Getting the AI to play as shrewdly as a human?

[onnel]

Originally Posted by injury log

The trading thing is messed up. I think things were changed in this version to solve a problem in OOTP8, but it's introduced a new problem. Last version, you could easily stack a trade with the AI, getting ten players in exchange for one, because the AI assigned nil value to anyone below a certain threshold. So you could set up what would be a fair one-for-one exchange, then add on nine semi-good prospects. One talent boost later, and you'd find you'd traded for a superstar.

If that was the only issue then it seems a much better fix (one which would avoid the problems we are now ostensibly seeing) is for the AI to put no value on additional players offered by the player, but to still include them at whatever value is appropriate (at least something - the chance that they will become a superstar) when trading them away.

I'm assuming the problem must be more complicated, however.

[Isura]

I know some recommend using 'favor prospects' but I find this makes the AI even dumber with trades. They totally don't value solid 26-32 year old vets (even the arbitration/cheap ones).

[The Wolf]

I do the same thing. Sometimes you eat the bear, sometimes the bear eats you.

[PhillieFever]

Well,I finally took the plunge.I'm in a historical universe with recalc so what I did first was turn off player development completely so that recalc was the only thing that determined how the players would perform,and then I changed the player evaluation settings to the one's recommended by Wolf and then finally,I turned off ratings,potential ratings and stars so the only thing I now have to go on are stats and scouting reports.As an added wrinkle I have players retiring according to history so there's always the chance that I trade for somebody who's all set to retire.The reason I did it now was so that I never see any of the ratings for next season,they will be hidden from the start.I don't have to worry about the draft because I have players entering the league with the team they broke in with.Its the best I could do to keep the league as historically accurate as possible while still making it as challenging as possible.

[Here24]

I'm so glad this thread is still alive. I have gone to all stats and I find it very difficult. I've found the game entertaining, but never have I described it as difficult. I wasn't sure if difficulty would lead to frustration or deeper enjoyment, but the more I play this way, the more enjoyable I find it.
The seasons take longer, but that may change as I get more comfortable with this way. Thanks again to everyone who contributes toward a more enjoyable OOTP experience - from developers to the average joe.

[StyxNCa]

Originally Posted by Isura

I think feeder leagues have stats

I don't have feeder leagues. I never figured out how to set them up etc. Seemed like too much bother.

[Afino]

Originally Posted by StyxNCa

I don't have feeder leagues. I never figured out how to set them up etc. Seemed like too much bother.

I always thought they were too. Then I did it, made sure I had enough teams to populate my draft pool, and I don't even think about them now.

[OldFatGuy]

Lots of good suggestions in this thread.

Now I've got one.

IMPROVE THE AI SO ITS CHALLENGING TO A HUMAN PLAYER.

[Afino]

Just one thing before I make the switch for my upcoming season (I'm excited!)


Wouldn't a talent change randomness of 10 make drafting automatic? I mean, that would more or less assure all 1st round picks are stars and less low-round picks could pan out someday.

[PotatoPeeler]

Originally Posted by OldFatGuy

Lots of good suggestions in this thread.

Now I've got one.

IMPROVE THE AI SO ITS CHALLENGING TO A HUMAN PLAYER.

If you play this way, it will be. That's the whole point of the thread.

[OldFatGuy]

Originally Posted by PotatoPeeler

If you play this way, it will be. That's the whole point of the thread.

I understand that. But, given that one of its mottos is "It's your game, play it your way" I would like to play it with ratings, with no house rules, and have it challenging. Chess AI used to be a laughing stock but after much work chess AI was developed to not only challenge humans, but beat the world chess champion.

While I expect baseball AI will never get that good, the fact is it can get a whole lot better. And it is weak in this game, as it is in every game I've ever purchased. I could list so many ways it's weak but then this would be a 15,000 word post so I won't. As far as roster management goes, it's weak but it's the best I've seen. As far as in-game strategy goes, it's weak and it's not the best I've seen. In fact, in-game management wise Sierra's BBPro series made better decisions, and that was over 10 years ago. So, I stand by my comment.

MAKE BASEBALL AI TO CHALLENGE A HUMAN OPPONENT.

