Originally Posted by Kropotkin
Pretty much what Skowinski and bk brkr baker said. He did things that he just wasn't supposed to be able to, like jumping off a 250 and going straight on to win a 500cc race. Dennis Noyes tells a story about Erv Kanemoto, his crew chief at the time. Erv would send Spencer out with a set up, Spencer would go out, ride around for a bit, hit lap record pace, and come back in again. Erv would ask "So, how did it feel?" and Freddie would say "Pretty good, I guess." So Erv would make some changes, send Spencer out again, and Spencer would hit the lap record, come back in and tell Erv that that was "Pretty good." There was almost no way of knowing whether the bike was better or worse, Spencer was lapping at the same pace, it would sometimes take him a couple of laps extra, but there was no difference in times. These were the days before datalogging, of course.
So Erv figured out that Spencer would wear a fresh helmet every time he went out on a new run. Erv then started checking Spencer's helmet for sweat: if the helmet was drenched, Spencer had worked hard to make the lap time. If the helmet was not too sweaty, Kanemoto knew he was pretty close with his set up.
What's amazing about this is that Spencer was capable of running lap record pace despite the bike set up, not because of it. He always rode to the limit of the tires, it just took him a little longer sometimes to figure out how to reach the limit of those tires.
To me, Stoner was capable of achieving the same, with the key difference that he could also put into words what was working or not. Look at his time on the Ducati, and you see he could find a way of reaching the limit of the tires, where other riders on the bike could only ride to the limit of the bike. That's a pretty remarkable achievement.