OOTP Developments Forums - View Single Post - How does the AI handle MY minor league rosters?

SteveP · 11-13-2009, 01:41 PM

My only comment on this question is that the AI has a tendency to bounce borderline players from one level to the other in the minor league structure quite often. Some OOTPers get quite incensed about this and I think end up managing promotions/demotions for this reason alone. Also, many OOTPers are frustrated by the fact that the AI does not always accord the same value to some players that they themselves would. However, it is useful to consider that:

1. These are borderline players at the time. They could arguably be at either level.

2. The AI is incapable of carrying out a long-term strategy for developing specific players. Players develop or they don't.

3. A certain amount of churn (more than we would see IRL) is useful to the performance of the AI. The AI "learns" by trial-and-error. If something seems to be working out OK, the AI will continue it. Otherwise, it makes a change. It will keep making frequent changes until things are working out OK. The AI doesn't have the ability to do any analysis of all this. It just keeps trying different things until it gets better results.

If you can live with all of that, then you can save yourself a lot of time and effort by letting the AI manage things. Otherwise, you have to do it yourself.

11-13-2009, 01:41 PM	#3 (permalink)
SteveP Moderator Join Date: Aug 2008 Posts: 2,776 Thanks: 6 Thanked 261x in 198 posts	My only comment on this question is that the AI has a tendency to bounce borderline players from one level to the other in the minor league structure quite often. Some OOTPers get quite incensed about this and I think end up managing promotions/demotions for this reason alone. Also, many OOTPers are frustrated by the fact that the AI does not always accord the same value to some players that they themselves would. However, it is useful to consider that: 1. These are borderline players at the time. They could arguably be at either level. 2. The AI is incapable of carrying out a long-term strategy for developing specific players. Players develop or they don't. 3. A certain amount of churn (more than we would see IRL) is useful to the performance of the AI. The AI "learns" by trial-and-error. If something seems to be working out OK, the AI will continue it. Otherwise, it makes a change. It will keep making frequent changes until things are working out OK. The AI doesn't have the ability to do any analysis of all this. It just keeps trying different things until it gets better results. If you can live with all of that, then you can save yourself a lot of time and effort by letting the AI manage things. Otherwise, you have to do it yourself.