Bug GT1 AI uneven pace spread

Discussion in 'Community Support' started by MattStone, Mar 1, 2020.

  1. MattStone

    MattStone Well-Known Member

    Joined:
    Mar 6, 2016
    Ratings:
    +169 / 0 / -0
    The GT1 class is my favorite class of car atm but after many hours of racing this class I've noticed (and have put up with) the large difference in pace of the field AND uneven grouping

    And it's not just happening on one track; I think it's all tracks
    at first I thought is was because of incidents but I don't believe this is the case

    I race 25 opponents, use the full 20 minute qualy time and a 40 min race at 114ish strength

    That pattern I'm seeing is this and applies to both quali pace and race pace
    The pace of the cars are clumped into 3 groups
    Top 3
    close together even close spread
    Pos 3-6 close together with an even close spread but around 1 sec slower than the top3
    The rest of the field starting at position 7 are then evenly spread BUT position 7 is 1-2 seconds behind position 6

    The result is you get 3 different races in one and by the end of a 40 minute race there ends up been a HUGE gap between each group. If my strength is in between of one of these gaps I end up having tha cars in front of me pulling away at 0.5 seconds per lap whilst I'm pulling away from the car behind me at 0.5 seconds a lap; a very lonely race :(

    The only way to have a fun competitive race with the AI in this class is to find an AI strength that places you in the bottom 1/4 of the field

    I love this class of car and it's one of the few classes that don't suffer from the dredded cars spinning bug so I can play most of my tracks

    I've attached screen shots of the qualy pace of the field after a 20min qualy; at no time did I hold up the field.
     

    Attached Files:

  2. Maskerader

    Maskerader Well-Known Member

    Joined:
    Oct 6, 2019
    Ratings:
    +352 / 0 / -0
    Are these 1-3 and 4-6 guys always the same? Like, Martin+Lee+Simpson are the top 3, and so on?
     
  3. MattStone

    MattStone Well-Known Member

    Joined:
    Mar 6, 2016
    Ratings:
    +169 / 0 / -0
    good question; tbh I haven't taken notice of that
    Next time I race I'll check that out
     
  4. Maskerader

    Maskerader Well-Known Member

    Joined:
    Oct 6, 2019
    Ratings:
    +352 / 0 / -0
    On your screenshot I also noticed that all top6 are in Koenigseggs (while Rufs and Gumperts are mainly on the back of the field). Might also be some BOP issues, not only with AI?

    I wonder what happens if you start a race with a custom selection of cars where there are no Koenigseggs at all.
     
  5. MattStone

    MattStone Well-Known Member

    Joined:
    Mar 6, 2016
    Ratings:
    +169 / 0 / -0
    I'll give that a try tonight but even in those top 6 Koenigseggs you still have the 1 second game which is pretty consistent!
     
  6. MattStone

    MattStone Well-Known Member

    Joined:
    Mar 6, 2016
    Ratings:
    +169 / 0 / -0
    Ok I think it is a BOP thing
    Selecting just one car solved the spread problem

    THE TEST
    10 min qualifying with just one brand of car on track
    AI strength 114 with 24 cars,
    player stayed in garage to avoid interfering with ai

    RESULTS OF 1st PLACE CAR
    Koenigseggs 1:31.1

    0.5 second between 1st and 2nd but even spread after that of around .1ish of a second)
    Saleen 1:34.9 (even Spread)
    Grumpert 1:35.4 (Even spread)
    Nissans 1:35.5 (even spread)
    Ruff 1:35.6
    0.4 between 1st & 2nd and then closer gaps after that
    A couple of groups of two to three cars close together holding each other up)
    Ford 1:35.6
    Pretty even spread but a larger gap between 1st and 2nd​

    GROUP TEST WITH OUT KOENIGSEGGS
    I did a final test of 24 cars with an even balance of makes of cars but without the Koenigseggs
    RESULT
    Nice even spread but predictable as per the results above; the saleen held the top 4 but the rest fell very close behind
    Noted that if a saleen came out of the garage behind a slower car he was held up
    Slower cars did tend to group up the faster cars behind it resulting in a mix bag in the bottom 2/3rds of the pack
    The longer the qualifying session the more chance the faster cars were able to put in an unimpeded flying lap resulting in a final qualifying position more indicative of the results of the cars BOP highlight in my individual car tests above

    I than qualified in 4th position and done a 40 minute race
    Closest racing I've ever had in the GT1s, it took 10 minutes to get into position 2 only to have it taken back from me and in the end came 2nd... I'm still grinning from the fun :D

    So well done @Maskerader for seeing what I'm 95% sure is the problem

    So BOP down the Koeny a lot and the Saleen (just a bit) and it will be perfect
     
    • Like Like x 1
    • Useful Useful x 1
    Last edited: Mar 3, 2020
  7. Maskerader

    Maskerader Well-Known Member

    Joined:
    Oct 6, 2019
    Ratings:
    +352 / 0 / -0
    Glad I could help. :)
     
  8. MattStone

    MattStone Well-Known Member

    Joined:
    Mar 6, 2016
    Ratings:
    +169 / 0 / -0
  9. MattStone

    MattStone Well-Known Member

    Joined:
    Mar 6, 2016
    Ratings:
    +169 / 0 / -0
    @Alex Hodgkinson

    OK after doing several hours of GT1 racing without having the Koenigseggs participating I'm finding that even with the AI set to maximum strength of 120 I'm still able to start at the back and work my way to 1st place in 25 min race on most tracks

    With a field of Koenigseggs I can work my way up to about 10th place starting from the back

    So I don't think we need to lower the BOP of the Koenigseggs, we need to lift the performance of all the others to match the Koenigseggs; and plus a little bit more.

    My racing strength is normally around 110-116 with other series of cars
     
    • Agree Agree x 2
    • Useful Useful x 1
  10. Alex Hodgkinson

    Alex Hodgkinson KW Studios Developer

    Joined:
    Jun 5, 2017
    Ratings:
    +1,948 / 0 / -0
    Hello! Thank you for this thread, very useful data here I can work with.

    Unfortunately, there is no way you can tweak this yourself as it's all defined behind the scenes.

    I have put it on my list of things to do though.
     
    • Like Like x 1