Ok, i have a bml editor. In which the file changes ai riders skills? I want to slow or accelerate the rider.
find someone willing? Thanks
find someone willing? Thanks
MGP17 Changing skills of riders
- Thread starter Matej Hanák
- Start date
Hey, so you have convert the data.mix to a folder, then in the folder go to data> vehicles>AI and then open the SKILLSCATALOG.BML
And edit the attributes.
Here's my reddit post about what the values mean. I've been playing with it for a week or so now, it is interesting.
Basically, Raceperformance determines their lap times per race when you are racing.
The highest value is 1, having it higher does not do anything, lower reduces race lap times.
Error Probability is how likely they are to crash.
Aggression is how likely they are make tough passes or dive up the middle. I like it set to 5, which is high but it makes racing fun.
Braking is always 1. I havent' played with it.
Time Generation is how fast the rider can go if you simulate time or skip qualifying.
Race Time generation is how fast the rider can go if you simulate a race, for example if you retire from a race in the middle of it. 1 being the fastest speed and the number the slower they'll be. I put Tito Rabat to .97 to test and retired from a race. His fastest lap was 20 seconds faster than the rest and he was able to win.
https://www.reddit.com/r/motogp/com..._determines_ai_skill/?st=jc122193&sh=a7784624
And edit the attributes.
Here's my reddit post about what the values mean. I've been playing with it for a week or so now, it is interesting.
Basically, Raceperformance determines their lap times per race when you are racing.
The highest value is 1, having it higher does not do anything, lower reduces race lap times.
Error Probability is how likely they are to crash.
Aggression is how likely they are make tough passes or dive up the middle. I like it set to 5, which is high but it makes racing fun.
Braking is always 1. I havent' played with it.
Time Generation is how fast the rider can go if you simulate time or skip qualifying.
Race Time generation is how fast the rider can go if you simulate a race, for example if you retire from a race in the middle of it. 1 being the fastest speed and the number the slower they'll be. I put Tito Rabat to .97 to test and retired from a race. His fastest lap was 20 seconds faster than the rest and he was able to win.
https://www.reddit.com/r/motogp/com..._determines_ai_skill/?st=jc122193&sh=a7784624
Oops... did not want to post anything, at first.
I just read this "error probability", thought - what is that? I completely missed it! Very useful addition. Of course, default values are too small. Realistic ones, for most crashing riders, should be around 0.36 - 0.42! I mean riders like N.Antonelli in Moto 3, S.Cortese, L.Baldassarri in Moto 2. For the likes of J.Mir - yes thats OK, ~ 0.05.
They even changed respawn time to 8 secs for AI. So AI riders really lose something when they crash.
Hopefully next step in the future will be full sequence scene - rider runs to bike, etc. like in SBK2000.
I just read this "error probability", thought - what is that? I completely missed it! Very useful addition. Of course, default values are too small. Realistic ones, for most crashing riders, should be around 0.36 - 0.42! I mean riders like N.Antonelli in Moto 3, S.Cortese, L.Baldassarri in Moto 2. For the likes of J.Mir - yes thats OK, ~ 0.05.
They even changed respawn time to 8 secs for AI. So AI riders really lose something when they crash.
Hopefully next step in the future will be full sequence scene - rider runs to bike, etc. like in SBK2000.
Last edited:
