JBirney-
Since when does a game have to be good for its benchmarks scores to be valid?
I was replying to-
Don't you get frustrated when even a 9600XT beats a more expensive 59xx series Nvidia card in DX9?
The only game out we have seen this in is TRAoD(even if we include FarCry). In order to be frustrated by the 9600XT besting the 59xx boards you have to be playing TRAoD- do you own that game? It came with my Audigy2, and it is incredibly bad by any standard.
Was Q3 a good game (note not the engine, the game itself)?
I liked Quake3 better then UT.
Halo shows the same results when you use Custom time demos. But yet people also through out those results because its PC version was a "bad game"
Halo doesn't show anything like TRAoD(9600XT besting a 5950), although you should dismiss the way in which Halo is 'custom' benched. The AI in Halo isn't on a set path and it will be a random encounter each time you play through.
oldfart-
If a game is poorly coded, maybe
Play the game is all I can say to that. Check out the, errrr, 'stunning' visuals
The performance TRAoD offers versus the visuals is easily the poorest trade off I have seen in a couple years(and this is running on a R9800Pro), it is a bad engine.
KOTR may be an example of bad coding. Here is a post from B3D that explains the problem. I dont play the game and have no personal knowledge if this is accurate.
I can only assume that that post is about NWN, not KoTOR.
So more than a month after BioWare had stated ATI didn't support a Pixel shader for shiny water, we were finally playing with shiny waters.
Shiny water was never a problem with KoTOR, it was a NWN issue. For the record the NWN engine is screwed up, for some time the Ti4200 was besting the highest end R3x0 and NV3X board in benches.