Why isn't the AI more aggressive on the waiver wire?

[texasmame]

I see many a player on the waiver wire that would help teams (and they have the space, both cash and roster-wise) yet they just sit there - unclaimed.


I'd really like to see the AI get more aggressive in this respect.

[Bobfather]

How much are the contracts?

[texasmame]

Originally Posted by Bobfather

How much are the contracts?

Anywhere from minimums to big contracts to guys who have already had their entire contract picked up by another team - runs the gamut.



It's nonsense.

[tonnage]

I just watched a very good young player go through the waiver process without being claimed. I looked at the team that put him on waivers and there's no way the player should have been waived. What's crazy is nobody claimed him. I would have claimed him in a heartbeat but felt that in the real world I never would have been given the chance so I let him go. The waivers process needs work for sure.

[ExpansionManager]

I was thinking there should be something in the AI logic, on top of the general evaluation, that prevents top 100 prospects from being recklessly thrown around. First they get put on the 40 man when they're neither major league ready or rule 5 eligible, then they get waived while they still have option years, all to make room for a #5 starter in AAA, without having to waive the backup catcher. Even in the case of an actual bust, I'd rather see a "bad" trade where a second team gives them a chance, than the original team throwing them away for nothing.

[texasmame]

Originally Posted by ExpansionManager

I was thinking there should be something in the AI logic, on top of the general evaluation, that prevents top 100 prospects from being recklessly thrown around. First they get put on the 40 man when they're neither major league ready or rule 5 eligible, then they get waived while they still have option years, all to make room for a #5 starter in AAA, without having to waive the backup catcher. Even in the case of an actual bust, I'd rather see a "bad" trade where a second team gives them a chance, than the original team throwing them away for nothing.

Don't even get me started on Rule 5. I disabled that many versions ago when a contending team put Gehrig, in his prime, on waivers so they could keep some middling R5 pick.



Nope!

[LeftyNate]

I actually had an AI team claim a player I wanted during this save in 26. I was glad to see that. I usually look for bullpen help on the waiter wire. I’ve just played through the offseason before the 2026 season and feel like I haven’t seen as much “good” talent on the waiver wire. Mostly 45 ovr. That’s pretty representative of real life, guys on the fringes. But I’m fine with there not being as much claiming as real life. (Though it seems like my players historically have been claimed more often than I’d think.) As long as I’m not seeing many 50+ ovr players.

I will say I did adjust “player evaluation” to include more statistics this year. So maybe that’s had a hand in limiting how many players are going to waivers. (Drove me nuts seeing a pitcher put on waivers when he has a 3.50 ERA. That wouldn’t happen irl.)

[Sweed]

Originally Posted by LeftyNate

I actually had an AI team claim a player I wanted during this save in 26. I was glad to see that. I usually look for bullpen help on the waiter wire. I’ve just played through the offseason before the 2026 season and feel like I haven’t seen as much “good” talent on the waiver wire. Mostly 45 ovr. That’s pretty representative of real life, guys on the fringes. But I’m fine with there not being as much claiming as real life. (Though it seems like my players historically have been claimed more often than I’d think.) As long as I’m not seeing many 50+ ovr players.

I will say I did adjust “player evaluation” to include more statistics this year. So maybe that’s had a hand in limiting how many players are going to waivers. (Drove me nuts seeing a pitcher put on waivers when he has a 3.50 ERA. That wouldn’t happen irl.)

I think that is a good point on player evaluation being more stat heavy, and one I hadn't thought of. I use a stat heavy evaluation (25/25/25/25) and rarely see any players on the wire that I feel shouldn't be there. Maybe that is why? IDK.

To the point of the thread I've been happy with how the AI does or does not claim waiver players. When it does take a player I can normally see how it improves their team. The ones that don't get claimed? They are usually the "same" as the player the AI team already has or worse. At the end of the day a "forced" increase in AI claim frequency may be doing something just to do it. The AI shuffling players of like ability with no real improvement to the team.

To be sure I have no idea how OOTP claims compare to real life and have no data to say OOTP is within real life expectancy. I would guess OOTP does come in low, but is that a bad thing? Perhaps in real life teams are more likely
to speculate that "this guy could break this bad streak of bullpen play. We need to do something." Where computer AI sees two players that are virtually the same borderline player, ie no reason to make a change.

My 2 cents.

[SirMichaelJordan]

Originally Posted by ExpansionManager

I was thinking there should be something in the AI logic, on top of the general evaluation, that prevents top 100 prospects from being recklessly thrown around. First they get put on the 40 man when they're neither major league ready or rule 5 eligible, then they get waived while they still have option years, all to make room for a #5 starter in AAA, without having to waive the backup catcher. Even in the case of an actual bust, I'd rather see a "bad" trade where a second team gives them a chance, than the original team throwing them away for nothing.

AI eval settings can negate this.

[ExpansionManager]

Originally Posted by SirMichaelJordan

AI eval settings can negate this.

