The only way that those numbers will have value is as a measure of relative rather than absolute improvement. If we try hard to compare near-identical conditions in the same rat with the variable being substance "X" and we find a significant change, that flags it for a closer look.
It is not a blinded trial and is not intended to be. It plays to our strengths - namely that we can be sloppy but still act as a fast way of identifying places for "real" scientists to look. If "X" cuts my time to first on by 25%, that might inspire another rat to try the same and they might report 50%, and another 10% etc. If ten rats report improvement but they are all over the map, that's great. We did something positive instead of twiddllng our thumbs.
I'm going to try one just as a test of this approach. I have here a bottle of Melissa officiallis (lemon balm). Brand is Nature's Way and it contains 490 mg per capsule. It has a long tradition as a remedy for stress and there are even a few human trials on PubMed - all positive. Reccomended dosage is three caps per day so I'll start there. I will take the first with my first meds, the second at noon, and the third at 3:00 PM
I'll use this morning as my baseline. I slept reasonably well, got up about 5AM, came on in one hour, but began to falter after about an hour and a half (which got me to thnking about this). During that time I had a boiled egg and a rice cake with peanut butter and a cup of coffee. A typical morning for me. I will try to reproduce those conditions anytime I gather new measurements.
In theory, if I see improvement in those numbers, then it is worth trying to refine the experiment in some way.
Hey,it keeps me off the streets!