(2019-02-09, 12:36 PM)Chris Wrote: But under a more general model I would still expect the gradient of the curve to show a minimum midway through the event.
I'm not understanding why this is, but then, I don't know how or why you chose your model either - not that I'd necessarily understand if you explained, but it might be worth "putting out there" for those who are capable: you have already drawn Roger Nelson to this thread.
(2019-02-09, 12:36 PM)Chris Wrote: But what wouldn't change, unless I'm mistaken, is the theoretical power-law dependence of the effect size Z upon the event duration N.
Again, without a lot of work, effort, and guidance, I'm unlikely to understand your explanation of that either, but it might be worth explaining for others.