Warrior General Games

Thought I would test out some AI for a championship game so I ran these two scrims the same day using the same settings between two teams I own gameplanned against each other.

http://glb.warriorgeneral.com/game/game.pl?game_id=2828196

http://glb.warriorgeneral.com/game/game.pl?game_id=2828198

It was super helpful.

private game

It has come to my attention these games are private and I'm dumb. They were:

L 10 - 27
W 70 - 0

Not just the score differential but in one the total yards were roughly the same, and in the blowout it was 800+ Boss yards to less than 100 Pwny yards. They were wildly different games.

morale 4tw

It should be noted that I scrimmed these teams 42 times throughout the season. These two games were the worst loss for Boss (there were only 3) and the best win (by a score of +8).

Originally posted by crazy454
morale 4tw

Which wouldn't matter if nothing changed between games (as far as play calling and skill #'s not changing) which, according to the OP, nothing between the games were changed.
Not that, if you did this 10 times, every single game would end up p'much the exact same but a wild variance such as you're attempting to show does make one scratch one's chin.

Originally posted by Theo Wizzago
Which wouldn't matter if nothing changed between games (as far as play calling and skill #'s not changing) which, according to the OP, nothing between the games were changed.
Not that, if you did this 10 times, every single game would end up p'much the exact same but a wild variance such as you're attempting to show does make one scratch one's chin.

People been sayin' it for many seasons now, sim is broken hense why so many left...no stability in the sim

I dont really think the sim is broken. Like I said I ran 42 scrims between them this season, mostly testing QB tactics. All the scores were relatively reasonable and I even ran numbers that fit patterns.

But these games when I was testing the new AI were just..... extreme...

Originally posted by Donk3yMan
I dont really think the sim is broken. Like I said I ran 42 scrims between them this season, mostly testing QB tactics. All the scores were relatively reasonable and I even ran numbers that fit patterns.

But these games when I was testing the new AI were just..... extreme...

were there any major adjustments to the OAI to exploit a certain DAI?

Tends to happen most often when you have a play hit early that the defense can't account for, and you have auto adjust strongly. Then this can lead to morale spiral and everything works at that point.

In the more competitive game, a lucky deflection or diving tackle early on could be why that particular play didn't come up as often, or possibly ever.

Just something to look for in the scouting as I can't see the results.

Originally posted by F00tballJunkie
were there any major adjustments to the OAI to exploit a certain DAI?

No exact same settings both games

Originally posted by Team Nucleus
People been sayin' it for many seasons now, sim is broken hense why so many left...no stability in the sim

Actually, pretty much nobody has been saying that.

If anything, people have been complaining that the sim is unchanged--that too much of it is a known book that forces people into cookie-cutter builds.

Originally posted by Donk3yMan
Thought I would test out some AI for a championship game so I ran these two scrims the same day using the same settings between two teams I own gameplanned against each other.

http://glb.warriorgeneral.com/game/game.pl?game_id=2828196

http://glb.warriorgeneral.com/game/game.pl?game_id=2828198

It was super helpful.

then try to turn off auto adjustment and see how it goes for two games.

Originally posted by Tomcic
then try to turn off auto adjustment and see how it goes for two games.

Yeah I should have given that a go but I ran those scrims pretty close to game time and then didn't care when we lost, now I have to refill one of the rosters. Realistically though I wouldn't launch into the casual pro LC with no AA so I didn't even consider it for testing.