The AI In-Game Manager

The AI In-Game Manager (Full Version)

All Forums >> [Current Games From Matrix.] >> [Sports] >> PureSim Baseball

Message

KG Erwin -> The AI In-Game Manager (10/27/2006 1:00:44 AM)
After a hiatus of a few weeks, I decided to get back into playing. I'm a game-by-game manager, and I've noticed that the AI really isn't too shabby at managing. I'm using the latest beta patch, with 50-man rosters (all real players, with the accelerated-entry feature ON) in the 1946 season. Now, it DOES call up on pinch-hitters at some odd times, but it seems to know when to call a steal or go for the hit & run or when to play the outfield deep or when to make defensive substitutions. As for defensive subs, I suppose that's why the odd choices for pinch-hitters sometimes occurs. I know that many guys prefer simming out games, but you oughta at least try a few in the play-by-play mode. Shaun has done some interesting tweaking in AI managerial tendencies, and overall I believe it's working well. I have a damn good team (IMHO), but yet I'm just 3-2 on this new real-player season, and I've already had two extra-inning nail-biters. I had one 11-inning game that ended up 1-0. This is in 1946! I took full advantage of the AI GMs after spring training, and made a series of trades to beef up the Pirates. So, I end up with a 19-year old Robin Roberts (who is very good but not yet fully developed) on my staff. In any case, I'm having fun with this latest incarnation, so it seems (at least to me) that very few issues are left to be ironed out before the patch can be released.

Deltadog -> RE: The AI In-Game Manager (10/27/2006 2:18:39 AM)
You are right A-1 manager does do a fairly good job. A few things it does that puzzles me are: 1. Intentially walk a lesser hitter to get to the opponents clean up hitter who is also leading the league in homers and batting average. 2. Has his best 2 hitters sac bunt and then a lesser hitter strikes out. 3. Trailing by 10 runs in the 8th or 9th and makes defensive substitutions. When a team is that far down that late in the game they need offense. Defense is a minor consideration.

KG Erwin -> RE: The AI In-Game Manager (10/27/2006 2:37:51 AM)
quote: ORIGINAL: Deltadog You are right A-1 manager does do a fairly good job. A few things it does that puzzles me are: 1. Intentially walk a lesser hitter to get to the opponents clean up hitter who is also leading the league in homers and batting average. 2. Has his best 2 hitters sac bunt and then a lesser hitter strikes out. 3. Trailing by 10 runs in the 8th or 9th and makes defensive substitutions. When a team is that far down that late in the game they need offense. Defense is a minor consideration. Yeah, but that's because the AI manager takes a tunnel-vision view of conditions AT game-time. It isn't sophisicated enough ( and realistically CAN'T be) to look at the big picture in season terms. Now, periodically during the season, I think the AI re-evaluates his teams' performances, and makes adjustments. I think this is about the best we can hope for, and it's decent now . You can try to outwit the AI all you want in-game , but the stat model and the vagaries of the parks can still screw-up the best-laid plans of a human player. The human isn't throwing the pitches or swinging the bats.

KG Erwin -> RE: The AI In-Game Manager (10/27/2006 3:08:03 AM)
Let me give you an example -- I played a game against the Cardinals at Sportsman's Park that went to 15 innings. The score was tied 6-6, and Stan Musial was up with 2 outs and no one on in the bottom of the 15th. Stan had not been hitting well, , so I stupidly decided to pitch to him, rather than pitch around him. I was feeling cocky, as my reliever seemed to have command of his pitches. Stan the Man hit it out, and that was that. Cardinals win, 7-6. My bravado did me in. I really felt like a dumbass. [:o] Conversely, in a game at Forbes Field, I was facing the Reds, and it was scoreless until the bottom of the 11th. Enos Slaughter and Ralph Kiner both singled, and then Bill Salkeld grounded out, putting runners at second and third. George Kell was next up. I called a hit and run -- an almost sure thing. Kell singled, and the game was over. The Pirates won it 1-0 and that helped remove the pain of that bad decision vs the Cards. I've said it before, and I'll say it again. Every managed game can become a drama in of itself.

wlfcards -> RE: The AI In-Game Manager (10/27/2006 9:51:14 AM)
There is still a problem with not the computer having enough catchers and the limits it places on game management. The computer team I played started a totally exhausted catcher. Their only back-up catcher is the starting 1B. As a cather his range is 22 and arm 36.

KG Erwin -> RE: The AI In-Game Manager (10/28/2006 7:43:55 AM)
quote: ORIGINAL: wlfcards There is still a problem with not the computer having enough catchers and the limits it places on game management. The computer team I played started a totally exhausted catcher. Their only back-up catcher is the starting 1B. As a cather his range is 22 and arm 36. What year are you starting with? My historical 46 Pirates had a surfeit of catchers, so I had to release a couple of them. This included some long-time vets (Floyd Baker and Al Lopez). They just didn't fit into my future plans.

wlfcards -> RE: The AI In-Game Manager (10/29/2006 12:25:20 AM)
"There is still a problem with not the computer having enough catchers * * *What year are you starting with? I am in Sandlot mode. I think I started with 2005 but experimented with bringing in players from other years prior to certain free agent drafts. Through various version of PS I am about 20 years from my initial draft. As stated in another post, in my last year with the most recent beta, my results are too highle skews. I had 2 pitchers with 600+ KOs, 1 with 500+, 3 with 400+ and 3 with 390+. 9 pitchers had ERAs of 3.00 or better. Had 1 hitter bat 422 and 2 bat in the 390s. HR leader had 57, followed by a 50 and 14 with 40 or better.

Page: [1]

Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI

1.75