AI Bullpen Logic

[kdutch98]

Re: AI Bullpen Logic

I tried this and the first reliever they put in was the 2nd setup spot (worst reliever) when they were up 2-0 in the 7th ��

It works better for simmed games.

In games I play, I move the worst rated reliever from the #2 set up spot, to the #2 LRP spot. Then once the series vs that team is over, I move him back to #2 SU

This does result in that reliever sometimes being put into the game, but normally only in "mop up" situations, where that team is either way ahead, or way behind

[Mephiskapheles]

Re: AI Bullpen Logic

It works better for simmed games.

In games I play, I move the worst rated reliever from the #2 set up spot, to the #2 LRP spot. Then once the series vs that team is over, I move him back to #2 SU

This does result in that reliever sometimes being put into the game, but normally only in "mop up" situations, where that team is either way ahead, or way behind

Who goes in the 2nd setup spot?

[kdutch98]

Re: AI Bullpen Logic

Who goes in the 2nd setup spot?

No one. Leave it blank for games you play.

That way, if there is a "set up" scenario in the game, the right guy will be used by the CPU.

[Cycloniac]

Re: AI Bullpen Logic

I tried this and the first reliever they put in was the 2nd setup spot (worst reliever) when they were up 2-0 in the 7th ��

Update:
Here's some box scores from some played games using the above setup, along with the bullpen setups themselves:

https://www.twitch.tv/videos/1814242572

I like the way everyone else was used. Maybe moving the SU2 to LRP2 would work better? Downside is the fact that LRP2 may get too many innings during simmed games.

[tytyboogie]

Re: AI Bullpen Logic

What number do you think works best?

ive used 4 then 7 and that works decent. but maybe 8 is the way to go. .so 1 other thing cuz the posts before me on this topic which has alawys been my lone gripe pretty much every year wishin it would be adressed and all im also tryin the 13 man with 2 lrp and 1 setup guy but last year i went with 12 mans and just 1 lrp and found way too many usages of them among others but my question is are you guys seeing in a played game the cpu put in the lrp2 slot guy at all? and during a series lets say they used lrp1 in the 1st game and hes maybe %50 will they be smart enough to now use lrp2 or they just goin right back to the same lrp 1 guy? just curious

[Ghost Of The Year]

Re: AI Bullpen Logic

While trying to get a better understanding of how the logic works in MLB23, I see that the Twins have Jhoan Duran listed as a Closer, yet the AI CPU slots him as the setup man, while it takes regular reliever Dylan Floro and designates him in the closer slot. That is all fine and dandy, I have no problem with the better pitcher closing games. HOWEVER, here's where the AI loses me. Floro's OverAll is 76 while Duran's is 83. That's not really close in my personal opinion. A couple points I can understand. But clearly Duran is the better pitcher, OverAll. Then I start looking at their Clutch rating, Duran 99, Floro 72. Again, not close. Then I look at further attributes:
Duran H9 90, K9 84, BB9 58, HR9 69 - that totals 301 ''points''.

Floro H9 72, K9 62, BB9 75, HR9 83 - 292 total ''points''.

What is at play here causing Floro to be a better option as closer? According to Floro's 2022 Show WAR stat, he earned a 1. Duran earned a 1.7. WAR (Baseball Reference gave Floro .8 WAR and Duran 2.8) Also, these are the last live rosters, which factor in the 2023 season. Duran WAR for '23 was 1.9 and Floro looks to be zero WAR.

Thank-you if anyone with a better grasp than me, knows what is at play here. Headscratcher.

[Ghost Of The Year]

Re: AI Bullpen Logic

While trying to get a better understanding of how the logic works in MLB23, I see that the Twins have Jhoan Duran listed as a Closer, yet the AI CPU slots him as the setup man, while it takes regular reliever Dylan Floro and designates him in the closer slot. That is all fine and dandy, I have no problem with the better pitcher closing games. HOWEVER, here's where the AI loses me. Floro's OverAll is 76 while Duran's is 83. That's not really close in my personal opinion. A couple points I can understand. But clearly Duran is the better pitcher, OverAll. Then I start looking at their Clutch rating, Duran 99, Floro 72. Again, not close. Then I look at further attributes:
Duran H9 90, K9 84, BB9 58, HR9 69 - that totals 301 ''points''.

Floro H9 72, K9 62, BB9 75, HR9 83 - 292 total ''points''.

What is at play here causing Floro to be a better option as closer? According to Floro's 2022 Show WAR stat, he earned a 1. Duran earned a 1.7. WAR (Baseball Reference gave Floro .8 WAR and Duran 2.8) Also, these are the last live rosters, which factor in the 2023 season. Duran WAR for '23 was 1.9 and Floro looks to be zero WAR.

Thank-you if anyone with a better grasp than me, knows what is at play here. Headscratcher.

Edit - one last thing I just thought to consider, as a Closer, Floro's OverAll is 78 and Duran's OverAll as a regular reliever is an 80, so that would narrow the gap if you view it that way. And stamina is a dead heat at 27.

[jcar0725]

Re: AI Bullpen Logic

While trying to get a better understanding of how the logic works in MLB23, I see that the Twins have Jhoan Duran listed as a Closer, yet the AI CPU slots him as the setup man, while it takes regular reliever Dylan Floro and designates him in the closer slot. That is all fine and dandy, I have no problem with the better pitcher closing games. HOWEVER, here's where the AI loses me. Floro's OverAll is 76 while Duran's is 83. That's not really close in my personal opinion. A couple points I can understand. But clearly Duran is the better pitcher, OverAll. Then I start looking at their Clutch rating, Duran 99, Floro 72. Again, not close. Then I look at further attributes:
Duran H9 90, K9 84, BB9 58, HR9 69 - that totals 301 ''points''.

