3bet vs Hero

[hacacare]

Hi,

I play HUSNGs and I am trying to build this stat that shows the average 3bet percent of the player population I play against. So here is the expression:

(cnt_p_3bet_def_opp / cnt_steal_att_2_to_3x) * 100

where cnt_steal_att_2_to_3x is this:

sum(if[tourney_hand_player_statistics.flg_steal_att AND tourney_hand_player_statistics.amt_p_raise_made BETWEEN (2 * tourney_blinds.amt_bb) and (3 * tourney_blinds.amt_bb), 1, 0])

I have built this column to exclude those hands where I openpush, so there is no opportunity for my opponent to 3bet (I cant find a column for opportunities for my opponents to 3bet).

I got some reasonable stats, but I do not know if I got what I want. So, does this stat give the opponent population's 3bet percent? Is there a way to test my stats if they actually filter the hands (opportunities and events) I want?

[hacacare]

Also, would be this correct if I wanted to filter only those 3bets that are not allin 3bets?

(cnt_p_3bet_def_opp_can_4bet / cnt_steal_att_2_to_3x) * 100

where cnt_p_3bet_def_opp_can_4bet is:

sum(if[tourney_hand_player_statistics.flg_p_3bet_def_opp AND tourney_hand_player_statistics.flg_p_4bet_opp, 1, 0])

[kraada]

Your first stat is what percentage of the time that they faced a 3bet having made a 2Bet between 2BB and 3BB (which includes 2BB and 3BB).

The only problem with doing it this way is that cnt_p_3bet_opp (and your cnt_p_3bet_def_opp_can_4bet is that you aren't looking for only times the player also made a 2bet between 2BB and 3BB so you are comparing a few apples to oranges. You're not only looking at the times they faced a 3bet when they made that same size steal, so you could see numbers greater than 100%. A simple way to think about it is imagine a person who normally opens to 4bb - if we have 1000 hands on him and he opened to 3bb once the denominator is 1. If he was 3bet 50 times in the 1000 hands with his normal open you would see 50 / 1 or 50 here. That's not what you want to see I'd imagine

[hacacare]

I think I did not explain it well . So I want a stat that is not a stat for a specific opponent, but for the player population I play against, since many times I do not have enough hands to get accurate stats for one player. Lets say I played 9 hands against an opponent so far and I am facing his first 3bet. This would give 20% percent 3bet, but obviously this is not an accurate number because of the small sample. In this case I would use the population tendencies for my decisions as the best approximation I have. Probably the regular way would be to create an alias which includes all the players, however I would try to find an easier way for it. Since I always play HU, all stats on opponents I have are from the hands they played against me. So I thought instead of creating a stat for them, I would create a stat for me. Whenever they 3bet, I am facing a 3bet, so my 3bet-facing stat should be actually their average 3bet. I never raise preflop more than 3x (only if I raise allin), and obviously never below 2x, this is why I thought this should be OK (they cant 3bet if I openshove preflop, thats why I need to filter somehow). Is that correct this way - do I get the actual 3bet % of the population? Is there any other way to get a population stat easier?

The ideal would be if I could weight the stat for specific player with population tendencies (Bayes theorem), but for that I would need the standard deviation (SD) for a given stat. Can I somehow get a specific stat out of the database for all the players? In a format where I would get for example all the 3bet stats for everyone (exept for me) together with the SD (or without it, separate stats for everyone in some exportable format - so I could calculate SD). Without having to check it for everyone one by one, obviously.

[hacacare]

OK, I see, this last one was a noob question - there is an all player report, I think this will solve all my problems

Amazing software, congrats