Just a little more talk about weighting ranges:

Also warning sizing is terribad by me in this hand
Villain is fish 35/0 over 20
PokerStars No-Limit Hold'em, $0.10 BB (5 handed) - PokerStars Converter Tool from FlopTurnRiver.com

SB ($15.14)
BB ($9.62)
UTG ($54.34)
Hero (MP) ($24.18)
Button ($10.42)

Preflop: Hero is MP with K, A
1 fold, Hero bets $0.40, 1 fold, SB calls $0.35, BB calls $0.30

Flop: ($1.30) 4, K, 5 (3 players)
SB checks, BB bets $0.30, Hero raises to $1.40, 1 fold, BB calls $1.10

Turn: ($4.10) 5 (2 players)
BB checks, Hero bets $2.50, BB calls $2.50

River: ($9.10) J (2 players)
BB bets $5.30 (All-In), Hero?


K so in this hand we get to the river and in reevaluation we may need to construct a range here for our villain. Here he can probably have a flush, a FH, a king, or some other random combo. Thing is when we weight each portion of his range we need to realise he's probably playing this way very often when he has a flush/FH (well jj atleast) , but he's playing kings this way much less frequently. This makes the math alot trickier, but for evaluating this kind of spot accurately, we need to take out combos of like kx due to the unlikelyhood that he plays it this way.

Picking how many combos to take out is where things start getting really really fuzzy. One method is to if for example we assumed villain played a king this way 10% of the time we would take out 90% of the kings in his range. Problem with this is it skews our range because we would next need to multiply the % he plays other hands in his range like this times their number of combos.

Basically best we can do is make educated approximations and get as close to the true equity of our decisions as we can.