Sorry, end of rant. For now. I rant about it two or three times a year and it does no good. I will end by saying the same thing I've said before. If someone out there develops a baseball game with good AI I'll gladly spend $100, $200, or maybe more. If you think that's a joke, it's not. Look at those who've purchased every version of OOTP. 9 versions now. I'm betting they've already put in several hundred dollars and still have weak AI.

[84Hammy]

Originally Posted by Afino

I always thought they were too. Then I did it, made sure I had enough teams to populate my draft pool, and I don't even think about them now.

True that! Also I find with stats only, I am not influenced by "potential" or what I think a player should do based on their ratings. I sim every game, doing about 5-15 games a day, so most seasons take only 2-3 weeks to complete, and I find this season that I have made the fewest roster moves, trades, etc...

[Afino]

I've taken the plunge into stats-only land officially now.

I won 5 championships in 6 years, but with my team aging, payroll bloating, and me wanting to try this "new" way, I decided to rebuild.

I started by trading a couple of older stars for prospects and picks (mind you, I'm playing with "hard" and "favor prospects") and calling up some of my younger talents. But are they ready???

My team actually is still expected by the owner to make the playoffs which isn't surprising considering my top starter is still around, I picked up another solid starter in a trade, my bullpen is still the best in the league, and I have three good power bats.

Time to find out

[The Wolf]

Originally Posted by PotatoPeeler

If you play this way, it will be. That's the whole point of the thread.

"

[The Wolf]

Originally Posted by Here24

I'm so glad this thread is still alive. I have gone to all stats and I find it very difficult. I've found the game entertaining, but never have I described it as difficult. I wasn't sure if difficulty would lead to frustration or deeper enjoyment, but the more I play this way, the more enjoyable I find it.

And it just keeps on getting better.

[t-bone shuffle]

Originally Posted by PotatoPeeler

If you play this way, it will be. That's the whole point of the thread.

Is it your contention that playing this way will eliminate, or at the very least significantly reduce the number of boneheaded in-game decisions the opposing AI Manager might make? For example, intentionally walking a batter with a runner on second (first base open) yet leading by three runs late in the game.

Are you saying that playing stats only will help with issues like this?

[PotatoPeeler]

Not at all. I'm saying that by turning ratings off and only basing your decisions on statistics that it will make the game more challenging.

[highandoutside]

Just to quickly add something to what has been a great thread. First, stats-only is a great way to play the game. I cannot recommend it enough. Second, to increase the difficulty even more, I put a strict salary cap on my team of $50 mil (other teams are not limited). I adjust the other teams finances primarily through their television and merchandise contracts. For instance, the Yankess, Mets, Cubs etc (the usual suspects) always have a ton of money to spend. The Tigers, Cardinals etc are midrange. I play as the Pirates and give myself almost no TV revenue (7 mil) and limit the merchandise, so that if I decide to ditch the salary cap I will still have to be very careful about the players I invest money in. I find that playing the game this way is very enjoyable because you must get the most value possible out of evey dollar spent.

[StyxNCa]

Originally Posted by Afino

I always thought they were too. Then I did it, made sure I had enough teams to populate my draft pool, and I don't even think about them now.

I read posts that were not agreeing with other posts on this topic and it all just got too complicated to try to figure out. Just when I thought I had figured it out somewhat, someone else posted something that made what I had figured seem wrong.

So, basically what is being said is that drafting in stats only doesn't work without feeder leagues.

[Afino]

Originally Posted by StyxNCa

I read posts that were not agreeing with other posts on this topic and it all just got too complicated to try to figure out. Just when I thought I had figured it out somewhat, someone else posted something that made what I had figured seem wrong.

So, basically what is being said is that drafting in stats only doesn't work without feeder leagues.

Well, it would be REALLY difficult.

Yes, with feeder leagues you have SOME stats to base opinions off of. I know my draft pool just came out, and going into the stats of the feeder players and actually doing the research is a LOT more immersive. That part I like already.

However, without feeder leagues, all you can do is sort by columns in the potential, pay attention to scout recommendations and positions. It doesn't "not work". It's just a very inexact science.