What's considered the best settings, I play with current ratings off so my AI eval is primarily stats, with a little bit of ratings. At any rate I think the AI is mostly solid, until that one outlying decision that would never happen in real life.
For an extreme example, let's say OSA falsely rates a minor leaguer as the #1 prospect in baseball, a can't miss Ted Williams or Walter/Randy Johnson, but their controlling teams scout knows the truth that they're minor league trash for life. The current AI seems ok with cutting or waiving a player with no future, where a real GM would put them on the trade block and field offers from rebuilding teams.
Obviously most scenarios and transactions are less extreme than that, but I think the AI would rather throw away a player that isn't in their plans than "lose" a "bad" trade.

[uruguru]

Originally Posted by SirMichaelJordan

AI eval settings can negate this.

As I understand it, AI Eval Settings determine the overall (current) rating of a player, though, not the potential rating. Therefore young players whose primary value lies in their potential ratings should not be affected so much by the AI Eval Settings.

[WhiskyTango]

Originally Posted by uruguru

As I understand it, AI Eval Settings determine the overall (current) rating of a player, though, not the potential rating. Therefore young players whose primary value lies in their potential ratings should not be affected so much by the AI Eval Settings.

If so then this may explain another issue, besides but similar to waivers, how prospects are handled. Meaning, they aren't. They get overlooked and released.

[DennisDenver]

Originally Posted by SirMichaelJordan

AI eval settings can negate this.

Suggestions on what to set these at currently? I actually have never messed with this before but now it kind of makes sense.

I did search came up via google on OOTP 23 forum that they kinda all settled on "55-20-15-10"

[SirMichaelJordan]

Originally Posted by ExpansionManager

What's considered the best settings, I play with current ratings off so my AI eval is primarily stats, with a little bit of ratings. At any rate I think the AI is mostly solid, until that one outlying decision that would never happen in real life.
For an extreme example, let's say OSA falsely rates a minor leaguer as the #1 prospect in baseball, a can't miss Ted Williams or Walter/Randy Johnson, but their controlling teams scout knows the truth that they're minor league trash for life. The current AI seems ok with cutting or waiving a player with no future, where a real GM would put them on the trade block and field offers from rebuilding teams.
Obviously most scenarios and transactions are less extreme than that, but I think the AI would rather throw away a player that isn't in their plans than "lose" a "bad" trade.

That's under the impression that a team wants said bad player. OSA is just media; an opposing team's scout may have also rated him poorly. OSA isn't accurate, and I believe they don't exceed average accuracy. A team would have to find a package deal to include that guy; they usually do that, and changing trade settings to favor prospects helps this too. Heavily favoring prospects is good because most coaches in the game are set to favor vets heavily, so it balances the game out.

[SirMichaelJordan]

Originally Posted by uruguru

As I understand it, AI Eval Settings determine the overall (current) rating of a player, though, not the potential rating. Therefore young players whose primary value lies in their potential ratings should not be affected so much by the AI Eval Settings.

Other players around him will be rated differently, determining how he's handled.

[SirMichaelJordan]

Originally Posted by DennisDenver

Suggestions on what to set these at currently? I actually have never messed with this before but now it kind of makes sense.

I did search came up via google on OOTP 23 forum that they kinda all settled on "55-20-15-10"

It's really a personal preference, but I am known for the 25/25/25/25 setting.

Right now, I am using 0/50/25/25.

I suggest playing around until you feel the logic is better.

[uruguru]

Originally Posted by SirMichaelJordan

It's really a personal preference, but I am known for the 25/25/25/25 setting.

Right now, I am using 0/50/25/25.

I suggest playing around until you feel the logic is better.

10/60/30/0 for me.

[Bobfather]

Originally Posted by texasmame

I see many a player on the waiver wire that would help teams (and they have the space, both cash and roster-wise) yet they just sit there - unclaimed.


I'd really like to see the AI get more aggressive in this respect.

Are these players being signed after they clear waivers. If a team signs a player as a waiver pickup then the inherit the whole contract. After waivers they can sign the player for a minimum. This is closer to real life. A player passes through waivers, then catches on with a new team in a day or two. If a team claims a player usually the two teams work out a trade for a PTNL or cash.

[Paul Reuschel's Mustache]

Originally Posted by uruguru

10/60/30/0 for me.

When do "current year" stats end? Does current year equate to the calendar year (i.e. up to December 31st)? Does current year end once the offseason starts? Or does current year extend all the way up to the start of the next season?

[Rain King]

Originally Posted by Paul Reuschel's Mustache

When do "current year" stats end? Does current year equate to the calendar year (i.e. up to December 31st)? Does current year end once the offseason starts? Or does current year extend all the way up to the start of the next season?

Pretty sure that it flips as soon as you move to the Offseason and is "season" based not calendar based.

It has also been mentioned by developers that it understands sample size (i.e. a guy who is 5 for 10 isn't going to be considered a superstar just because you have a heavy emphasis on current year stats).

I've used something along the lines of 30-30-25-15 for quite a while and think AI teams behave pretty realistically.