Hey, thought I would come to the math guys for this one.

I'm trying to do some database analysis. I see that opponents in the BN position raise approx 37% of the time. They obviously use multiple sizes.

I don't believe they balance each of those sizes correctly, and I'm looking to exploit that. But can I? I believe with certain assumptions, yes. But I'm not sure. Here is the information and the assumptions I've made:

I have 5000 hands where the BN had the opportunity to RFI (1500 players)
I have 2000 hands where they RFI'd (700 players)

I have 140 players who raised 400 times for 2bb
I have 280 players who raised 900 times for 2.5bb
I have 280 players who raised 700 times for 3bb

Assumptions
- No one mixes their preflop open sizing
- No one limps

Ok, that's the info. The way I tried to work it out:

700 players raised, 280 raised for 3bb sizing.
(Number of players who raised for 3bb) / (Number of players who raised) = 40%

So I come to the conclusion that 40% of players play raise-or-fold for 3bb. Is this a correct conclusion, given that I don't know what the other 800 players (who open folded) are doing, and given my assumptions?

If the bolded is a correct extrapolation, we can see that they raise disproportionately less than other players (40% of players only make up 35% of total raises) . So from here we can work out the exact RFI %. I haven't got to this bit yet. I have to know if I can correctly state the bolded.