by David Watts

I might be wrong, but I still say the biggest flaw in the "always start the highest rested" feature is that it operates based on rotation size and not on the basis of what it's meant to do, that being always starting the highest rested pitcher.

If you have rotation size set to 5 pitchers, always start the highest rested will mainly look to skip the 5th starter when possible. If you set the rotation size to 6, it will look to skip skip the 6th starter. I don't think rotation size should factor into how "always start the highest rested" works period. I think the game should add another feature that tells the game how many starters we want the AI to prioritize in terms of getting the highest rested starts. If I'm playing during the dead ball period this might be the starters listed 1 and 2 on my staff, or maybe the top 3 starters. I feel this would work along with the ability to have starters pitch in relief to eliminate the need for game to have bullpens during the periods in baseball when teams didn't have them.

When playing 1901 right now I can get decent results if I set the game to a 5 man rotation, always start the highest rested, starters allowed in relief and set bullpen size to only 1 pitcher. That one pitcher in the bullpen will still get way too many appearances, but it's not nearly as bad if I have 2 or 3 pitchers in the bullpen.

Set it to 6 starters with zero pitchers in the bullpen and things get ugly fast. If a starter makes 30 total starts in a season it's a miracle. Start the highest rested needs to eliminate the rotations from the equation.

Yes, highest rested doesn't do what would be expected. Who could disagree that with equal rest the better pitcher should be selected?

[Charlie Hough]

So, what can we do to provide the development team with some good parameters that it can use to improve this? Can we come up with a definitive logic that could be programmed? Do we need different parameters or even a different option depending on historical era and rotation sizes and upcoming schedule?

The best way we can potentially get this fixed is to propose a solution that's as specific as possible and gives the developers the right parameters or requirements for the AI to use in making its decisions.

[uruguru]

Originally Posted by David Watts

I might be wrong, but I still say the biggest flaw in the "always start the highest rested" feature is that it operates based on rotation size and not on the basis of what it's meant to do, that being always starting the highest rested pitcher.

The term "highest rested" needs to be defined properly.

If your #1 starter threw 80 pitches 5 days ago, and your #5 starter threw 20 pitches in relief for that same game, they are BOTH 100% rested, which presumably means either can be started without a fatigue penalty.

But the highest rested pitcher is still your #5 starter.