Dear Rick,
thanks for your fast reply.

>Your stimuli seem fairly dense, so there isn't so much room for
>randomizing the ISI. Maybe that is why the range of results is
>a bit tight.

Right, we have to do a trade-off between runs length (so consequently rather short ISIs... to get enough power with a high number of trials) and experiment duration....

About the seed:
No this is how we run the script... MMhh but the seed was intended to add +1 to the original seed to get different seeds per iteration...That's why we chose

....
@ seed = $seed + 1
....
-seed @seed

Yeah, I know, the 'python' prefix is strange, but the code seems not to be executabe without it....

So, the bottom line is 50000 iterations 'should' be enough and the rather 'small' differences in average effectivity are probably due to the dense timing?

Best,
Dorit