• Welcome to the largest (sim) racing website in the world!
    Blurring the line between real and virtual motorsports.


Discussion in 'rFactor' started by saultenian, Jan 24, 2011.

  1. saultenian


    I have noticed many posts concerning AI settings on rFactor tracks, especially in regards to offline racing being too slow or too fast. I offer this simple tutorial that I use on many tracks I have loaded and find it very satisfying. If this is already known, then, well, ignore it!

    Open the .AIW file in notepad or wordpad and scroll down to the following lines:


    These are the AI strength settings for each track. The WorstAdjust is for 80% AI strength, MidAdjust is for 100%, BestAdjust is for 120%. If you run with the AI strength at 100% and they are too slow, increase the number (larger for faster), if they are too fast, decrease the number (smaller for slower).

    To test the changes, go to RaceWeekend, load the track and hit the “finish session” button. You will see the AI times increase or decrease as you change the AI settings. I tend to base this on my best time, but sometimes find that setting the AI at my speed tends to speed them up too much. I set them down slightly, as I like to run races with me at the back. In the end, this can be a bit tedious, but I have found that my offline racing benefits greatly when I do this. There can be BIG differences by just adding .1 to the number.

    Please note that you can go back to the default settings for the track by re-imputing the numbers from above, as these tend to be the settings that all tracks come with by default.
    Good luck!

  2. Samuli Leinonen

    Samuli Leinonen

    If you don't like all the AI cars using same racing line and would like to them behave a little less robotic, try to add some AI mistake from your .plr file.

    A range of (intentional) AI mistakes from 0.0 (none) to 1.0 (sometimes). Anything above 1.0 multiplies the frequency.

    Currenlty mine is setted to 2.0. It's nice when you're making videos and want to see some action on track. :cool:
  3. John Sydenstricker

    John Sydenstricker

    Thanks. I'll give this a try.
  4. April Dillon

    April Dillon

    I'll deffinately try out the mistake thing. Is there anything that can be done to make the AI more conservative when overtaking? They cause a lot of accidents by picking poor overtaking places, or just plain stupidity.
  5. Wayne Reed

    Wayne Reed

    I would think you might need to redo the racing line for them not an easy task but not to sure on that.
  6. Kris Vickers

    Kris Vickers
    Hardware Staff

    That would be pick a lower aggression setting.

    There are also settings in the PLR file you can play with;

    And change these for better racing offline too. This wont cause online missmatches like it would editing the AIW;

    These settings might not neccesarily work for you, but these are the ones you want to play about with :)
  7. Ken O'Keeffe

    Ken O'Keeffe

    Hi All,

    You might have all heard of this but have used a AI Learning guide i found on RSCNet forums by Rob,
    basically you change a setting in your player config file as follows,

    Autocalibrate AI Mode = 0 ( modify this to 1) save and exit. Remember to return this to 0 when finished.
    Start rFactor change settings to
    AI Drivers 1
    Race grid position 1
    Race start 8am
    Race time scale = none
    Mech failures = off
    Standing start

    and rules to
    Fags, fuel usuage, tire wear to off or none.
    Private testing = on

    Then choose vehicle you want to race and track, go in and add one AI driver.
    AI driver will say keep out of way human and start practising ( you can speed this up by pressing control T or A i think)
    Once AI is finished this will create a config file in that vehicles team folder ( you can move or copy this to the main vehicle folder so all the race teams using that vehicle can use)

    Remember to put everything back as it was and should help with AI driving better.

  8. Cioccolata1278


    Two questions.
    1. I am trying to do the "AI Learning" guide with National Stock Car mod. When I do all the steps, the computer player never gets anywhere near 0.0 meters. It always says "Off by 950.33 ie meters", never anywhere closer to 0.0. I race a second a.i. driver to get closer, but they never do. What am I doing wrong?

    2. When I watch them run there laps, they are never anywhere near the lines that stock cars take on race lines. They never race up near the wall on straightaways and take a near impossible line inside on corners. Is that part of the problem with them not getting close to the meter thing? Will this only work for road courses and such?
    Any help would be appreciated. I like racing offline and stock cars.

  9. jnl.42


    You have to let them keep running laps, as they run laps the distance will drop, and whenever you run a second driver he doesn't pick up where the other one left off he starts over.
  10. Cioccolata1278


    I let the driver run 36 laps and they never get any closer. The closest is 146.44 meters so far. I cannot find the config file in the cars I am using after they race as well.
    From the forum I read about doing this, they say if the first one does not make 0.0 meters, boot him, add a a.i. and he will finish off.
    So what am I doing wrong. I have run about 5 or 6 drivers, not at once. I do one, close rfactor down and when I cannot find the config file, I try again, but it never works.
  11. jnl.42


    When you start off one driver it creates an ini in that driver's folder so to get the other driver to pick up where he left off you'd have to go and cut-paste the ini into the other driver's folder. Strange that it doesn't get very close, maybe it's an issue with the track? Try it on other tracks and see how you go. And I'm pretty sure the config file goes into the folder where the .veh file is for the driver.
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.