I want to test to what extent people can anticipate downswings, and I have a simple idea how to make this.

First, estimate your advantage over your opponents. Suppose you are offered a near-coinflip with the better part of it, like vs . At what point would you refuse it? Like 55 vs AKs, or 33 vs AKs, I don't know, make a reasonable guess. Then you estimate the bankroll that you need for this. (it's better if you would precise which probability of going broke you accept in your estimation.)

I did some calculations and I'll post the results later, if there is some interest.

Once again, I'm confused.

If I KNOW they have AKs and I have a black PP <KK, what's the difference between 22 and QQ?

If I have a good br and my mindset is in the right place, then I'll take them all day long.

I think if I could have a 50 buyin BR the chances of going broke would be very (<1%) small ?

I wrote a similar C++ simulation to this a few weeks back, I can always update it with your equity numbers and rerun it and make graph over the probability of going broke vs X number of buyins.

I thought you had 50.3 % equity with [2s] vs [Ks]?
And that you have 53.9 % equity with [Qs].

Now that is a huge difference.

With 22 and 100 BB stacks you win 50.3% of a 200 BB stack each time (no rake I assume) +EV of 0.6BB. I.e. in 100 hands i win 60 BB/100

With QQ, I have + EV of 7.8 BB, i.e., in 100 hands I win 780 BB

Although the difference is only a few percentage points, the winrate difference is huge. Basically the EV over 50%, 3.9/0.3 = 13 times.

I would assume that I would need a huge BR to be able to have less than 10% risk of ruin with 50.3% equity. I am too lazy to do the math, but I do not think 200 buy ins of 100 BB is enough.

With QQ and 53.9% I think you need a lot less. say 50-75?

Zmej, do the math for me. plz.

I dont really get the discussion, how often are you gonna deduce that he has AKs without putting money in the pot...and once they'res money in the pot folding a coinflip would be -EV...correct me if im wrong but this seems like a moot discussion.