AI to AI trade logic is awful

[Thundercrack]

Originally Posted by R. Hansh

Sweed,I am playing an historical season starting in 1970, development is off, trading is on hard and in the middle for prospects, I am the GM of one team,in the month of July I had 54 trades and it got worse during the off season so I turned trading off. My question is where do I find the trade frequency in the game?

Since so many people are seeing this avalanche of trading in different setups and style of play, it’s obvious no one actually tested this during beta. I have no doubt they will fix this, but how does something this glaring get through

[HRBaker]

Originally Posted by Thundercrack

Since so many people are seeing this avalanche of trading in different setups and style of play, it’s obvious no one actually tested this during beta. I have no doubt they will fix this, but how does something this glaring get through

Related story...


When I was till working, I was in IT and partially responsible for upgrading the company (about 5,000 PCs) to each new version of Windows. Once we would get the final commercial product, we would set up testing in all departments, running our specific custom programs against the new version to find out what didn't work. We would then modify our programs to avoid the issues and make sure everything worked across all departments before we rolled out the new Windows version.

In one of these upgrades, I gotr called into the IT's department head's office asking why, after 3 months, we still couldn't roll out Windows. He actually asked me if it was because the Windows release was so bad or if my people weren't capable. (I bit my lip).

I went on to explain that no matter how much testing Microsoft did before they released the product, there were simply too many specific, custom, variables in the world to test without access to those variables. Once the product was available - THAT"S WHEN that kind of testing would happen, and that was why we were still working on it.

OOTP is no different. The program, after 24 years has added more options that, I sure, Markus ever imagined were possible. EACH of these variables can have unexpected results on the game engine, and MANY of them won't get tested until the customer base (HUGELY larger than the Beta team) actually tried it out.

This is life. It even happens with Microsoft products, and they have a ton more people than OOTP.

[Thundercrack]

Originally Posted by HRBaker

Related story...


When I was till working, I was in IT and partially responsible for upgrading the company (about 5,000 PCs) to each new version of Windows. Once we would get the final commercial product, we would set up testing in all departments, running our specific custom programs against the new version to find out what didn't work. We would then modify our programs to avoid the issues and make sure everything worked across all departments before we rolled out the new Windows version.

In one of these upgrades, I gotr called into the IT's department head's office asking why, after 3 months, we still couldn't roll out Windows. He actually asked me if it was because the Windows release was so bad or if my people weren't capable. (I bit my lip).

I went on to explain that no matter how much testing Microsoft did before they released the product, there were simply too many specific, custom, variables in the world to test without access to those variables. Once the product was available - THAT"S WHEN that kind of testing would happen, and that was why we were still working on it.

OOTP is no different. The program, after 24 years has added more options that, I sure, Markus ever imagined were possible. EACH of these variables can have unexpected results on the game engine, and MANY of them won't get tested until the customer base (HUGELY larger than the Beta team) actually tried it out.

This is life. It even happens with Microsoft products, and they have a ton more people than OOTP.

I get it. Should have learned my lesson last year about switching over too soon

[OutS|der]

Originally Posted by shakturi101

Can you name a game in the strategy/management market that has been able to do this well? I've played paradox games (EU4/Stellaris), civ games, Football Manager, and the most popular complaint across all communities in the game is the inability to create an AI that can adequately challenge a human. I'm not an expert on how to create good AI and don't know what is going on under the hood, but if all these companies, including some with lots of resources, are struggling to create a challenging AI, it would seem it's not really that easy to just "fix."

The games like EU4/Stellaris/Civ just have the AI cheat in higher difficulties by giving them bonuses, which does help a lot, but once you get past the initial phase where the human is weaker, the human is able to snowball and become insanely powerful. Also, playing with huge advantages to the AI sometimes requires specific actions that feel exploity because the actions only make sense with the knowledge of the difficulty settings you are using and, therefore, break immersion.

