The pool exploit is research for a future LeakPrice product; it is described in the “Soon” block on the home page. The video shows one spot: a 3bet pot BB vs BTN, flop Q♠J♥9♥. The pool is the regs on the BTN from the LeakPrice database: WPN, iPoker and Chico, NL200–NL5000.
What is shown
- The pool is locked whole. Every node from the hands of the database: 3,336 hands on similar flops and 20,934 on other boards. Where a node has few hands, the frequency is taken from the same move on the street, with no data — from the solver. Of the pool’s 63 nodes, 35 are locked from their own hands, 28 from the move on the street.
- Where the pool deviates from the solver. Flop, BB checks: the BTN checks back 60% vs the solver’s 69% (941 hands). Turn after the BB’s cbet and a call, BB checks: the BTN bets 33% pot 15% of the time vs the solver’s 34%.
- What it is worth. The BB’s EV per hand of the spot: the solver as it is against this pool — 13.31 bb, the exploit — 15.88 bb. That is 2.57 bb more, 11% of the flop pot. It is the win in one spot, not a winrate.
- A check on the other half of the database. The exploit is built from one half of the hands and plays against the pool locked from the other half: on hands it has not seen, +1.60…+1.90 bb over GTO. On 8 flops, 67–91% of the win remains. This checks that the exploit is not fitted to the randomness of the hands; it is not proof of the pool model.
Why locking one flop is not enough is in the article “Solver vs the pool”. Right now the LeakPrice subscription includes a full database review and a pool review once a month, plus the account, where you order it: leakprice.com/cabinet.