MTM2.com

Hi folks
I just reinstalled MonstertruckMadness2 on my "new" computer (Duron 1200MHz). Everythings works fine and looks nicer than ever before, but during the races, the AI performs extremely bad, with almost all trucks making detours to miss checkpoints!

I remember of similar problems with GPL, where the AI went mad if you increased frame rate beyond 36fps. Maybe this is related?

Does anybody observe the same problem, and is there a way to tweak the AI?
I'd like to practice for the next LAN party.

Cheers
Rocketalces

The AI's performance has nothing to do with your system settings, the track builder is at fault here by not taking the time to tweak the bots to follow the course correctly.

[edit]

there is no way for you to edit the AI's performance, so your pretty much stuck with what you get,... even with tracks that do have "good" AI paths set are still no competition for experienced racers, so your best bet for LAN practice is to go solo to find the best line and to find any short cuts.

if you'd like to find out how you stack up against other "live" racers check out the mtmg's track page, fast lap records are posted there per track as well as recommended settings (go to the "garage" link)

mtmg tracks



[This message has been edited by Malibu350 (edited 09-09-2002).]

you play GPL? whats your GPL rank? I'm still +28..... lol..

For a LAN party last week I installed MTM2 on a AMD K-6 400MHz.

By curiosity, I let the AI race the track and, surprise, they drove much better laps than on my machine. The weird "detour to avoid checkpoint" behavior was noticeably absent. So there must be something system related.

Cheers
Rocketalces

PS: Psyco, my GPL-handicap is -39

------------------

Nice I'm guessing you visit the racesim central forums? You've probobly seen me around..... BrandonB2003.. king of the off topic forum (besides norbs and his post a babe thread....... lol..)

------------------
"If it don't hurt, you're not havin fun"

"Bloody hell, I don't Want To Be A Millionaire, I don't want to see people's Greed, I don't want to be a Survivor, Shipwrecked or left on a bloody Treasure Island, and I don't want to know who Big Brother is watching!" - Drive2Survive

"When I read about the evils of drinking, I gave up reading."
Henny Youngman
On The Roof Racing
My Personal Site

Hmm, I see...
Well, a touch of humour from time to time can't hurt in those "GPL is the only racing game worth to be played".
In LAN-parties, MTM2 is king: even the girls raced with us. With GPL, this is rather - unlikely to happen.

Now this brings me back to the topic:
the MTM2 AI-behaviour seems system dependent.

Anything about that?

Cheers Rocketalces


------------------

The AI's are going to perform the same, track to track no matter what machine you're running on..

give me a track name that seems to perform better on your 1200mhz but performs poorly on a lessor machine and I'll run some tests myself on a p4 2ghz vs a p1 233mhz for fun.

what is gpl? a golf game?

It's the other way round: the AI performs much better on the slower machine (K6-400) than on the Duron 1.2Ghz
Two examples (always on professional level):
1) The height: on the Duron 1.2GHz, the AI trucks apparently try to drive around checkpoint 4, thus only 1 or 2 (out of 7) trucks complete the lap in a decent time. In a multilap race, none manages to finish in a decent time.
2) Crazy 98: lap times on K6 are about 5 sec better than on the Duron

GPL is Grand Prix Legends, the famous Sierra racing game released in 1998 and still unequaled for its physics modelling and its difficulty. If you have a couple of 100 hours to spend, try it! Otherwise, try it only if you have a taste for frustration .



------------------

AI tests


Test #1
Crazy 98
20 laps
professional
manual shift
1800 gears
soft suspension
shallow tires

......................P2 266MHz 32mb ram.......P4 1.6GHz 256mb ram
.......................Fast Lap.....Total ET........Fast Lap.....Total ET

Bearfoot...........00:49.74.....18:45.96........00:53.05.....18:04.54
Bigfoot.............00:49.93.....19:05.53........00:52.98.....18:13.90
Boogey Van.......00:53.58.....23:45.19........00:52.96.....18:38.30
Bulldozer...........00:49.88.....18:18.10........00:53.09.....18:39.12
Carolina Crusher.00:43.26.....17:33.74........00:53.03.....18:02.72
Grave Digger......00:46.28.....18:53.65........00:53.00.....18:18.29
Executioner.......00:50.15.....17:57.34........00:53.03.....18:12.48

=================================================

Test #2
The Heights
10 laps
professional
manual shift
1400 gears
soft suspension
medium tires

......................P2 266MHz 32mb ram.......P4 1.6GHz 256mb ram
.......................Fast Lap.....Total ET........Fast Lap.....Total ET

Bearfoot...........02:21.48.....26:31.74........02:29.77.....26:02.53
Bigfoot.............02:22.21.....24:52.90........02:32.35.....28:30.90
Boogey Van.......02:26.04.....29:39.67........02:04.19.....28:16.64
Bulldozer...........02:22.69.....27:50.21........02:35.02.....26:32.85
Carolina Crusher.02:24.32.....25:06.82........02:30.80.....25:54.10
Grave Digger......02:25.16.....26:23.23........02:33.37.....26:07.40
Executioner.......02:20.33.....28:12.05........02:32.38.....25:33.48


=================================================


Crazy 98, fastest recorded lap (non-AI) 00:40.98
The Heights, fastest recorded lap (non-AI) 01:36.36



[This message has been edited by Malibu350 (edited 18-09-2002).]

Hehehe Mal. I guess he is right. The AI is system dependent.

Still funny though, even if they are by 3-4sec slower per lap on a p4 they are alot more constant (as u can see by total time).

P.S. That Boogey van was dang fast on a p4 in The Heights.

------------------
"But the music is not too outstanding as the sound effects (Like the footsteps mario makes when walking on brick)" - review Super Mario Sunshine by KronoKross

[This message has been edited by Fila (edited 19-09-2002).]

OK, now we know that the AI is system dependent, but dependent from what?
I made some experiments with the monitor refresh rate (which is linked to the frame rate) but could not observe any effect neither on the fast nor the slow system.

Any ideas other ideas?