Hold on now, all I said was that I do not have to recall each and every instance of Tebow making a big run or throwing an incomplete pass because the stats show it happened a lot. That doesn't mean I never saw it, or didn't see it a lot, only that I don't have to dredge up ALL of them from my all too human memory to support my general memory of it happening a lot. It means that even if a few big runs/bad passes stand out more than others in my memory, or I imprinted too much on the thirty yard sack against KC or the 51 yard over the shoulder spiral to Thomas against Pitt, checking the stats double checks my memory to give me a more objective and thorough, and thus more accurate, portrait than a few snapshots from memory.
When judging by stats more is better; an accurate picture requires all the stats. Looking at some but not others can aid focus on some things, but often skew perceptions, too. That's why cherrypicking is ill advised. If/when we look exclusively at a given set of stats because they are unique to a particular situation we must bear in mind that limitation. We can't extrapolate from them to more general situations unless we factor in the other stats equally relevant to those situations.
I agree stats only support or undermine what we see (or think we do.) I'm pretty sure I already said as much.
However, if they were wholly subordinate rather than complementary to the eye test they could ONLY support, NEVER undermine, it; any stat conflicting with the eye test would automatically be wrong. The biggest stat is wins/losses, in games, seasons and championships. It's shown many great athletes badly lacked what mattered despite passing the eye test throughout HS and college careers, into the Combine, onto practice fields and even through training camp. Looks can deceive, especially accompanied by a lot of back story.
I don't watch college ball either; I consider it "amateur" in all the worst senses of the word (ironically, not in the most literal sense, because the best "college" players aren't getting paid with a degree (at least) to play.) Double that for the SEC, then square it for the dozen or so chosen schools the SECAA deems worthy of nomination for election as each years "national champion." I'm pretty much the last guy who'd climb on a FL option QBs bandwagon; I was all in with Cutler, horrified when McDumbass drafted Tebow and fully expected him to crash and burn the moment he set foot on a pro field.
He didn't, and surprised my eyes enough I want to see how this ends. Odds are still against him because he spent four years playing for a typical SECAA coach in a typical SECAA system, but he's come a lot farther than I ever expected, and certainly much faster than I ever dared hope. It's not as pretty as Testaverde or Quinn, but I'm confident it'll end differently; whether better or worse I can only guess at the moment.
Stats are just measurements, but the measurements THEMSELVES are unbiased. The limitations are that they're useless without context and measurements for one thing can't be applied to another.
Is a fifteen foot snake long? If it's a cobra, yes; if an anaconda, no. Based on its fifteen foot length, how well can it drive a car?
Interesting analogy though; I wonder how many who think a higher Int% and lower TD% is acceptable if completion percentage rises ALSO think a higher GDP makes increased child mortality and lower life expectancy OK?