CPU Bullpen Management

[BrewCrew823]

Re: CPU Bullpen Management

I am curious if you noted same result with the two guys in LR slots being MR guys (i.e. 18-30 stamina)?

Some of the guys in the LR slot had stamina around 30 as they were MR guys in that spot. They did still go over two IP per appearance. At the same time it increased the other LR’s IP because he was well rested more often.

[BrewCrew823]

Re: CPU Bullpen Management

I want to say that with injuries near max PLUS the workaround I tried with manually injuring the most used MRP, the most pitchers I've seen the CPU use is around 15-20.

So that right there tells us without managing the CPU yourself there’s really no way to get that close. Rough guess that’s probably 100+ innings that have to go somewhere.

[sharkdaballer]

Re: CPU Bullpen Management

I think the bullpen is really only set up for 12 guys and not 13. I’m going to continue to tinker with the sliders and try to find something that works out. I don’t want to end up only carrying 12. I do use 30 team control in my franchise. Without babysitting the CPU teams
I don’t think you are going to any final results that will anyone happy.

I believe this is true. From what I've seen, the CPU will often call up a position player when a pitcher is injured and put on the IL, leaving the roster with 12 pitchers.

I think its also fair to point out the 3-batter-minimum rule hasn't been used irl other than spring training and teams' bullpen strategies were going to change this season anyway.

[garry1221]

Re: CPU Bullpen Management

👆🏼 Very true statement.

I simulated through one season 30 plus times last night. I had injuries turned off simply to make simulating go faster. I used all sorts of combinations of SP stamina, RP stamina and manager hook. I tried using two long relievers for each team then switched to two set up guys.

I found that manager hook really makes the biggest difference in how many innings your starter goes. I set it to ten for one of the sims and Brandon Bailey (Orioles long relief) and Trevor Williams (Pirates long relief) had over 120 appearances with 260 plus innings. It doesn’t really effect total innings of any relievers except long relief.

Overall two guys in the long relief spot causes mass problems. Those guys become so well rested that the CPU constantly goes to them. I’d gets results of those two spots having over 250 IP combined even if one of them had a stamina below 30 frequently.

Two guys in the set up spot works better when RP stamina is higher and one of the guys had a much better overall than the other. Some teams didn’t use set up 2 that often and I found cases with less than 8 holds for that guy.

I think the bullpen is really only set up for 12 guys and not 13. I’m going to continue to tinker with the sliders and try to find something that works out. I don’t want to end up only carrying 12. I do use 30 team control in my franchise. Without babysitting the CPU teams
I don’t think you are going to any final results that will anyone happy.

With injuries off, IP is greatly inflated across the board. Last season I simmed with injuries off a couple times before realizing the data was compromised due to no injuries. I let aches and pains occur in my tests. I'd only replay things if month long injuries happened. Just my .02 there.

I'm really starting to believe Hook is as much about SP performance as it is about a manager's trust in his 'pen. I've found the higher the hook, the earlier a SP will be pulled. Beyond box scores and end of season stats, I have no visual proof on that though. Hook at 0, i believe makes the decision completely based on situation, without any other factors. In my sims i've seen everything from getting shelled early and pulled to throwing a complete game.

[knich1]

Re: CPU Bullpen Management

This would be a really good time for someone from SDS to pop in here and explain the ratings and why we are facing the challenges we are with getting realistic stats...

[ninertravel]

Re: CPU Bullpen Management

holy crap I actually seen the CPU go for a 4 out save yesterday and for the first time EVER the home team used the closer in the 9th innings in a tied game. this is what managers actually do do! (the smart ones anyway)

why can't these sort of things happen more often?

I am sick to death of the LR guy coming in early regardless how tired they are. with the 3 batter rule too now I can't change hook them straight away if the CPU does it like I used to be able to.

we need a option to turn CPU BP decisions off so we can do them all and warm up who we want/. this would stop a lot of comes backs. and the reason people think their is a comeback code

[BrewCrew823]

Re: CPU Bullpen Management

With injuries off, IP is greatly inflated across the board. Last season I simmed with injuries off a couple times before realizing the data was compromised due to no injuries. I let aches and pains occur in my tests. I'd only replay things if month long injuries happened. Just my .02 there.

I'm really starting to believe Hook is as much about SP performance as it is about a manager's trust in his 'pen. I've found the higher the hook, the earlier a SP will be pulled. Beyond box scores and end of season stats, I have no visual proof on that though. Hook at 0, i believe makes the decision completely based on situation, without any other factors. In my sims i've seen everything from getting shelled early and pulled to throwing a complete game.

I wanted to see who the CPU selects to go with based on spot in the bullpen. You’re right, data without injury isn’t complete.

SP performance is definitely a huge factor. I’ve seen starters in the high 60s/low 70s get almost 200 innings at times. They had a really good ERA. The higher the hook the lower number of innings on the guys that suck. The number of complete games goes down with a higher hook too.

We probably could run circles around this. The other factor is what roster we are choosing to use. Everyone has their ratings systems for rosters and I’m sure that can change things. Until they choose to put some real work into franchise mode, this probably won’t get fixed. Stupid Diamond Dynasty.

[Unlucky 13]

Re: CPU Bullpen Management

holy crap I actually seen the CPU go for a 4 out save yesterday and for the first time EVER the home team used the closer in the 9th innings in a tied game. this is what managers actually do do! (the smart ones anyway)

why can't these sort of things happen more often?

I am sick to death of the LR guy coming in early regardless how tired they are. with the 3 batter rule too now I can't change hook them straight away if the CPU does it like I used to be able to.

we need a option to turn CPU BP decisions off so we can do them all and warm up who we want/. this would stop a lot of comes backs. and the reason people think their is a comeback code

I actually had the CPU bring in their closer with the score tied in the bottom of the 7th a few days ago. Their bullpen was depleted, and it was a logical move, but still very surprising and cool to see the game make it happen.

[baconbits11]

Re: CPU Bullpen Management

So whats the consensus on slider settings for this?

[Cycloniac]

Re: CPU Bullpen Management

So whats the consensus on slider settings for this?

This issue goes beyond slider settings

[The Kid 24]

Re: CPU Bullpen Management

This issue goes beyond slider settings

Do your sim sliders not yield good results anymore?

[Cycloniac]

Re: CPU Bullpen Management

Do your sim sliders not yield good results anymore?

They do. They give the best results the game will allow. I would just like to push the results a bit closer to real life bullpen inning distribution over a full season. Sliders alone can't do that.

Some teams are perfect. Others are not.

[The Kid 24]

Re: CPU Bullpen Management

So whats the consensus on slider settings for this?

I'm going with Cy's settings for playing games:
CPU SP Sta - 8
CPU RP Sta - 2
Manager Hook - 3

And also Cy's settings for simming games... I forget what the #'s are tho.

[tytyboogie]

Re: CPU Bullpen Management

They do. They give the best results the game will allow. I would just like to push the results a bit closer to real life bullpen inning distribution over a full season. Sliders alone can't do that.

Some teams are perfect. Others are not.

Sir can you speak to what you do beyond sliders as far as your BP slots and your RP's stamina ratings? Ive been so engulfed with more roster edits daily ive hardly dipped into actual franchise games. im 16/30 teams done with fangraphs data for actual pitches /speed/and proper usuage. im curious as to your slotting cuz i dont use SU2 at all and if you do that then you have LRP2 in use if you do 13 pitchers. ive been wanting to see if using 2 LRP slots helps with cutting down the CPUs overusage of LRP1 ...i thought i noticed this in my first simmed box score a day after just playing a team myself that had there LRP at half energy they chose LRP2 to use that day in there next game. I just hope the CPU doesnt alawys try to use LRP1 frist no matter what when they have a LRP2 full energy..(curious cuase in spring training the cpu would alawys use all LRPs first before putting in anyone else it was annoying hoping thats not the case in real games with multiple LRPs)And also whats your LRPs stamina ratings? ive been using 60 for LRp1 and 45 for LRP2.

[Cycloniac]

Re: CPU Bullpen Management

Sir can you speak to what you do beyond sliders as far as your BP slots and your RP's stamina ratings? Ive been so engulfed with more roster edits daily ive hardly dipped into actual franchise games. im 16/30 teams done with fangraphs data for actual pitches /speed/and proper usuage. im curious as to your slotting cuz i dont use SU2 at all and if you do that then you have LRP2 in use if you do 13 pitchers. ive been wanting to see if using 2 LRP slots helps with cutting down the CPUs overusage of LRP1 ...i thought i noticed this in my first simmed box score a day after just playing a team myself that had there LRP at half energy they chose LRP2 to use that day in there next game. I just hope the CPU doesnt alawys try to use LRP1 frist no matter what when they have a LRP2 full energy..(curious cuase in spring training the cpu would alawys use all LRPs first before putting in anyone else it was annoying hoping thats not the case in real games with multiple LRPs)And also whats your LRPs stamina ratings? ive been using 60 for LRp1 and 45 for LRP2.

I personally don't change the setups anymore.

I don't have the patience to micromanage the CPU to such a degree for an entire season, and I'm not yet convinced that putting their bullpen in a certain configuration would fix the underlying issue.