- 376
- Nürburgring
- lukasspeed
As a follow-up to my previous test I ran another one with a bit different setting: It's again the 10 lap Ascari race in S-class, the car is again the 585 PP MP4-12C on sports soft (ABS:1, else off if that matters). I made sure to have again the Ferrari FXX, the Ferrari Enzo and the McLaren MP4-12C as leading AI, but couldn't get the exact same grid order, Bugatti on pole again though.
But this time my goal was to drop back to a 30 sec gap to the leader during the first laps and then maintain that gap of 30 sec until the end of lap 8. After the pitstop I came a little close as the gap dropped to 23 sec until I got the first split time to adjust to 30 sec again.
So at the end of lap 8 I was in 6th place ca. 30 sec behind the leader and only 2 laps to go. If the AI looks for my pace and slows down to let me win, at that point I would look deeply in need of it. Of course I wasn't and took the lead midway in the final lap.
These were the laptimes (hoping it works with formatting):
laps me 12C FXX Enzo
01. 2.27 2.20 2.17 2.21
02. 2.26 2.24 2.22 2.24
03. 2.27 2.24 2.28 2.25
04. 2.30 2.26 2.27 2.25
05. 2.32 2.22 2.22 2.24
06. 2.35 2.28 2.28 2.37
07. 2.28 2.41 2.51 2.50
08. 2.28 2.27 2.27 2.29
09. 2.06 2.28 2.27 2.30
10. 2.08 2.28 2.29 2.30
Total racetimes:
1. 24.24 me
2. 24.35 MP4-12C
3. 24.43 FXX
4. 25.01 Enzo
So while the race was a bit slower overall then my other tests, I don't think it's really significant. I also don't see any spooky things in the laptimes. In my first test the AI was a little quicker in the second part (to be precise in laps 6-8, later they were slower) despite my laptimes being very very similar in the first 5 laps and only laps 6-8 very different. Still: At the beginning of lap 9 I was so far back I see no chance the game could believe I still had a chance at that point.
So there might be some minor adjustment to player speed in that s-class race, but nothing really significant. It could be they speed up when you are close?
While maintaining my gap and cruising between the AI I could watch them a lot though. I can't remember the GT5 AI having so much trouble with worn tyres. If I remember right in GT5 they became very slow with worn tyres but could drive them basically down to 1 before making mistakes from it. In this race in GT6 the Ferrari 458 couldn't make it out of any slow corner without half spinning the car from on throttle oversteer after lap 3 on a set of tyres. Others with more power are even worse. They step on throttle without any "feel" for it. These mistakes create a big random factor in such a race, because they effect the other cars around too, who can either pass or are held up. One would have to run a lot of races to study any systematic in that.
[Edit:] After thinking more about it, my theory is AI is at least in S-class similar to GT5 b-pec Bobs. They overheat, cool down too much and they can get tired towards the end. They could even have different "character" or level each race and the combination of Bob-character and grid puts the random element in those races.
But this time my goal was to drop back to a 30 sec gap to the leader during the first laps and then maintain that gap of 30 sec until the end of lap 8. After the pitstop I came a little close as the gap dropped to 23 sec until I got the first split time to adjust to 30 sec again.
So at the end of lap 8 I was in 6th place ca. 30 sec behind the leader and only 2 laps to go. If the AI looks for my pace and slows down to let me win, at that point I would look deeply in need of it. Of course I wasn't and took the lead midway in the final lap.
These were the laptimes (hoping it works with formatting):
laps me 12C FXX Enzo
01. 2.27 2.20 2.17 2.21
02. 2.26 2.24 2.22 2.24
03. 2.27 2.24 2.28 2.25
04. 2.30 2.26 2.27 2.25
05. 2.32 2.22 2.22 2.24
06. 2.35 2.28 2.28 2.37
07. 2.28 2.41 2.51 2.50
08. 2.28 2.27 2.27 2.29
09. 2.06 2.28 2.27 2.30
10. 2.08 2.28 2.29 2.30
Total racetimes:
1. 24.24 me
2. 24.35 MP4-12C
3. 24.43 FXX
4. 25.01 Enzo
So while the race was a bit slower overall then my other tests, I don't think it's really significant. I also don't see any spooky things in the laptimes. In my first test the AI was a little quicker in the second part (to be precise in laps 6-8, later they were slower) despite my laptimes being very very similar in the first 5 laps and only laps 6-8 very different. Still: At the beginning of lap 9 I was so far back I see no chance the game could believe I still had a chance at that point.
So there might be some minor adjustment to player speed in that s-class race, but nothing really significant. It could be they speed up when you are close?
While maintaining my gap and cruising between the AI I could watch them a lot though. I can't remember the GT5 AI having so much trouble with worn tyres. If I remember right in GT5 they became very slow with worn tyres but could drive them basically down to 1 before making mistakes from it. In this race in GT6 the Ferrari 458 couldn't make it out of any slow corner without half spinning the car from on throttle oversteer after lap 3 on a set of tyres. Others with more power are even worse. They step on throttle without any "feel" for it. These mistakes create a big random factor in such a race, because they effect the other cars around too, who can either pass or are held up. One would have to run a lot of races to study any systematic in that.
[Edit:] After thinking more about it, my theory is AI is at least in S-class similar to GT5 b-pec Bobs. They overheat, cool down too much and they can get tired towards the end. They could even have different "character" or level each race and the combination of Bob-character and grid puts the random element in those races.
Last edited: