OOTP Developments Forums - View Single Post

Cobby · 07-09-2019, 02:43 PM

I find the whole trading debate fascinating - to the point that I've wanted to try my hand at writing my own AI trading logic (...one of these days).

But. Right now I'm using Normal/Neutral. On that setting I find the AI to be pretty good -tactically good at least, not necessarily strategically.

There are a few important things to note however:

This is a fictional league
It's currently 1893 - (I started in 1871), and the reserve clause is in effect.

Since it's a fictional league, there is no name-recognition effect like what you have if you play historical or modern. I often wonder if much of the outrage over bad AI trades doesn't stem from this name-recognition effect.

Second, since it's in the reserve clause era, there are no bad contracts to worry about and most teams are in great shape financially. This makes the AI's job a lot easier and it seems to do a great job in this era - when talent evaluation and projection are the only things it has to consider. The financial considerations are less important.

My observation is that the trade AI seems to get better with every version. In past versions my favorite strategy would be to pry top prospects loose from the AI teams. Do this enough and you've got a constant pipeline of young players and a surplus of older talent just about to decline that you could trade for more prospects.

That strategy no longer works. At least not so far. The AI won't part with a top prospect for *anything* it seems. I shudder to think how tightly the AI would hold on to prospects if I had it set to "favor prospects". It's nice that I have to adapt my strategy, but I think maybe it's gone overboard. Sometimes I think that if I cloned a team's top prospect and put 5 clones on my team and offered them for that prospect that it still wouldn't take the deal. (I may actually try that)

What is working for me now though is to trade for younger established players. These, the AI will part with. It won't trade a 21-23 year old prospect for anything, but it will trade good players in the 26-28 year old range. I still have to give up a lot of talent to get them though, so I don't feel that I'm fleecing the AI too badly.

Remember though - 19th century baseball is odd. Hardly any home runs, so power is less important. Not many strikeouts, so avoid K's is not so important, and not too many walks either. And especially, the lack of financial worries - so what I'm saying and observing may not apply at all to most of you....

One more observation: if you make the trade, it means that you thought it was a good trade. This is almost by definition - few people will purposely make a bad trade. So there is a bias toward thinking that you've fleeced the AI. There is a recent thread where the poster was worried that he had fleeced the AI. But after other posters had chimed in - it wasn't so clear anymore who it was that had been fleeced. It may be the AI is better than you think.

07-09-2019, 02:43 PM	#10
Cobby Minors (Triple A) Join Date: May 2016 Posts: 252	I find the whole trading debate fascinating - to the point that I've wanted to try my hand at writing my own AI trading logic (...one of these days). But. Right now I'm using Normal/Neutral. On that setting I find the AI to be pretty good -tactically good at least, not necessarily strategically. There are a few important things to note however: This is a fictional league It's currently 1893 - (I started in 1871), and the reserve clause is in effect. Since it's a fictional league, there is no name-recognition effect like what you have if you play historical or modern. I often wonder if much of the outrage over bad AI trades doesn't stem from this name-recognition effect. Second, since it's in the reserve clause era, there are no bad contracts to worry about and most teams are in great shape financially. This makes the AI's job a lot easier and it seems to do a great job in this era - when talent evaluation and projection are the only things it has to consider. The financial considerations are less important. My observation is that the trade AI seems to get better with every version. In past versions my favorite strategy would be to pry top prospects loose from the AI teams. Do this enough and you've got a constant pipeline of young players and a surplus of older talent just about to decline that you could trade for more prospects. That strategy no longer works. At least not so far. The AI won't part with a top prospect for anything it seems. I shudder to think how tightly the AI would hold on to prospects if I had it set to "favor prospects". It's nice that I have to adapt my strategy, but I think maybe it's gone overboard. Sometimes I think that if I cloned a team's top prospect and put 5 clones on my team and offered them for that prospect that it still wouldn't take the deal. (I may actually try that) What is working for me now though is to trade for younger established players. These, the AI will part with. It won't trade a 21-23 year old prospect for anything, but it will trade good players in the 26-28 year old range. I still have to give up a lot of talent to get them though, so I don't feel that I'm fleecing the AI too badly. Remember though - 19th century baseball is odd. Hardly any home runs, so power is less important. Not many strikeouts, so avoid K's is not so important, and not too many walks either. And especially, the lack of financial worries - so what I'm saying and observing may not apply at all to most of you.... One more observation: if you make the trade, it means that you thought it was a good trade. This is almost by definition - few people will purposely make a bad trade. So there is a bias toward thinking that you've fleeced the AI. There is a recent thread where the poster was worried that he had fleeced the AI. But after other posters had chimed in - it wasn't so clear anymore who it was that had been fleeced. It may be the AI is better than you think.