It bloody well works!

Written on Wednesday October3, 2018

Small sample sizes and all those caveats.

I’m tossing out the number of figures being born as a threshold. That will pop out of any of them at any time, and it’s just luck. One of them starts reproducing very quickly, and happens to find a window of relatively few mutation’s, and then skate on through to the finish line.

I retested the first five populations, snagged with a threshold of 100 and 100,000, if memory servs me. Looking at the most mutations given to a population, side by side results, first five and latest five.

First
excluding mu2
largest most 464 mu3
smallest most 84 mu1
Mean most 274
average most 308.4 with mu2 296.833__

latest
excluding mu5.3.1.2
largest most 663 mu5.3.1.3
smallest most 388 mu5.3.1.0
mean most 525.5
average most 524.25

You might have noticed, that mu3 did the best. Why didn’t I use mu3; why mu5 instead? Because I was using 100 extinctions for testing averages, with 1000 extinctions, mu5 was the second best, with a 61 average and a most of 401, compared to 68 and 663 for mu3.

Whatever… it still worked.