I suspect many of the differences we're setting have little to do with hardware configurations, but can be accounted for, with considerable detective work, with settings deviations. This is why I recommend using the presets (fastest, fast, normal, etc.) and the default game settings (light flares off, vsync off, everything else on). They are easily obtained, easy to return to, and easy to remember. Unless you have modified your config file through the console (I disabled the blue line at the bottom of the screen, as I found it distracting), it'll give everyone a better idea of how we're all doing.
Unavoidable and desirable variables (hardware specs, vanilla or Max or G200, clock speed manipulations, resolution, color depth, etc.) should, of course, be posted. Vsync should always be disabled and the *latest* version of the game should be used.
Everybody has favorite settings for playing and benchmarking, for whatever reasons, and that's great. But as it stands now, as far as our methodology goes, we're having both "validity" and "reliability" issues. We are measuring different things, and we're having problems reproducing results from tester to tester.
The game provides us with a series of preset standards or defaults or whatever you want to call them. If we used them, I believe it would make the task at hand, comparing results, so much easier.
Paul
paulcs@flashcom.net
Unavoidable and desirable variables (hardware specs, vanilla or Max or G200, clock speed manipulations, resolution, color depth, etc.) should, of course, be posted. Vsync should always be disabled and the *latest* version of the game should be used.
Everybody has favorite settings for playing and benchmarking, for whatever reasons, and that's great. But as it stands now, as far as our methodology goes, we're having both "validity" and "reliability" issues. We are measuring different things, and we're having problems reproducing results from tester to tester.
The game provides us with a series of preset standards or defaults or whatever you want to call them. If we used them, I believe it would make the task at hand, comparing results, so much easier.
Paul
paulcs@flashcom.net
Comment