Post by Deleted on Sept 4, 2014 15:06:41 GMT
So, I've been wanting to do one of these for a while. Mainly because I'm not pleased with most of the "testing" that's being done atm. Broughy and GUN1T both run perfectly valid tests, but I feel they are still lacking some input, which is why I will be doing something a bit different. Instead of just running laps on the same track over and over, or having a bunch of different tests for acceleration, damage, torque and top speed to give the cars validity, I want to be testing them in terms of how they are actually used over time as you get used to them. The idea is to run a fully modded version of every single car and motorcycle in the game through 10-15 minutes, over a large variety of tracks.
So far my idea is that I'll start out by using some of my own tracks that I know well, I'll be using my Divine Nine playlist which I have fairly consistent lines on. However, as these tracks are built for super cars they wont alone be offering the full picture. That's why my aim is to over a longer time period, do approx. 100-150 different tracks with each vehicle. I will be recording each run, and note down which car, the total time, the best lap time and the average lap time. (Yes, both total and average lap)
I'm then aiming to rate the cars mainly based on their avg. finish on the list in terms of the best lap time. With a seperate list of ratings that will go by their total time added up through all the races done. And then use the average lap times to see if there is any patterns as to what type of track the various cars excel on, and provide a rating based on the type of track as well. I'm assuming it will be similar types of cars excelling at similar types of tracks, and some excelling as good all-over cars.
I'll also try to get a list over which cars have the most first places and such.
The advantage of doing it this way is that it provides testing across the classes, and it provides a much larger statistical analysis, whilst not going too indepth. Considering my familiarity with the tracks, and the amount of tracks done, any accidents can then be considered such as lack of consistency in the car. And hopefully when done it will provide a much clearer picture on which cars are most consistent, which cars are fastest (regardless of consistency), etc, etc... It should also then be quite easy to maintain as new cars come out, as I'll just need to run the tracks with new car as well. I'll be starting out with the sports class, and then I'll be moving on to other classes afterwards.
So that's my idea, however, ideas are often flawed. What I need your help to, is finetuning this idea into something that you all would appreciate and find useful and trustworthy information. I am also wondering wether or not you think I should do the same thing racing with other people, through the various categories of racing (GTA, Rally, Non-Contact and Standard) and if you think this extra factor of having other drivers knocking you about is an important one to measure on the cars as well. As some cars could prove to be more consistently stable when knocked around, whilst others go flying, and that could also be an important factor affecting your choice.
Basically I just want input, how do YOU think I should be doing this to get the most reliable data? I know it's a big project, but I'm not planning to finish it any time soon either. It's something I believe I would enjoy doing, and I'm willing to take it on so long as I can keep it simple like this, as it's basically just running a playlists with different cars and recording each race. And how big do you think my data collection should be before it starts offering information of real value?
Cheers!
- Fachuro :-)
So far my idea is that I'll start out by using some of my own tracks that I know well, I'll be using my Divine Nine playlist which I have fairly consistent lines on. However, as these tracks are built for super cars they wont alone be offering the full picture. That's why my aim is to over a longer time period, do approx. 100-150 different tracks with each vehicle. I will be recording each run, and note down which car, the total time, the best lap time and the average lap time. (Yes, both total and average lap)
I'm then aiming to rate the cars mainly based on their avg. finish on the list in terms of the best lap time. With a seperate list of ratings that will go by their total time added up through all the races done. And then use the average lap times to see if there is any patterns as to what type of track the various cars excel on, and provide a rating based on the type of track as well. I'm assuming it will be similar types of cars excelling at similar types of tracks, and some excelling as good all-over cars.
I'll also try to get a list over which cars have the most first places and such.
The advantage of doing it this way is that it provides testing across the classes, and it provides a much larger statistical analysis, whilst not going too indepth. Considering my familiarity with the tracks, and the amount of tracks done, any accidents can then be considered such as lack of consistency in the car. And hopefully when done it will provide a much clearer picture on which cars are most consistent, which cars are fastest (regardless of consistency), etc, etc... It should also then be quite easy to maintain as new cars come out, as I'll just need to run the tracks with new car as well. I'll be starting out with the sports class, and then I'll be moving on to other classes afterwards.
So that's my idea, however, ideas are often flawed. What I need your help to, is finetuning this idea into something that you all would appreciate and find useful and trustworthy information. I am also wondering wether or not you think I should do the same thing racing with other people, through the various categories of racing (GTA, Rally, Non-Contact and Standard) and if you think this extra factor of having other drivers knocking you about is an important one to measure on the cars as well. As some cars could prove to be more consistently stable when knocked around, whilst others go flying, and that could also be an important factor affecting your choice.
Basically I just want input, how do YOU think I should be doing this to get the most reliable data? I know it's a big project, but I'm not planning to finish it any time soon either. It's something I believe I would enjoy doing, and I'm willing to take it on so long as I can keep it simple like this, as it's basically just running a playlists with different cars and recording each race. And how big do you think my data collection should be before it starts offering information of real value?
Cheers!
- Fachuro :-)
