AI Uses Long Relievers Too Liberally

[realstar]

Long relievers are pitchers a manager puts at the back of the bullpen because they are either not so good or in a slump. They are supposed to be used WHEN THE TEAM IS AHEAD BY A LOT.

It's inappropriate to put in a long reliever because the team is behind, supposedly by enough that there's no chance of making a comeback. WTF? Talk about a morale killer.

Apparently AI thinks it's time to put in a long reliever whenever opposing team gets ahead by 3 or more runs. WRONG. Games can become sabotaged when poor relievers are put in too early.

Ex: Starter goes awry & allows 5 runs in the first 2 innings. In the 3rd inning, the starter gets in trouble again, so with one out & one runner on, a reliever is brought in. A long reliever? WRONG. That reliever allows the runner to score & then another run in the 4th inning. AI brings in another long reliever in the 5th & he allows 3 more runs to score. Now the score is 10 to 3. Meanwhile, our team scores 5 runs in the 7th inning. Runs that could've won the game if better relievers were brought in. That's called an AI SABOTAGE.
__________________________________________________

-------->>>>> ***NOTE: The new title of this post should be:

LESSONS LEARNED ABOUT LONG RELIEF ROLE. Because. . . . .

The OP has since re-evaluated the issue. That would be me. Consequently, this original post should be thought of as an INCORRECT WHINE. My bad. My apology for complaining before thoroughly thinking things through. However, this thread my still be used as a kind of tutorial. Because, as we see below, I learned how wrong I was & through that process have learned very important issues about managing bullpens & about how to better play OOTP as well. I believe this is actually an excellent example of the inherent usefulness of this forum & I thank all participants for their fine efforts to teach me how to be a better baseball manager.

[Brad K]

Can't quite tell when your team scored the first three runs. I'm assuming for the purposes of what follows that you're the home team down 5-0 in the top of the third. Since you say the starter is in trouble again I'm putting the base runner on second. From gregstoll.com, your win probability is 13%. If you're the visiting team with the same circumstances in the bottom of the third your win probability is 8%.

I don't see much reason to quarrel with AI choices here. I'm generally please with how the AI game day manager runs the game. And I'm playing v 11 which presumably isn't as good as the newer ones.

[NoOne]

what you are thinking of is a "avoid high leverage.". long relievers play in tight games quite often. the score isn't the determining factor of their use... timing is... early in the game? long relief - hence the use of "long" in the name, too.

they are supposed to be used when the SP doesn't make it to the typical point in a game you expect to use a MR. That situation may be behind by alot, tied or even ahead by alot... quite random for a long-reliever, actually. obviously a good manager with 2 viable long-relieve options will use the "better" one in tighter situations, unless they need work badly, and the weaker option when the game is out of reach early.

am i saying they are as good as the su/cl? nothing should lead to that assumption... but, they are definitely a different beast than a 1inning reliever. they typically have a 3rd or more pitch and definitely require more stamina.

i think you need to change who you are using for your LR... if you feel hopeless when the LR goes in, you need to invest more into that role, maybe... or pick talent differently for that role etc... SP prospects from AAA are awesome in this role, if they are nto yet fully ready for a starting gig in the majors. (2 developed and 1 half developed pitch or maybe slightly underdeveloped control, but enough for a reliever etc etc)

my worst bullpen pitchers are my 3rd MR even 2nd MR is subpar too on occasion... not my #1 long-relief guy. he is typicalyl a vey capable player becaue he's going to get 70-100ip for me. i don't want to march some chump out there for 70+innings. those 2nd/3rd MR guys get 25-50ip barring injuries.

try getting a more competent LR. if he's good, i even put "high leverage" as secondary role... i want the AI picking him over the 2nd/3rd MR in a MR situation that's tighter than normal, but still below my #1 MR/high lev guy (if LR has high lev, i make sure my #1 mr has high lev as secondary -- in addition to ability determining that role being assigned, of course).

you will likely never see a real life manager, a competent one, bring in a "better" reliever (su/cl best MR -figure top ~4) in the 2-4th inning... i'd even say through the fifth inning except in extreme rare cases of injury and fatigue (forced hand will occur eventually). in 162 games, you aren't blowing a good rested arm when you are down 3 or more early and the SP (for wathever reason doesn't matter) has left early.

