Inspired by Ryan Soucy's thread about AI inconsistency, and well, a lot of other threads about it too. I figured I'd make one thread where we can list all of the car/track/AI level combos that appear to cause issues.
For example, we all know Silverstone is a track where the AI are notoriously difficult seemingly for every class of car. So an entry for that would read:
In this scenario, +1.8 seconds is the absolute closest you could get to the pole time when the level is set at 100 percent difficulty and 50 percent aggression.
In order for this to work, it will require a lot of people doing the same track. It can't be one person does Spa, then someone else does Fuji, because we can't equate talent into a metric like this at all. What I'm proposing is that we make a list of 10 tracks, and all of us run a qualifying session at 100 difficulty / 50 aggression and post the results here.
While it's an imperfect system, it should give people who want to use PC2 a baseline for AI adjustment on a set number of tracks. It doesn't matter what your actual talent level is, as long as the results can interpret an equal talent across multiple tracks. Once we have 10 people on 10 tracks, we could assemble a chart and see what the average looks like. Then the results could be posted here and users would have a kind of workaround for at least a partial list of tracks.
Here is my initial list that I plan to begin testing later today. If anyone wants to join me in posting their results or sign up to helping with this project, post in here. I'm not particularly quick, so don't expect fast times. It's not a competition about speed though, it's about developing a community resource, which to me, is an integral part of what Race Department is about.
I will be using the Porsche Cayman GT4 and the Formula C (both) for these tests, but feel free to use whatever car you like as long as you specify what it is. However, I would suggest against using cars with little comparison to others in the game, such as Formula X / Aston Martin Vulcan / etc.
For example, we all know Silverstone is a track where the AI are notoriously difficult seemingly for every class of car. So an entry for that would read:
- Silverstone - [+]1.8 seconds
In this scenario, +1.8 seconds is the absolute closest you could get to the pole time when the level is set at 100 percent difficulty and 50 percent aggression.
In order for this to work, it will require a lot of people doing the same track. It can't be one person does Spa, then someone else does Fuji, because we can't equate talent into a metric like this at all. What I'm proposing is that we make a list of 10 tracks, and all of us run a qualifying session at 100 difficulty / 50 aggression and post the results here.
While it's an imperfect system, it should give people who want to use PC2 a baseline for AI adjustment on a set number of tracks. It doesn't matter what your actual talent level is, as long as the results can interpret an equal talent across multiple tracks. Once we have 10 people on 10 tracks, we could assemble a chart and see what the average looks like. Then the results could be posted here and users would have a kind of workaround for at least a partial list of tracks.
Here is my initial list that I plan to begin testing later today. If anyone wants to join me in posting their results or sign up to helping with this project, post in here. I'm not particularly quick, so don't expect fast times. It's not a competition about speed though, it's about developing a community resource, which to me, is an integral part of what Race Department is about.
- Silverstone (GP)
- Le Mans (Full)
- Brands Hatch (GP)
- Spa (GP)
- Red Bull Ring
- Knockhill
- Fuji
- Long Beach
- Laguna Seca
- Daytona (Road Course)
I will be using the Porsche Cayman GT4 and the Formula C (both) for these tests, but feel free to use whatever car you like as long as you specify what it is. However, I would suggest against using cars with little comparison to others in the game, such as Formula X / Aston Martin Vulcan / etc.
