GT4 AI: OMG, it's even worse than I thought.
- Thread starter RaidoGT
- 91 comments
- 26,443 views
- 1,608
- Here and there...
- SUBMANIAC
I should've mentioned that I play on the PAL version, so this is 344 kmh, or 213 mph ^^'
Obelisk
Premium
- 9,792
- United States
- RMedia_Obelisk
I did a 2 lap Arcade race on the Nürburgring Nordschleife in a stock Mazdaspeed 6, 268 HP and 1500 kilograms, no assists, on SM tires. I did only the first lap of the race.
The rabbit is a 2001 Honda Civic Type R, starting in first place. It's followed by a Peugeot 205 Turbo, 2004 Civic Type R, 1990 Mazda RX-7 GT-X, and a Mitsubishi LEGNUM in that order.
Here's how the test lap went:
Sector 1: Passed the Mitsubishi just before going into the chicane prior to the left-right-left at the end of Hatzenbach. Passed the RX-7 just before Hocheichen and passed the '04 Civic going up the hill. Time gap to leader: +0'02.916, 3rd place.
Sector 2: Passed the 205 Turbo going through Schwedenkreuz and spotted the rabbit at Aremberg. The Peugeot is mad and hot on my rear bumper coming out of Aremberg (approx 3 car lengths).
Time gap to leader: +0'01.736, 2nd place
Sector 3: 205 fell behind right at the start of Fuchsröhre. Caught up to the rabbit as I was coming out of Adenauer Forst. Time gap to leader: +0'00.488
Sector 4: Passed the Civic as we started Metzgesfeld. It kept up with me through the first half of Kallenhard and fell behind in the second half. Time gap ahead: -0'01.832, 1st place.
Sector 5: Civic fell behind. There were no signs of any of the cars speeding up.
Gap: -0'03.341
Sector 6: Gap: -0'05.024
Sector 7: Gap: -0'07.682
Sector 8: Left the track just a bit at Wippermann, two outside tires on the grass and two on the kerb. Gap: -0'08.718
Sector 9: Gap: -0'09.794
Sector 10: Very minor scare at Galgenkopf, two tires left the tarmac on the outside of the turn. Recovered quickly. Gap: -0'11.403
Sector 11: By the time the Civic hit this sector, I was three turns from the finish. In addition, the 2004 Type R that was in 3rd place passed the 205 Turbo on Döttinger Höhe. Gap: -0'13.378
Sector 12/Finish: My lap time: 8'36:318. Gap: -0'13.997. Rabbit's lap time: 8'50.315
My test seems to indicate that there is no rubber-banding whatsoever. If the rubber-banding was present, I would not have finished 14 seconds ahead of the Civic, and there would have been more exchanges between the AI cars.
The rabbit is a 2001 Honda Civic Type R, starting in first place. It's followed by a Peugeot 205 Turbo, 2004 Civic Type R, 1990 Mazda RX-7 GT-X, and a Mitsubishi LEGNUM in that order.
Here's how the test lap went:
Sector 1: Passed the Mitsubishi just before going into the chicane prior to the left-right-left at the end of Hatzenbach. Passed the RX-7 just before Hocheichen and passed the '04 Civic going up the hill. Time gap to leader: +0'02.916, 3rd place.
Sector 2: Passed the 205 Turbo going through Schwedenkreuz and spotted the rabbit at Aremberg. The Peugeot is mad and hot on my rear bumper coming out of Aremberg (approx 3 car lengths).
Time gap to leader: +0'01.736, 2nd place
Sector 3: 205 fell behind right at the start of Fuchsröhre. Caught up to the rabbit as I was coming out of Adenauer Forst. Time gap to leader: +0'00.488
Sector 4: Passed the Civic as we started Metzgesfeld. It kept up with me through the first half of Kallenhard and fell behind in the second half. Time gap ahead: -0'01.832, 1st place.
Sector 5: Civic fell behind. There were no signs of any of the cars speeding up.
Gap: -0'03.341
Sector 6: Gap: -0'05.024
Sector 7: Gap: -0'07.682
Sector 8: Left the track just a bit at Wippermann, two outside tires on the grass and two on the kerb. Gap: -0'08.718
Sector 9: Gap: -0'09.794
Sector 10: Very minor scare at Galgenkopf, two tires left the tarmac on the outside of the turn. Recovered quickly. Gap: -0'11.403
Sector 11: By the time the Civic hit this sector, I was three turns from the finish. In addition, the 2004 Type R that was in 3rd place passed the 205 Turbo on Döttinger Höhe. Gap: -0'13.378
Sector 12/Finish: My lap time: 8'36:318. Gap: -0'13.997. Rabbit's lap time: 8'50.315
My test seems to indicate that there is no rubber-banding whatsoever. If the rubber-banding was present, I would not have finished 14 seconds ahead of the Civic, and there would have been more exchanges between the AI cars.
Future Testers...
Please post the video (or personally I'll assume it never happened)... Please use the GT Mode NOT arcade, the Option Stream for proper comparison loaded to 900hp+ with medium race tires, and use the Endurance Hall 24 hour (not the 4 hour), NTSC version prefered... Thanks
Please post the video (or personally I'll assume it never happened)... Please use the GT Mode NOT arcade, the Option Stream for proper comparison loaded to 900hp+ with medium race tires, and use the Endurance Hall 24 hour (not the 4 hour), NTSC version prefered... Thanks
Last edited by a moderator:
- 36,697
- Scotland
- GTP_daan
Because the A4 took its time getting past the other AI.
After it got past, it got up to its natural speed and reeled you in.
I would call it "driving faster than you in clear air."
If it was rubber banding, why did you pull away along Dottinger hohe at the end? Surely rubber banding wouldn't allow that?
- 1,608
- Here and there...
- SUBMANIAC
- 36,697
- Scotland
- GTP_daan
The A4 DTM is a quick car. The Option stream Z is not, no matter how much power you give it.Ok... whatever... I guess I can't drive... Believe what you wish... Don't text and drive and have a nice day...
The A4 catches you up on the twisty bits because it's a race car. The Option stream Z is not.
You pull away at the end because the Option stream Z has 800bhp. The A4 DTM does not.
I don't see rubber banding. I see 2 different types of car showing their strengths and weaknesses at different parts of the track.
- 1,608
- Here and there...
- SUBMANIAC
Even stock, the Option 350 can easily win against the DTM cars, mostly because of the massive power it makes though.
- 61
- Finland
So..I decided to try the Nurburgring 24 hour endurance.
Test 1:
Player car: Formula Gran Turismo
Track: Endurance hall, Nurburgring 24 hour endurance.
Starting grid: Vauxhall Calibra Super Touring, Audi A4 DTM, Celica GT300, MR-S GT300, Silvia S15 GT300, Player.
Player time: 5'51.553
Calibra time: 7'30.115
Audi A4 time : 7'33.599
Afterwards I went to drive these two cars to see how fast the AI is actually going.
Cars: Audi A4 Touring Car (DTM) '04 (Stock with oil change, racing medium tires)
Opel Calibra Touring Car '94 (Stock with 0 km, racing medium tires)
WAIT WAIT WAIT that's not the Vauxhall! Well..apparently the Opel Calibra has 421 horsepower in the garage while the Vauxhall Calibra has 449 hp in the car dealer. So the Opel should be either equal or slower..?
Calibra time trial: 6'58.601
Audi A4 time trial: 6'39.208
According to these results the AI are driving at grandma pace even if the player is driving a Formula Gran Turismo.
Test 2:
Player car: Honda Today G
Track: Endurance hall, Nurburgring 24 hour endurance.
In this test the A4 was replaced by the Audi TT-R DTM.
In addition, both the Calibra and the Audi had to overtake an ASL Arta Garaiya GT300.
Player time: 12'41
Calibra time: 7'32.863
Audi TT-R: 7'35.490
Final conclusions:
Earlier I did the tests in the TVR race, which showed no rubberbanding. Much like the TVR test, the AI seem to be completely oblivious about the player. If the player is very fast, they average 7'30. If the player is very slow, they average 7'30.
All evidence I find points to there being no rubberbanding in Gran Turismo 4.
PS. I have terrible quality video proof for this Nurburgring testing, if required.
Test 1:
Player car: Formula Gran Turismo
Track: Endurance hall, Nurburgring 24 hour endurance.
Starting grid: Vauxhall Calibra Super Touring, Audi A4 DTM, Celica GT300, MR-S GT300, Silvia S15 GT300, Player.
Player time: 5'51.553
Calibra time: 7'30.115
Audi A4 time : 7'33.599
Afterwards I went to drive these two cars to see how fast the AI is actually going.
Cars: Audi A4 Touring Car (DTM) '04 (Stock with oil change, racing medium tires)
Opel Calibra Touring Car '94 (Stock with 0 km, racing medium tires)
WAIT WAIT WAIT that's not the Vauxhall! Well..apparently the Opel Calibra has 421 horsepower in the garage while the Vauxhall Calibra has 449 hp in the car dealer. So the Opel should be either equal or slower..?
Calibra time trial: 6'58.601
Audi A4 time trial: 6'39.208
According to these results the AI are driving at grandma pace even if the player is driving a Formula Gran Turismo.
Test 2:
Player car: Honda Today G
Track: Endurance hall, Nurburgring 24 hour endurance.
In this test the A4 was replaced by the Audi TT-R DTM.
In addition, both the Calibra and the Audi had to overtake an ASL Arta Garaiya GT300.