the ai struggles with long-term thought on the bullpen. i think they 'have' to use position players more than they should. but, that's not a common occurence or anything... it takes a special situation for it to rear its ugly head.

i think it relates to how they have coded the days off for positon players... they count up time in a way that's not intuitive... e..g a position player's day's off shcouldn't be every X/x days. it should be X days since a day off. so pen ai should look ahead for when the next day off is... if long ways it takes a more conservative approach (on top of how it chooces fatigued pitchers and when)... you code a few different approaches and it should all be kosher.... expecially after how many years has this game been developed? this isn't year one anymore... you can't use that type of excuse, lol. handling the pen and days off for positon players should defintiely be better... there's defintiely room to improve, but in most instances the bullpen is handled mostly fine- or at least within reason for a video game.

[RchW]

I can only wish the AI would use long relievers. I end up with 2/3 of my bullpen overused and 1/3 never used, with no relationship to results. That means a guy with a 2+ WHIP gets used as if he's Mariano Rivera and others with great stats get 1 inning every 6-8 days

One thing for sure; don't assume what you see in your set up applies across the board. I just want the AI to actually use my RP in some way approaching their effectiveness.

[NoOne]

Originally Posted by RchW

I can only wish the AI would use long relievers. I end up with 2/3 of my bullpen overused and 1/3 never used, with no relationship to results. That means a guy with a 2+ WHIP gets used as if he's Mariano Rivera and others with great stats get 1 inning every 6-8 days

One thing for sure; don't assume what you see in your set up applies across the board. I just want the AI to actually use my RP in some way approaching their effectiveness.

short and sweet - worry less about he semantics of the roles and more about what results. if the words don't fit your perception of how a bullpen should be labeled, it shouldn't matter if you get results that match what you want. kiss principle is always best way to go: keep it simple, stupid. if it's not doing what you want, then change is required for different resuts.

try a different method of assigning roles in the bullpen. i find simpler is better, but sometimes not possible.

i avoid secondary roles as much as possible. long reliever will have one if he's better than atpyical LR... maybe my best MR (exclude lefty spec, i'll use that on occasion without worry, but not often with my "better" relievers. i don't want a top-4 arm being used for 1 out. - the manager's going to follow their sliders on this regardless, i'd wager. so, the less it happens the better, imo)

heh, remind me to mention - the manager could be a big reason for using the pen in ways you dislike. play with sliders and see if you can find a "template" for a coach that does what you want. (i just edit my coaches; i say jump, and they ask how high?)

i can definitely stratify my middle releivers as i want... i do have a problem when the setup guy i prefer is not shared by the ai (higher ratings on the one i don't like as much). my workaround is to put the ai preference as 8th or later and the one i like 7th, and he gets used more that way. typicaly i put both to 7th or later so that tehy get used more, overall.

i'll give my prefered setup guy a high lev secondary and none for the one i like less - just another option to test if the innings given out aren't what i want with the other setup of roles.

i tyicaly have 4 solid relievers. cl, 2su and 1 mr (plus a good LR but that's a different role, imo, so i get differetn pitchers that are more like starters, plus fills the emergency sp role too). the cl/2su guys are nearly equal, but the MR guy is typically a bit lower rated than the cl/2su.

anyway i only set 1 MR to use more often and the other 2 to use less often.

i won't use avoid high leverage unless he's a total dumpster fire. i will use high leverage on my cl, su su, and best MR, sometimes the LR too.... if i'm more conerned about # of saves, i'll not use high leverage on the closer.

when i have an awesome lr, i'll use them as the 3rd high leverage secondary role.

i defintely limit my 2 worst bullpen arm usage by doing this. when i have 5 SP that go 200innings each, roughly, the bottom 2 pitchers only get 20-30-40 innings each, barring injury. my closer will get 60-70some (rarely higher, almost never lower), and setup kinda ranges based on whether the guy i prefer is the better rated one (see above). the one i like can be as high as 80-90ip, but 60-80 is probably normal... and the 2nd setup guy is 60-70 typically. top middle reliever is 50-70ish. and my LR is 70-100. (with a strong SP and no injuries, figure the lower values more often than the higher values) i can make it work like clockwork as far as seasonal usage -- individual game use i try to ignore, lol. there's defintiely some issues there, but not game breaking by any means.

i think i suggested a depth chart fix for the bullpen.. the way it is now is sloppy.

a depth chart for each role and getting rid of most of the secondary nonsense would make it simpler and more intuitive. you could leave small stuff like emergency SP and lefty specialist roles for secondary to save space. the less common stuff can be left as secondary roles.

if they had an eye to the future they coded it in a way that makes the only change necessary is recoding the gathering of the info for the lists the AI uses. the rest of the process should be the exact same. *it shouldn't use the depth charts themselves, but pull the data to a "working" area (text arrays? whatever vocab that fits better). all that changes is how those arrays are populated, after that point it's using the same old logic .

