Originally Posted by 2advanced
If it's a reputable site, then why do they have the Non-Mantle results for Thief and BF4? Again, the problem with all this elementary style benchmark toting is that without a few bad ones to muddy the results, you're simply back at intentionally trying to bash an otherwise good performer.
Simply removing 3 of the results from that horrid graph changes the difference by an outstanding 7% between the i5 and FX.
I personally feel offended that you would think that we're stupid enough to believe that this kind of garbage is representative of anything other than a failed attempt at ballooning the performance gap.
They do have mantle results for BF4 and Thief:
(these however were obviously made after the games were launched because the mantle patches were late)
Not that it really matters, mantle makes the 4670K win over the 8350 in BF4 instead of losing by a per cent or two. And in Thief the 4670K's advantage closes from 50% to 30%. So in one game mantle made it worse for the 8350 compared to the 4670K and in one it made it better. The average percentages would still be the same give or take a per cent.
Besides it's 23 games, all the latest ones. And it doesn't leave out any of them. So it's by definition not trying to muddy the results or skew the numbers. That spreadsheet is just taking all the games available on that site and averaging the results. It's about as objective as you can get.
But of course, the argument is that we should remove some of the games. I'm sorry but since I like to be objective I included all the games. And as I said in my first post in that thread, that's the advantage of the intel offerings. You don't have to remove any games, they give you consistently good performance.
What do you think we should use for comparisons between the parts?
Only certain games? Is there anything wrong with using all of them? If you want to remove outliers you also have to remove them from the other end of the spectrum...Edited by Alatar - 3/31/14 at 8:03am