Another way games can mitigate this issue is by having natural challenge (that don't necessarily break immersion) built into certain starting points for a human. For example, in EU4, you can start as a really small nation with little resources, which can make it really hard to ultimately compete with the big countries like France and the Ottomans. In Football Manager, you can start in a really low tier of Football (like the sixth tier in England) with really low finances and try to work your way up all the way to the top of the league from there.

OOTP has this as well, but most people, rightly so, want to play in MLB. But even playing the MLB team with the lowest resources (and they even get revenue sharing lol) pales in comparison to the natural resource gap you can create for yourself in a game like EU4 or Football Manager. As much as baseball fans complain about the resource disparity in MLB, it pales in comparison to the European soccer leagues, especially for teams in the lower leagues.

Ultimately, in all these games, if you are a competent human player who understands the game's mechanics, you will consistently make better decisions than the AI in every area, which will lead to the human outperforming the AI based on their starting point. But playing with the A's with an endboss of the Yankees/Dodgers isn't quite the same in difficulty as a non-league English club with an endboss of Manchester City.

I assume a big factor in decent AI is system requirements need to be kept low to attract the biggest audience. If minimum requirements was say a 9900k and 32gb ram I assume would be able to run more complex AI.
Then again I know nothing so as usual I'm just likely making an ass of myself by assuming

[PSUColonel]

Originally Posted by OutS|der

I assume a big factor in decent AI is system requirements need to be kept low to attract the biggest audience. If minimum requirements was say a 9900k and 32gb ram I assume would be able to run more complex AI.
Then again I know nothing so as usual I'm just likely making an ass of myself by assuming

FM does a pretty darn good job of AI simulation when you’re using full detail for the leagues. This requires a lot of CPU resources as I specifically purchased a machine to churn it out. Is it perfect? Nope…but it does a very reasonably good job.

[HoustonGM]

https://www.youtube.com/watch?v=SV_8smWvq6U

I'm not one to point out a couple random weird bad trades as evidence that the engine needs to be overhauled because as Lukas has said, bad trades happen, that's fine... but the offseason of this video at the default settings is just simply not realistic in any way.

[R. Hansh]

Originally Posted by Sweed

I did take another look at my game and just to correct what I posted, I set my trades to "very low" not "low". No idea on the effect, but I do manage every game (play slowly day to day), so I believe this will do nicely for me as I progress through ST. I'd like to think by the time I'm done with that we'll probably have heard back from the development team with some tweaks.



Thanks for answering. Saved me having to look through menus again. Even after playing for years I still can't just go right to many of the things I want to.

Thanks David.

[MikeG719]

so if we wanted to start a game today.....best bet is to set trades to very low, hard mode, and 65/20/10/5?

[tonnage]

Originally Posted by MikeG719

so if we wanted to start a game today.....best bet is to set trades to very low, hard mode, and 65/20/10/5?

That's the way I play!!

[Guthrien]

Originally Posted by HoustonGM

https://www.youtube.com/watch?v=SV_8smWvq6U

I'm not one to point out a couple random weird bad trades as evidence that the engine needs to be overhauled because as Lukas has said, bad trades happen, that's fine... but the offseason of this video at the default settings is just simply not realistic in any way.

It's glaringly obvious to anyone playing it I'd think. The posts trying to justify some of the player value are painful. I don't expect perfection, and 23 by and large worked really well so I have every faith this will get fixed. There's no way anyone should think a simulation this big can be perfect out the gate. I'll grant people it's so obvious there should have been some red flags.

I still think the AI is offering more interesting trades. Say 20% vs 0% in 23. But current state is not playable unless you want total bizarro world (which hey, sometimes you do).

[Thundercrack]

Originally Posted by Guthrien

But current state is not playable unless you want total bizarro world (which hey, sometimes you do).

Indeed. We’re a week out from release. Let’s hope we’re not saying the same thing a month out

[krsult29]

In what world would the Baltimore Orioles trade Rutschman to the Braves for RF - Chris Sharpe (POT 30), RP Collin McHugh (POT 45) and C Andrew Keck (POT 30). Trade AI is wacked this year.