Floro H9 72, K9 62, BB9 75, HR9 83 - 292 total ''points''.

What is at play here causing Floro to be a better option as closer. According to Floro's 2022 Show WAR stat, he earned a 1. Duran earned a 1.7. WAR (Baseball Reference gave Floro .8 WAR and Duran 2.8) Also, these are the last live rosters, which factor in the 2023 season. Duran WAR for '23 was 1.9 and Floro looks to be zero WAR.

Thank-you if anyone with a better grasp than me, knows what is at play here. Headscratcher.

It's often not very logical you really need 30 team control to fix each teams lineup and pitching staff. I'm annoyed that this is an ongoing issue but I'm grateful 30 team control is at least an option.

I've not seen your scenario very often but I have seen the cpu put its best RP in the closer role. Which I guess is fine I just usually edit his position to CP.

And don't get me started on the cpu placing guys with low overall rating on the 26 man roster and dropping higher rated players to the minors....simply because the lower rated players are out of options.

If I didn't have 30 team control I would get way too annoyed .

[Ghost Of The Year]

Re: AI Bullpen Logic

It's often not very logical you really need 30 team control to fix each teams lineup and pitching staff. I'm annoyed that this is an ongoing issue but I'm grateful 30 team control is at least an option.

I've not seen your scenario very often but I have seen the cpu put its best RP in the closer role. Which I guess is fine I just usually edit his position to CP.

And don't get me started on the cpu placing guys with low overall rating on the 26 man roster and dropping higher rated players to the minors....simply because the lower rated players are out of options.

If I didn't have 30 team control I would get way too annoyed .

It annoys me but I rather prefer to stay away from the 30 team control route. I'm just going to pretend Rocco Baldelli doesn't know what the hell he's doing. But thank-you JCar, I suspected it was flawed logic, but wanted to ask anyway.

[djflock]

Re: AI Bullpen Logic

While trying to get a better understanding of how the logic works in MLB23, I see that the Twins have Jhoan Duran listed as a Closer, yet the AI CPU slots him as the setup man, while it takes regular reliever Dylan Floro and designates him in the closer slot. That is all fine and dandy, I have no problem with the better pitcher closing games. HOWEVER, here's where the AI loses me. Floro's OverAll is 76 while Duran's is 83. That's not really close in my personal opinion. A couple points I can understand. But clearly Duran is the better pitcher, OverAll. Then I start looking at their Clutch rating, Duran 99, Floro 72. Again, not close. Then I look at further attributes:
Duran H9 90, K9 84, BB9 58, HR9 69 - that totals 301 ''points''.

Floro H9 72, K9 62, BB9 75, HR9 83 - 292 total ''points''.

What is at play here causing Floro to be a better option as closer? According to Floro's 2022 Show WAR stat, he earned a 1. Duran earned a 1.7. WAR (Baseball Reference gave Floro .8 WAR and Duran 2.8) Also, these are the last live rosters, which factor in the 2023 season. Duran WAR for '23 was 1.9 and Floro looks to be zero WAR.

Thank-you if anyone with a better grasp than me, knows what is at play here. Headscratcher.

So this is something i dont think is an issue. The cpu actually has some variance when it comes to the closer spot which is a good thing. I believe they added closer logic in as well where if you closer is performing bad they will put someone else in the role.

[djflock]

Re: AI Bullpen Logic

.

And don't get me started on the cpu placing guys with low overall rating on the 26 man roster and dropping higher rated players to the minors....simply because the lower rated players are out of options.

If I didn't have 30 team control I would get way too annoyed .

I dont mind this. I want the cpu to have have some sort of variance and not just do everything basic on a static rating.

[jcar0725]

Re: AI Bullpen Logic

I dont mind this. I want the cpu to have have some sort of variance and not just do everything basic on a static rating.

There is no logic to a team having a 50 rated RP on the 26 man roster while a 75 rated RP is sitting in AAA because the 50 rated RP is out of options.

[Ghost Of The Year]

Re: AI Bullpen Logic

So this is something i dont think is an issue. The cpu actually has some variance when it comes to the closer spot which is a good thing. I believe they added closer logic in as well where if you closer is performing bad they will put someone else in the role.

Yeah, I dont consider it a problem to the point I feel I need to change anything, I just enjoy understanding.

[DonkeyJote]

Re: AI Bullpen Logic

There is no logic to a team having a 50 rated RP on the 26 man roster while a 75 rated RP is sitting in AAA because the 50 rated RP is out of options.

Teams do carry worse players on their major league roster while sending better players to the minors because of options. Especially right out of Spring Training.

Sent from my Pixel 6 using Tapatalk

[jcar0725]

Re: AI Bullpen Logic

Teams do carry worse players on their major league roster while sending better players to the minors because of options. Especially right out of Spring Training.

Sent from my Pixel 6 using Tapatalk

100% true. I have to fix it for them. Otherwise, drives me nuts that cpu teams are not at full strength.

Unless you're talking about real major league teams. Then there is no way they would carry a guy with no ML pitching experience over an experienced fully healthy high rated player.