Player time: 12'41
Calibra time: 7'32.863
Audi TT-R: 7'35.490
Final conclusions:
Earlier I did the tests in the TVR race, which showed no rubberbanding. Much like the TVR test, the AI seem to be completely oblivious about the player. If the player is very fast, they average 7'30. If the player is very slow, they average 7'30.
All evidence I find points to there being no rubberbanding in Gran Turismo 4.
PS. I have terrible quality video proof for this Nurburgring testing, if required.
- 1,608
- Here and there...
- SUBMANIAC
@PS2 GT4 PAL and NTSC really aren't different in how they play... ^^' the only difference is internal framerate. NTSC *tries* to play at a constant 60fps, while PAL is limited to 50. There's also a slight resolution difference, but the internal mechanics are the exact same 
Also, when you look at some record posted on youtube, 5.51 is actually pretty slow xD
This guy has a 4.39.890 time with the FGT as well ^^' Damn he's quick..
Also, when you look at some record posted on youtube, 5.51 is actually pretty slow xD
This guy has a 4.39.890 time with the FGT as well ^^' Damn he's quick..
Last edited:
- 1,143
- United States
- Jake-Racer1221
The PAL & NTSC versions should have the same identical system of Artificial Intelligence. You know Polyphoney Digital is not "great" at programming the rabbit affect. Gran Turismo is pretty infamous for it's terrible A.I, once the player is ahead of the A.I they somehow give up on catching up with the player, and they go faster once the player is behind, which is why you need a significantly slower car for a challenge.Yep... Id prefer to see the video(s) "with specs" and to see those wonderful driving skills of yours turning a 5.51... or shall I just assume the Pal version of the game is wayyyyy different than the NTSC
Yep... Id prefer to see the video(s) "with specs" and to see those wonderful driving skills of yours turning a 5.51... or shall I just assume the Pal version of the game is wayyyyy different than the NTSC
Thanks Sub & Jake for those explanations...
Edit Added: Just saw the Minolta video... Not in endurance hall that's for sure and also it doesn't seem to show the actual start... ugly video if ya ask me, and with all the cheat disks etc out there on ebay like code breaker etc as well as editing clues... I have to question it's integrity... Certainly others can see that some of the turns being made at certain high speeds would normally result in a crash... Just imho... One things for sure, my video replays are way more realistic and classy looking...
PS. I have terrible quality video proof for this Nurburgring testing, if required.
The video data on my phone seems to be practically unusable...
I shall redo the entire test just to prove my point.
Figures... What a shame, cause I was really really really looking forward to seeing that 5 minute run around Nurburgring in your Formula...
No worries, doesn't matter to me, cause I'm finished with any further posts on this thread anyway...
Last edited by a moderator:
BobK
Premium
- 7,020
- Massachusetts, USA
- 1,608
- Here and there...
- SUBMANIAC
Okay let's not escalate the subject to flaming ^^'
We've all demonstrated in our own way that there's no apparent rubberbanding in GT4.
@PS2 GT4 trust me on that, this level of grip with a 88C-V is very manageable with Super Soft race tires. True that the run is probably in arcade mode, with upped power and reduced weight, but it's definitely manageable. And he's definitely as well on the very edge of losing control most of the time ^^'
Btw, the video dates back from 2008, hence the bad quality.
We've all demonstrated in our own way that there's no apparent rubberbanding in GT4.
@PS2 GT4 trust me on that, this level of grip with a 88C-V is very manageable with Super Soft race tires. True that the run is probably in arcade mode, with upped power and reduced weight, but it's definitely manageable. And he's definitely as well on the very edge of losing control most of the time ^^'
Btw, the video dates back from 2008, hence the bad quality.
Last edited:
Obelisk
Premium
- 9,792
- United States
- RMedia_Obelisk
For someone who's supposedly older (and has been playing Gran Turismo for 12 years...), you should have known better than to criticize people who took the time to substantiate their claims and counter-arguments.
And @BobK simply said for you to do your research on the game.
By the way, the admins and moderators can still see what you posted, so changing all of them to "delete" is a moot point.
- 1,608
- Here and there...
- SUBMANIAC
Shows that maturity sometimes does not come with age.... >.>
Anyway, I've driven my Gray/black FGT on the 24h Nurb' twice just a few minutes ago, both times I went easy (and still made 2-3 mistakes here and there 'cause I didn't soften up the suspension and the car was bouncing around lol) and both were 5.56 and 6.03 lap times. And way over 1 minute ahead of the others (M3 GTR, Calibra, Mercedes 190E)...
Anyway, I've driven my Gray/black FGT on the 24h Nurb' twice just a few minutes ago, both times I went easy (and still made 2-3 mistakes here and there 'cause I didn't soften up the suspension and the car was bouncing around lol) and both were 5.56 and 6.03 lap times. And way over 1 minute ahead of the others (M3 GTR, Calibra, Mercedes 190E)...
