AI tyre strategies and wear (in 100% race?)
Hi,
I've recently switched to 100% races.
And a very very strange behavior is that I noticed the AI, when choosing a 1-stop-strategy, goes for a longer stint on medium than on hards.
For example my Monaco 100% GP.
The suggested Strategy said
- Start on hard
- Pit in Lap 33 for Mediums
- Run Mediums for 45 Laps (or so) until the end
Now that seemed to be very weird to me, so I switched it around. Started on medium and then to hard in lap 33.
Now, in race, I had to actually pit 2 laps early because my mediums were done (around 50% but very slippery).
But the AI seemed to follow the strategy stated above. Almost all started on hards and pitted on the same lap or one lap after me.
And then they ran for 45 Laps on Mediums. Which means their tyre degredation must have beeen between 70-80% in the end, according to my data. Even my teammate did it.
Result was:
- In the first few laps, I was a little faster than the cars around my performance index, then it switched around.
- After they pitted for Mediums, they were significantly faster than me on hards.
- They stayed substantially faster until the very end of the race.
I lost around 1-2 seconds a lap to my teammate, and around 2-3 to other cars in my performance range.
This was so frustrating, as in lap 31 a Williams (last in performance charts) overcut me by one lap, and managed to pull a 40 second gap by lap 60.
Btw, I qualified P17 (around 0.5sec faster than my teammate) and I hang along quite well at P12 in the beginning, so I don't think it was a problem of AI level being too high. Maybe a little too high but I like the challenge, for qualifying it was perfect though ... race pace they seem to be faster than me in general, because I drive more carefully (no flashbacks), but definetly not 2 seconds slower)
In the end, I decided screw it, Im going to pit again for softs, because the cars behind me on mediums where piling up and I was defending all the time. Which made no sense, I hoped that by that time the mediums would be so degraded that my hards would be faster (as it was that way for me when they were on hards in the first stint).
On the new softs I managed to go around a second a lap faster than those guys, and also due to a lot of lapping by the leaders, I managed to get my old place back. (minus Vettel, who overtook me around lap 56 on 25 lap old mediums and managed to pull a 12 second gap in 4 laps). The distance to him in the end was 27 seconds, so if I calculate around 20 seconds for the pitstop, I only managed to gain 5 seconds to him in 17 laps on new softs vs. 40+ lap old mediums.
So to summarize:
- Suggested Pit Stop Strategy says "go 31 laps for hard, 47 for mediums, even if they will be at 80% wear in the end -> When I switch it around the expected time goes down significantly.
- AI takes this Pit Stop Strategy and seems to have no issues with Tyre Wear, even on 47 (!!) lap old mediums (they still pulled low 1:16s in the end, my best lap on fresh softs was 1:14.6 with a little traffic). In the end with worn softs I did about the same times as the ones on the 75+% mediums, and my softs were at about 40% wear)
- If I had done the same I would have gotten a puncture, at lap 31 I had 50% tyre wear and was driving 1:18-1:19s when the others were driving high 1:16s on hards)..
- The significant pace difference was not or not significantly related to AI Difficulty, because in the beginning with mediums vs. AI's hards, I was able to follow Magnussen in the faster Haas and stay within 1-2 seconds to him, while building a gap of 2-3 seconds to the Aston Martin behind me)
I don't know if this is only an issue in 100% race, but I figured it might be, because the issue has not been brought up as far as I see.
But this kinda proves to me that the AI has significantly (!!) less tyre degradation than the player.