What variables are used to choose the AI tires in S series?

[AstonDB5]

Hey guys, I beat the game a while ago and platinumed it, so like most, I usually just run free run or S series races over again. The race in question is the 10 laps of Silverstone. I did the race last week with my new XANAVI NISMO GT-R (If you dont already have this car.. buy it.. now..). I had Racing mediums on, and I took first place by lap 4. The AI in 2nd all race was the '97 BMW McLaren F1 GTR Race Car. I won the race with roughly a 25 second gap between us. I ran the race again today with my new Citroen GT Race Car on Racing Hards. The AI in the lead was using the same F1 GTR but this timw he was still 20 second ahead by lap 8. I narrowed his lead to 7 seconds by the end of lap 10 and thought that he was going ridiculously fast, considering how much I beat him by last time, and how close my lap times were in both cars (1:53 average in NISMO, 1:56 average in Citroen). So to lose by that much caught me off guard. I went in to the replay to look at both races data, his lap times in both, and tires used. Low and behold, when I used the GT-R with racing mediums, he had Racing mediums as well and was averaging 1:58. When I used the Citroen, he had Racing Softs and was averaging 1:53! I was under the impression that unless it was raining, the AI used Racing Hard. But I also have no idea, just figured that. So, is there any variables that decide what they use? Is it based on the players tires? Car? PP? or is it just random?

Thanks for any info guys!

 

[NOSWaster]

If you won the race by 25 seconds, then lost by 7 seconds, look at your lap times. 3 seconds per lap over 10 laps is a 30 second loss. Considering there was a 32 second disparity, that result makes sense.

The F1 GTR was most likely using the exact same tires as in the first race, which to me is sounding like it was on Racing Mediums and driving really slowly for that compound.