I think that the problem with the random-number-generator "simulations" is that they do not account for such things as reads, positional advantage, the ability to fold a big hand when needed and other poker skills. They also don't account for tilt.
There was a video recently where bogdan ran several simulations using cash games though.
The site he used is here:
Play with the site, and use a winrate of 5bb/100, and
start w/ a standard deviation of 90.
The worst run results in winning 30K bb
The best run results in winning 60K bb
The more hands you play, the more the spread between the "luckiest" player, and the "unluckiest" player.
To make things easy, let's call the luckiest guy "Phil Ivey", an the other one, "Me".
both of us have THE SAME WINRATE. but the spread is huge.
Now drop the SD to 5.
Note how now the runs are all grouped together at around 50K bb
The ultimate conclusion is that "Variables that you have no control over have an effect on the outcome".
Since they do, it is necessary to optimize those variables that you DO have control over, and let the chips fall where they may.
I submit that as you perfect your skills, your winrate will increase (yay), but also your SD be lower, and tend not to vary so widely.
Oh, and the question, "What is the biggest possible downswing?" can only be answered by another question:
"How much money do you have?"
Great thought-provoking post,
Thanks for that...
--VS