[PSUColonel]

If you ask me…the trade AI wasn’t bad last year at all…just needed to get rid of the “make this work now” button and that is now optional…just sayin

[Dr Naysay]

The other issue with people's complaints about trading which again I have literally just seen one posted on another forum has nothing to do with the trading engine but how some players are rated. People can't wrap their heads around
how a highly rated prospect can get such an awful return yet said highly rated player whose hit well at every level is rated 2.5 stars by OSA and so the AI aren't exactly as high on the player as we are.

There is definitely work that is needed to be done on the trading engine but some of the player ratings need to be sorted out too because it's affecting some of the AI trades as we perceive them.

[Cod]

Is this problem just effecting MLB (real-life) saves? Or do fictional leagues have the same problem?

[Cactusguy21]

Fictional too, I tested my European Baseball League on it and within two days the greatest player in league history was traded for a pair of 30/45 prospects

[Sweed]

Originally Posted by Cod

Is this problem just effecting MLB (real-life) saves? Or do fictional leagues have the same problem?

It is under the hood evaluations so would affect all leagues. The question is how much at what settings (?), at least until a patch arrives. In my game I've gone to "very low" and have not had a trade since doing so, but.. I am in spring training so not really expecting any either.

Originally Posted by Cactusguy21

Fictional too, I tested my European Baseball League on it and within two days the greatest player in league history was traded for a pair of 30/45 prospects

I posted earlier that in my imported game I saw almost twice the number of trades in the offseason in v24 compared to the same time frame and situation (first off season in a new version with both leagues imported directly after the WS).


While I did see a lot of trades none included any current superstar or even big star type players. I did see some trading of top prospects. The only "top prospect" trade I was involved in was Miami offering me a SP, the #10 overall on the list along with a "maybe a cup of tea RP". For that they wanted my top SP prospect that was a top #25 on the list, and not far behind the guy they offered in skills. My guy has no MLB service, is pretty much "can't miss". Their guy had about a half year of service and slight control issues with a 4 on the 8 scale, fully developed according to my scout. The rub? As soon as the trade was completed I got a "our players are concerned about "SP X's" affect on clubhouse moral". IOW a problem child. Skill wise the deal is in my favor, but overall? Maybe not. And no, I did not dig that deeply into the other trades.

What was interesting to me is all of these players were only traded once. IOW it seemed once the AI GM's "got their guy" they kept him. It's possible I missed something. As the issue became reported, known, and discussed I was in the midst of my offseason and just getting to the winter meetings. So, not only did I review every trade clicking on every player (as I always have, IE not in response to the issue) I also wondered if this was like "sign and release"? IE players were being traded multiple times? In my game this was not happening.

There seemed to be a lot of minor leaguers being traded for other minor leaguers. At first I didn't think much of it, IE the AI revaluating players/needs with a new version's algorithms and trying to fill needs. As I looked closer I found the same as the MLB deals, nobody being traded more than once.

IDK what that means for sure, but it seems the AI wasn't "trading just to trade", at least in my game. What does my experience mean in the broad context of the subject? Maybe nothing Other than answering Cod's question along with the effects I'm seeing in my game.

[Grafton19]

I don’t know if it was PSU or another OOTP OG who said years ago

It’s your universe make it your own. That’s why I do commissioner mode for trades until deadline. Then allow some CPU trades to test AI

[PSUColonel]

Originally Posted by Grafton19

I don’t know if it was PSU or another OOTP OG who said years ago

It’s your universe make it your own. That’s why I do commissioner mode for trades until deadline. Then allow some CPU trades to test AI

Not me lol

[tonnage]

Originally Posted by Cactusguy21

Fictional too, I tested my European Baseball League on it and within two days the greatest player in league history was traded for a pair of 30/45 prospects

I'm not seeing it as bad as most are saying here. In my fictional save seldom do I question a trade and I make sure to read every AI to AI trade that takes place. Again, my settings are 60, 25, 10, 5, I have trading set to the hardest level, and AI trading set to very low. I don't sim games, I play every game one-pitch mode. Works for me.