AI difficulty being inconsistent after the last patch

[Arfnix]

Has anyone had this problem? Had consistent pace versus the AI in Moto3 (before last patch), consistent pace in Moto2 (after last patch), but in MotoGP while the AI is laughable at some tracks, it's completely overkill at others. Examples:


- A1 Ring: playing it ridiculously safe and with an underpowered bike, I was lapping over 2.5s faster than the AI and totally dominated the race, my time in the rain was half a second faster than the AI times in the dry. ~103% IIRC.

- Mugello: the AI completely wiped my as­s in practice. I was playing in 105%, had to lower to 70% and I still wasn't even getting close to the development tests times (!), and Mugello is a track in which I perform well, unlike A1 Ring.

In the race, after setting the difficulty to its original value again, I could still get a podium because after the patch the AI is overpowered in qualy and underpowered in the race. No joke but with fully developed bikes, in practice, I wouldn't be surprised if the AI started lapping world record-like time, because the gap in performance between the AI and myself was around the same gap between me and the WR guys in time attack.


Taking this with the brilliant observations an user made, on this same forum, about the times for the development programs being completely out of what they should be, it's easy to conclude the game should be drastically "recalibrated." I am wondering if more people are going through this same situation since I did not see any other complaints about it.

Have fun and may the Rossi be with you.

[WOEaintME]

I find the AI, like myself and the actual riders in GP, have their good tracks and their bad tracks. Lorenzo is not great at Assen, 1 win and only 3 podiums in all his years of MotoGP. At Germany he's arguably even worse, never having won. By comparison, he only finished off the podium at Catalunya twice, both times 4th (though he has had some DNFs, including last year's). He never finished worse than 2nd at Mugello on the Yamaha.

You won't be great at every track. But at some you will be a nigh unstoppable force. It's just racing.

[abrensons]

I have not reached motogp yet but 2.5 secs advantage over fastest AI feels too much to have some "reasonable" explanation.

[Arfnix]

And now I won Misano at 120% by 7 seconds over MM and over 12 seconds on the rest of the field, didn't change anything in both driving and setups and my race was full of errors (I even recorded it but I don't see the need of posting it), so it was the same technique that went 3.0s seconds slower than the AI per lap at Mugello and dominated Austria. Things are pretty crazy.

Austin was another one in which I recall to be quite dominating as well but I won't remember if I bumped the AI to 120%, I think I didn't.

[Arfnix]

LCR Idemitsu btw.

[Luska]

Arfnix님이 먼저 게시:

And now I won Misano at 120% by 7 seconds over MM and over 12 seconds on the rest of the field, didn't change anything in both driving and setups and my race was full of errors (I even recorded it but I don't see the need of posting it), so it was the same technique that went 3.0s seconds slower than the AI per lap at Mugello and dominated Austria. Things are pretty crazy.

Austin was another one in which I recall to be quite dominating as well but I won't remember if I bumped the AI to 120%, I think I didn't.

Same here with me at Misano. The AI seem very very slow at that track.

[Arfnix]

And at Aragon it went to Mugello levels again :P. I am back to MotoGP 15 until the bugs are fixed really — if they are. Really like both MotoGP 20 graphics and physics but am definitely not patient enough to keep waiting for the patches and tweaking difficulty, also if the AI and the game difficulty overall changes at every update, it may not be a good idea to run a career mode since it's hard to keep it coherent: "this difficulty here, average finish P3 and patch and then average finish P17, where do I tweak the difficulty for now? Wait that track is different..."

I will certainly be very enthusiastic to test it once the game goes to its final state though, assuming the game is well calibrated again and the development test times are fixed according to what @WOEaintME said in the other thread.