Originally Posted by stevehnm
Lol - the fewer samples you have then yes, the more individual differences will vary the results. However, this data not only has some intrinsic value but it matches other data quite well. The 5MT appears to give about the EPA ratings. The CVT results pretty much mirror the first post in this thread, which was by professional driver(s?) trying to achieve the same roadway conditions.
OK, I get it. Different drivers, using different driving methods, over different terrains, and at different speeds will get different MPG averages. This means that your graph is totally meaningless regardless of any sampling as they do not address any or all anomalies in the data collected and used as the basis for the statistics. I think that you are trying really hard to "fit" the "statistics" to you own conclusions - inane and maybe dishonest IMHO - stats, damned stats, and lies "type of thing" rings true to me.