[f97tosc]

Yes for a given run of 4 KKs this is mathematically correct (taking your word on the 1/20 stat).

However, there are some deeper problems with the general procedure of noticing something unusual and then afterwards calculating the odds for that event only.

The problem is that there may be quite a few "surprising events" that you would have found highly unusual and which might have made you post at 2 plus 2. For example, what if the 4KKs vs. the AAs had been last month and not in the past 5 days? What if you had had 4QQs and faced KKs each time? Or what if you had flopped boats vs. quads four times? Etc?

And then after one of the remarkable events happens, you calculate the probability only for that particular event, and find that it is very low.

A better measure of "unlikeliness" is to *first* write down *all* the events that you would consider unlikely, and the measurement period, and then afterwards calculate the odds.