It's a bit nit picking here - I can see your point, but in all honesty collecting all that data is a bit too much.
It's just, battlefield 4 @100fps feels really terrible compared to quake or counterstrike at 100fps, you really can't tell i guess unless you compare the two - like playing on a console even. I was very dissapointed by how unsmooth it was
If Frametime variances don't matter, then anything over 60fps is borderline irrelevant on a 60hz screen, and getting 70 or 80 doesn't really matter. Only reason 100 feels more smooth than 70 is because of dips, yknow?
I don't usually check them, only in games that perform badly, and bf4 is one of them especially the beta on win7 - i didn't get to try anything else
Since it's so easy to collect if you're already benching (a few megabytes from hours of benching that's just taken automatically) i'd appriciate you throwing them at me, if it performs good i'll buy it. If it performs "meh" i'll probably still buy it, just less enthusiestically because i like my FPS games to be smooth and nicely responsive, as little of a step down from quake live as possible
THE MOST IMPORTANT DATA, in my opinion in the whole sampling of my data is the variances between the AVERAGE FPS on Win8.1 and win7
This is what i'd disagree with the most - If i can say, like in that shot - 3-5% of my frames failed to be rendered within 16.7ms (the refresh period on a 60hz screen) yet changing to windows 8 made it so only 1-2% of the frames failed to meet 60fps mark, that would feel like a massive upgrade in smoothness that i could blind test i feel, though average FPS change might be small or nonexistant
I'm not saying put a ton of effort into interpreting and showing such info - Just throw me a few files that are alongside the benches that you're taking anyway, i'll look into it and if there's anything really solid worth reporting on, it's easy, good, and occasionally really important info Edited by Cyro999 - 11/12/13 at 3:54pm