Reset Seed before Monte Carlo simulation?
complete
Bob Croucher
Hi. I noticed this last weekend. We don't yet have Strength of Schedule, but one decent metric is the difference in rank between before schedule and after schedule. Each time I slide the slider back and forth, I get new results. With 1000 samples it seems like the results would be more consistent. In the picture I show 2910 jumping all the way from 8th to 21st place. Maybe the seed should be reset.
A
Abhijit Gupta
complete
A
Abhijit Gupta
Yeah there was a bug that outputted garbage simulations before the schedule was released. Should be fixed now.
Bob Croucher
I can't think of any schedule in which Jack in the Bot could rank 21/27. So there may be some sort of error here.
Bob Croucher
It just hit me that the schedule hasn't been released yet, so perhaps we are getting results from random schedules. I do recall seeing these jumps between simulation results even after the schedule was released last week though. So it may still be worth investigating.