[slugger922cubs]

I think theres a misinterpretation here between long-reliever and mop-up. "Long relievers" "bridge" and eat multiple innings when a starter comes out early to 6-9th innings or when the bullpen is tired and needs rest. He may also pitch in extra innings if the pen is getting exhausted. High stamina not enough skill to necessarily be in the rotation. Has nothing to do with leverage. Think 2016 playoffs Mike Montgomery.
"Mop-up" is often the low skilled last reliever in the bullpen. He comes in when the game is out of hand one way or the other to give everyone else a rest for higher leverage opportunities. He may pitch multiple innings he may not.

[realstar]

Originally Posted by Brad K

Can't quite tell when your team scored the first three runs. I'm assuming for the purposes of what follows that you're the home team down 5-0 in the top of the third. Since you say the starter is in trouble again I'm putting the base runner on second. From gregstoll.com, your win probability is 13%. If you're the visiting team with the same circumstances in the bottom of the third your win probability is 8%.

I don't see much reason to quarrel with AI choices here. I'm generally please with how the AI game day manager runs the game. And I'm playing v 11 which presumably isn't as good as the newer ones.

Sorry for the delay. I was away from this forum for the past year. I remember this game I wrote in about & it continues to happen on a regular basis, so it's worth noting again. The score was 5-2 in the 3rd when the long reliever was brought in. One run was scored by each team in the 3rd, so it was 6-3. in the top of the 4th. Yes we were the home team.

I think it's inappropriate to bring in weak pitchers in that situation. But as another person posted below, I should use better quality pitchers (from AAA) for long relief. So I will address that issue in reply to them. Thanks for your input.

[realstar]

I would just like to say first that most all of your (NoOne) comments here are excellent & much appreciated.

Originally Posted by NoOne

what you are thinking of is a "avoid high leverage.". long relievers play in tight games quite often. the score isn't the determining factor of their use... timing is... early in the game? long relief - hence the use of "long" in the name, too.

Obviously. But the 'avoid high leverage' option was new to me & I have since used it a lot. Actually I avoid using long relief entirely b/c I don't trust AI's use of it. However, as you stated further down in your post, there is more to long relief than meets the eye. So I am learning here, and I thank you for that.

i think you need to change who you are using for your LR... if you feel hopeless when the LR goes in, you need to invest more into that role, maybe... or pick talent differently for that role etc... SP prospects from AAA are awesome in this role, if they are nto yet fully ready for a starting gig in the majors. (2 developed and 1 half developed pitch or maybe slightly underdeveloped control, but enough for a reliever etc etc)

my worst bullpen pitchers are my 3rd MR even 2nd MR is subpar too on occasion... not my #1 long-relief guy. he is typicalyl a vey capable player becaue he's going to get 70-100ip for me. i don't want to march some chump out there for 70+innings. those 2nd/3rd MR guys get 25-50ip barring injuries.

try getting a more competent LR. if he's good, i even put "high leverage" as secondary role... i want the AI picking him over the 2nd/3rd MR in a MR situation that's tighter than normal, but still below my #1 MR/high lev guy (if LR has high lev, i make sure my #1 mr has high lev as secondary -- in addition to ability determining that role being assigned, of course).

you will likely never see a real life manager, a competent one, bring in a "better" reliever (su/cl best MR -figure top ~4) in the 2-4th inning... i'd even say through the fifth inning except in extreme rare cases of injury and fatigue (forced hand will occur eventually). in 162 games, you aren't blowing a good rested arm when you are down 3 or more early and the SP (for wathever reason doesn't matter) has left early.

the ai struggles with long-term thought on the bullpen.

