JimboJ
Posts: 70
Joined: 10/8/2005 Status: offline
|
quote:
ORIGINAL: Amaroq Yeah, definitely. 1. Computer manager using its best lineup in the playoffs. 2. Computer managers double-switch-ing. 3. Intentional walks taking into account 'who's on deck'. 4. Trade evaluation function taking into account - 'delta' at the positions involved (need/opportunity cost) - long-term planning - number of years left on the contract 5. AI GM's trying to renew contracts prior to the trade deadline, and trying to trade away last-year players who are unwilling to renew. 6. Pinch-hitter AI taking into account the need to save some bench hitters for the pitchers' slot in extra innings (I think its too likely to pinch-hit for the #8 hitter with its last bench player, when the pitcher in the #9 slot needs to be pinch-hit for.) 7. AI making fewer moves based on 'stats' and poor performance; it feels like it will bench or demote 'stars' for statistically insignificant poor performance by June, for example. I usually find that my team 'breaks away from the pack' through June, July, and August, as the AI managers bench their starters or send them to the minors, and I rarely agree with the AI's starting lineups through a September stretch drive or playoff series. Great list! +1 for every one of these. Despite all the improvements Shaun has made to the in-game experience, I still get the feeling he is more concerned with stats and how they are presented than he is with making the games themselves more realistic and making the manager smarter. I'm not complaining, because I LOVE this game. I just want to see more focus on improving the in-game AI. I want it to the point where if I am managing a game against the computer with teams of equal talent, then I have to really struggle to beat it every time. Kind of like a chess program set on a high difficulty level.
|