i am not trying to offend. i meant:

Out of touch in this..........graphical presentations of data through modern methods focusing on efficiency and simplicity.

when the chart needs to represent small variances the scale should also be small. when you have data that has small variances its best not to have them clumped together. Its a lot more presentable to have the chart contain the differences that are most important with clarity.

Example:



there is really no need to show 0-100,000. If you did the data would be crunched and not centered. In this example we have the vertical axis starting at 100,000 and continues up in 50,000 increments which is great for the data it represents. The vertical axis starts about 50,000 lower than the lowest plotted value and its top line is at about 100,000 above the highest plotted value. This results in the data being symmetrical and easy on the eyes. It gives room for continued growth while keeping the data clear and presentable. People who work with graphs and carts strive for this. But i guess you could say that from this chart that 200,000 looks like double 150,000 and that there is an attempt to make the difference look bigger. I guess a person could say such things but its not anything t all like that. The graph focus is only on the difference between the independent statistics and nothing more. Its nothing deeper than this. Its just a more efficient way of keeping focus on what the graph is about.

another example:



this graph is not meant to trick anyone into thing that 2010 was double that of 2006. That is crazy.

You can see that its very proper to have your graphs showing whats important and not what isnt. Nvidia works with graphs every day. Tons and tons of graphs that focus on everything. A proper graph only shows what it wants to get across. If your not used to seeing graphs like these i guess???? well i still dont get it. I dont get how anyone can be saying whats been said. Clearly the scale is marked on the graphs. This is the way we do things. The way we are taught to.

look at the graphs here:

http://www.overclockersclub.com/reviews ... view/4.htm

they are so poor you each bar has to have the temperature wrote above the bar just to read it. although you might prefer this, its not very proper.

So lets look at it. the variance is from 785-984 but the vertical axis starts at zero. This clumps your data together and the differences cannot be clearly determined without the knacks above telling you what the temps are. it would look a lot better to have the vertical axis start at alot closer to the lowest variable. 750 to 1000 would be great for the data presented but if you wanted to have room for more data in time 500 - 1250.

Anyway, i am probably wasting my time but i was really trying to help bring understanding. The graph wasnt trying to mislead, even at the top in made clear to show "up to 30%" just case you missed the vertical axis starting point. it goes well beyond.

Ocre, I've understood your point from the beginning and see your point of view. This escapes you. What ISN'T happening is the reciprocal.
Now we must agree, to disagree. As is the case with many forum discussions.

never!!!!!!!!!!!!!!!!!!!!!!!