[Homer]

"To illustrate the problem of going all-in, let's say that I'm in a tournament of 100 people and 9 have been knocked out. I have AA in LP and before me one player goes all-in. Should I go all-in? The answer is probably not. Against a lower pair I'm a 4-1 favorite but finished #90 out of 100 20% of the time would kill my rating (unless I finished in the top 10% the rest of the time). So even in this very favorable situation I think that going all-in is incorrect."

That doesn't mean you shouldn't go all-in here. It means that pokerpages rating system is flawed. They shouldn't give you a percentage rank for each tournament you play in, they should assign point values to each of the top x (10 or 20) places. For any finish lower than x you get zero points. The percentage ranking system promotes weak-tight play. It is kind of like having a tournament with 100 participants and a $50.50 entry fee, in which first place gets $100 and 100th place gets $1, and every position in between gets a dollar more or less. In other words, the payout is linear. In real tournaments, very few positions get paid, and the ones that do get payouts don't get paid linearly, as the payouts are top-heavy. So in a real tournament you must take a chance with a hand in which you are a 4-1 favorite, because only the top few positions get paid. You must put yourself in a position to reach one of these places. In the pokerpages tournament you can play weak-tight and finish in the 80th percentile, which seems good, but really wouldn't cut it in a real tournament.

-- Homer