Incorrect performance comparison

[laurence1211]

I dont think these are correct, G-O-M seems to be a bit off comparing my gfx.

Dark Messiah of Might and Magic
Publishers Recommended System
Show full details.. What is this?
Processor: Intel Pentium 4 3000 Mhz / AMD Athlon 3000+
Comparing your CPU Intel(R) Core(TM)2 CPU T7200 @ 2.00GHz (running @1997MHz) with the requirement AMD Athlon(tm) 64 (running @1800MHz)
• Your processor is significantly more powerful in raw performance
Display Card: DX9-compliant 256MB
Comparing your display card NVIDIA GeForce Go 7900 GS with the requirement NVIDIA GeForce 6800
• Your display card is slightly less powerful in raw performance
Memory: 1024MB
You have 1024MB
Free Disk Space: 7GB
You have 55.82GB
Operating System: Windows XP

Crysis
Publishers Minimum System
Show full details.. What is this?
Processor: Intel Pentium 4 2.8Ghz
Comparing your CPU Intel(R) Core(TM)2 CPU T7200 @ 2.00GHz (running @1997MHz) with the requirement Intel Pentium 4 (running @2800MHz)
• Your processor is significantly more powerful in raw performance
Display Card: 256MB, Supported chipsets: NVIDIA GeForce 6800 GT or greater; ATI Radeon 9800 Pro (Radeon X800 Pro f
Comparing your display card NVIDIA GeForce Go 7900 GS with the requirement NVIDIA GeForce 6800 GT
• Your display card is slightly less powerful in raw performance
Memory: 1024MB
You have 1024MB
Free Disk Space: 12GB
You have 55.82GB
Operating System: Windows XP/Vista
You have Microsoft Windows XP

Edit: Noticed this as well..
Attack on Pearl Harbor

[Jarnis]

Hmm, can you describe a bit more in detail what you find inconsistent?

(and maybe provide the ID link to your GOM scan? - the URL used for displaying any one of those GOM listings)

I looked those over, and it all seems to be pretty much in order. There is less data about laptop graphics cards than desktop ones, but the performance estimation in this case seems accurate enough for the first two examples.

Remember that if the GOM "overall graph" indicates your system performs higher than the minimum/recommended system, that means overall performance is good enough to meet those requirements (assuming no component fails for a missing feature). Individual component comparisons may actually show that a certain component, when compared to yours, is more powerful on average - and you get the "slightly less powerful in performance".

Yet if the overall graph says you are over the comparison system in total, that would most likely mean that your CPU helps so that the sum of your parts is more than is needed, even if the video card is estimated to be slightly slower than the comparison card (across all possible system configurations in the 3dMark result database).

Overall bar compares whole systems (CPU, Videocard) while individual component checks compare individual component performance across all available configurations.

The last case seems like a bug - the card passes the performance requirements, so the red X is odd. Could use a link to those results.

[Neeyik]

He's talking about his GeForce Go 7900 GS being "slightly less in performance" than a GeForce 6800 GT. On paper this isn't the case, and a quick search through the ORB suggests this is so too. However, such data is open to variation caused by end-user overclocking and power-saving underclocking, giving the GOM the impression that the Go GS is slightly less powerful than the GT. It should be the other way... just!

[laurence1211]

He's talking about his GeForce Go 7900 GS being "slightly less in performance" than a GeForce 6800 GT. On paper this isn't the case, and a quick search through the ORB suggests this is so too. However, such data is open to variation caused by end-user overclocking and power-saving underclocking, giving the GOM the impression that the Go GS is slightly less powerful than the GT. It should be the other way... just!

You guys work on the weekend too, shocking..lol. Its not just the GT version, it thinks a non GT card is faster as well.. Anyway links as requested:
Attack on Pearl Harbor
http://www.yougamers.com/gameometer/10185/
Dark Messiah of Might
http://www.yougamers.com/gameometer/10140/
Cysis
http://www.yougamers.com/gameometer/10213/

[Neeyik]

We need the urls that you get after running the Game-o-Meter, not the ones that just link to the test itself - basically, run the GoM and then when you get your results, copy and paste the url in your browser's address bar here.

[laurence1211]

We need the urls that you get after running the Game-o-Meter, not the ones that just link to the test itself - basically, run the GoM and then when you get your results, copy and paste the url in your browser's address bar here.

Ok had to retest to get what you wanted,
Cysis
http://www.yougamers.com/gameometer/...sInfoId=325805
Attack on Pearl Harbor
http://www.yougamers.com/gameometer/...sInfoId=325815
Dark Messiah of Might and Magic
http://www.yougamers.com/gameometer/...sInfoId=325825

And now weirdly it says my card is slower than a 6600 in the pearl harbor test...LOL..whats going on?
And the Dark Messiah test has gone funny too...
Edit: Ran pearl harbor test on my desktop and the result was fishy as well(see full details)..
http://www.yougamers.com/gameometer/...sInfoId=326016

[Pasi_FM]

The results certainly seem to be wrong. So far it looks like from my investigations that the Game-o-Meter has a bug somewhere in calculating graphics cards' performance estimates and the values that currently get used for comparisons on the live site are not consistent with the benchmark data or the actual estimation model it should be using.

We'll continue sorting this out tomorrow and hopefully we'll find the exact cause soon.

[Pasi_FM]

The problem affecting some graphics card comparisons has been located and a fix is on the way.

[laurence1211]

The problem affecting some graphics card comparisons has been located and a fix is on the way.

[thumbup].........