Skewed graphs in 3d graphics industry

I've always resented non-scale correct graphs (unless clearly visualized with a break-line and still holding the correct relative scale) and the subjective representation of these graphs. I remember an xbit review where they did this scaling and purposefully mentioned a 2% advantage in nvidia's hand as a slaughter of the ati card and the ati card performing 11% better being represented as a marginal improvement.

This all, both graphics and text are solely marketing attempts to skew reality for everyone that is in the spur of the moment when it comes to purchasing a new piece of hardware . .after all, you want a card that obliterates the other instead of offering a marginal improvement... these graphs do just that.. represent a marginal improvement as an obliteration..

 

Heh.

 

Not to be rude, but pie charts are terrible if you have more than one option to compare. In a bifurcated graphics world it may work, but if you want to compare the 3850 to 3870 to 8800gt etc... then you are in a real mess and bar charts or a simple table will convery the information more easily. And you are representing less info unless you also label them with the FPS. And then why not just list the FPS to begin with?

 

Nice

 

/boggle Can not comprehend... Anyway, I didn't list the FPS in the charts I provided because they were an overall compilation of all the results (I stated this in the post). Basically, the charts I did would be found at the very end of a review to provide an accurate overall picture of the results. If you simply add up the various frames per second you end up with games/tests being of different values than the others (weighted differently). Thus you have to standardize the results. While, I could have provided the standardized numbers, they would've been essentially meaningless in this context.

EDIT: Since the above may be unclear, a small example of what I did:
HD3870 vs HD2900XT World in Conflict results in 24 and 20 fps respectively.
So if you want the result of this game to count the same as every other game, then 24/(24+20)=.545 and 20/(24+20)=.455. Do this for every game/test and add up the results for an overall picture. Obviously, providing the resulting values is not going to have much significance to the reader.

Now, as for comparing more than one option, it really isn't so difficult - you just make alot of pie charts and arrange and label them coherently.

You do have a point that they do not work very well for anything other than a head to head competition. However, if you just want to primarily compare two cards, I think a pie chart "summary of results" at the end of a review similar to what I provided in this thread could indeed be a very valuable addition to the consumer.

 

Damn people, not providing all the data still makes it useless.

being 50% faster in a pie chart makes it look neat, but it might still be a single digit pissing match underneath. A lower average of frames might conceal the fact that one of the products has serious problems with maintaining a high minimum frame rate.

Seen in that light, graphs with an X(t) and Y(fps) are simply best because one can easily deduct all useful information through a median for those who are interested and make things glaringly obvious for average Joe because his blue line clearly beats the other guys red line.

I just feel like I don't have all the information at hand anymore.. remember the old UT2003 demo's where nV's performance was lower in bot matches than in flyby's .. I just don't see that kind of benchmarking anymore. I want to see CPU load as well.. hell.. maybe even memory pages from perfmon so that not only the graphics card final output gets measure, but it's position in a systems eco-system as well.

 

Damn people... READ. I said it would be an addition at the end of the review. All the individual data would still be available to reader if they were interested.

You are confusing problems with the chart type with issues concerning the underlying data. You can make pie charts based on minimum fps, median low, etc...

 

A single pie chart can not give me all of those at once, at least not providing a clear, blink of an eye solution

 

The word summary.

And yes, the method I outlined gives you a reasonably accurate picture for a particular set of data. If you can't tell which slice of the pie is larger at the blink of an eye, then I might suggest you visit an optometrist.

 

Ahhh, I don't want to be personal but... you don't need to waste your time like that.
Some people learn one way and want to stay like it forever. Full scale is good to show absolute values. Point of this one was to show how one card is performing compared to other. Think like in full scale you have 0 for beginning, in this one performance of 2900 XT is used as zero. Just try. Suddenly everything will be simple and nice.

 

The point we're trying to make is that absolute values are a pretty good basis for clearly presenting relative performance--the only basis worth using if that's your intent, really. Unless you think something like this is more useful both in terms of most transparently conveying the data or best doing so in an at-a-glance format. The correct answer is no, a simple table would have been more accurate and concise. (OTOH, I think an agglomeration of data like this would be better served with a bar chart with the relevant %ages overlaid on the end of each bar.) To each his own, according to his intentions and limitations....

As for the implications of 2900 being zero, that makes comparison moot, no? Much simpler to make it one, which brings us back to my first paragraph.

 

I used zero to explain difference between this type of chart and full scale. As you can see here, people got so used to zero as referential point they refuse other ways. Yes, to have it allign to 1 is great thing, often better then then full scale. Which brings me again to disagreement. You can tell at glance percentage difference and have this relative difference for more games in one graph. It gets better with more games tested, you get easily average difference between the two. This one number, although not precise, is the one most wanted. Not card 1 have x fps and card 2 have y.
Your link is error 403.

 

No, it is actually only about bars, fps numbers are irrelevant

 

That usually happens when one links directly to an inline image. Just click in the address bar and hit enter (or click the green 'go' arrow in Firefox).

 

To be more specific- the bars carry still the same point, showing 3870 doing about 10% better then 2900 xt.

 

you're not getting that information from looking at the bars...