Misreporting Benchmark Results Everywhere!

[S0me1X]

/rant

Am I the only person who hates poorly done excel graphs to report benchmarks?

I am talking about those 3DMark/Sandra bar-graphs where a score of 6000 looks like it is twice as much as 5950 because the axis was set to 5900 to 6000. WTF! It could easily give someone the wrong idea that one card/cpu/whatever is twice as fast as another. It obviously is not the case.

 

[Howard]

It isn't really misreporting. I find it much easier and much more accurate to have an axis that includes only the relevant data when completing a graph for school work.

 

[Howard]

It isn't really misreporting. I find it much easier and much more accurate to have an axis that includes only the relevant data when completing a graph for school work.

 

[S0me1X]

it may be easier for you to graph on paper, but this is using excel or other spreadsheet software.

also, shouldn't a bar-graph comparing 6000 to 5950 show two bars of nearly same length? Numerically, they are very close; they should be graphically close, too.

 

[Wingznut]

For some time now, I've thought that bar graphs are not only overused, but poorly used in reviews. I think that a very large segment of the readers only look at the graphs.