That pretty much says it all. Using starting pitcher prospects for the long relief role instead of leaving them in AAA is the obvious choice & I should have thought of that before I whined about AI using weaker pitchers early on, in semi-close games. I see now, that basically AI is trying to tell me something: that I need better pitchers at the bottom of the pen. So the rule seems to be, if you assign any long relievers, you should be aware that AI will use them in situations where they need to be good pitchers. Long Relief therefore requires better pitchers than the bottom of the middle relief area, as you stated, and certainly better than Avoid High Leverage status. The mistake is assuming it's appropriate to put lesser pitchers in long relief. The way the bullpen is set up in OOTP lends itself to assuming that because long relief is at the bottom of the list. So that is the lesson learned. Just because it's listed last doesn't mean you should assume that's where to put your weakest pitchers. Excellent info in your posts. Thanks for your input.

[realstar]

Originally Posted by slugger922cubs

I think theres a misinterpretation here between long-reliever and mop-up. "Long relievers" "bridge" and eat multiple innings when a starter comes out early to 6-9th innings or when the bullpen is tired and needs rest. He may also pitch in extra innings if the pen is getting exhausted. High stamina not enough skill to necessarily be in the rotation. Has nothing to do with leverage. Think 2016 playoffs Mike Montgomery.
"Mop-up" is often the low skilled last reliever in the bullpen. He comes in when the game is out of hand one way or the other to give everyone else a rest for higher leverage opportunities. He may pitch multiple innings he may not.

That is an excellent point--use of the term "Mop-Up"-- and your statement is mostly correct. What you forgot to mention, is how this strategy is applied in OOTP. And I would argue that Long Reliever might need to be better than the bottom of the rotation. Furthermore, Mop-Up may, in some cases, be thought of as a synonym for OOTP's "Avoid High Leverage."

Long Reliever needs to have the quality & stamina of an extra starter (4th, 5th, 6th, 7th.. depending on the size of the rotation) and may be used as a starter replacement (called "Emergency SP" in OOTP) but who is also used in long relief situations. He therefore needs to be thought of as, not the Mop-Up pitcher, but as a pitcher who is as good, if not better, than the bottom of the rotation. In some respect, better, because he'll be brought in for difficult situations. (the standard of all relief pitchers)

(Note how the game offers Avoid High Leverage for Long Relief as well--THAT is the Mop-Up role for early games: several innings pitched. Where the Avoid High Leverage offered for Middle Relief is the Mop-Up role for mid/late games: in any case, less innings pitched.)

Whereas the bottom of the rotation may be used/or not (skipped) & if possible only used to start games against weaker opponents. So the starter at the bottom of the rotation could/should be a weaker (or even) choice compared to a long reliever choice.

This also stands as one of the best reasons for using a larger rotation (6 instead of typically 5, or, for vintage seasons, 5 instead of typically 4). The #6 or #5 extra starter in the rotation serves as the weakest pitcher in the stable who has long pitching stamina --yellow or green in OOTP--and is used only to start against weaker opponents & otherwise skipped. But since this requires manually manipulating the rotation in OOTP, you might as well bring in an Emergency SP from the pen, for those games against weaker opponents. The difference being if you want that weak, high stamina pitcher to be kept out of the bullpen & saved for starting those specific games.. and also it's easier than moving pitchers in & out of the rotation to replace and/or switching back & forth between rotation sizes during the season. There is also the question of whether it's allowed/legal/fair to have a different size rotation than the rest of the teams in the league. Because OOTP certainly does not alter rotation sizes of its AI teams.

Which brings us to NoOne's statement about "sweet & simple." Using such elaborate, alternative pitching staff might not be recognized by OOTP & therefore could be interpreted as weak or incorrect (?) by AI and somehow penalized. I've often wondered what AI 'thinks' about a user who uses alternative rotation sizes. On one hand, it shouldn't make any difference, as it's only a manager strategy. On the other hand, it's not really fair since OOTP doesn't include such managing strategy options for AI controlled teams.

Thanks for your input.

[realstar]

Originally Posted by NoOne

... - the manager could be a big reason for using the pen in ways you dislike. play with sliders and see if you can find a "template" for a coach that does what you want. (i just edit my coaches; i say jump, and they ask how high?)

Well said. If the game is not playing realistically, learn how to use the editor.