[tessarji]

I don't know if this is the type of answer you are looking for, but:

Many estimates indicate an standard deviation of about 17BB/100 for 6-8 player hold em poker. Let's assume your day was about 2000 hands online. The standard deviation goes as the sums of the squares so sqrt((17^2)x20) is about 76BB/100.

If you are exactly a break-even player, a result of -150BB would be at minus two sigma, so it would happen about 2.25% of the time you play 2000 hands.

If you are a winning player at 1.5 BB/100 and you lose 200 bets, that is an 'effective' swing of -230BB, which is close to -3 sigma. That should happen 0.12% of all 2000 hand sessions.

This is assuming a normal distribution which is fairly good estimate for medium term poker results.