As rhoudine say, the theory behind bankroll management is to reduce the risk of ruin for a winning player to an acceptable level. So lets say we want to play cash games in a local casino 2-3 nights per week, and the cheapest game has $1/$3 blinds. Then if we make an educated guess about our winrate and variance, we can calculate, how much money (bankroll), we need to set aside to be reasonably sure, we dont need to stop playing, because we run out of money. Maybe if we decide, that we can only accept a 1% risk of ruin, we need $12.000 (4.000BB), but if we can accept a 10% risk of ruin, we only need $6.000 (2.000BB).
However in
online poker most people do it the other way around. They let the size of their bankroll determine, which games they play, which also mean, they move up and down. And then bankroll management is not about controlling the risk of ruin, but rather about controlling the size of swings. So if you can accept larger swings, then maybe its fine to only have something like 50BIs for the most expensive games, you play, especially if you also play cheaper games to keep your average buyin down. So in this example you can play $10 tournaments now, but if you drop below $500, you have to stop.
You just need to be aware, that more aggressive bankroll management means larger swings. So if you would feel really devastated, if your $600 becomes $150 or even $300, then aggressive bankroll management is a bad idea. And the truth is, most people only like variance, when it goes their way. The majority tends to tilt, when they lose a significant portion of their bankroll, and then they either donk the rest of it away or stop playing.
Finally as rhoudine also said, field size matter for tournaments. The advice about 100BIs is tailored to small field MTTs, lets say 50-300 players. For single table SnGs you can get away with less, and if you mainly play large field MTTs (>300 player), you might want to consider something even more conservative than 100 BIs. Personally I use 50 BIs for SnGs and 100 BIs for MTTs with an occational "shot" above those limits. This has worked well for me for several years, but for other people other numbers might be better
