Here’s a breakdown of a massive $933,000 poker hand using insights from GTO Wizard AI.
In a jaw-dropping hand from season 14 of High Stakes Poker, a pot swelled to $933,000 in a classic “cooler” situation: a full house versus a better full house. The hand featured Justin Gavry and Brandon Adams playing with 1K/2K blinds and stacks around 230 big blinds deep. Using GTO Wizard AI’s 200 big blind MT ranges, we can analyze the complex decisions that led to this colossal clash.
Pre-Flop: A Standard Open and a Questionable Call
The action began with
Justin Gavry in the cutoff, who opened to 2.5 big blinds holding Pocket Deuces. The solver analysis confirmed this was a “pretty trivial open” and a “very reasonable” play.
The intrigue started when Brandon Adams in the big blind just called with Pocket Queens. According to the GTO Wizard ranges for this stack depth, Pocket Queens should be apure three-bet. However, the difference in expected value (EV) between three-betting and calling is relatively small. While calling is a more passive line, it was deemed “pretty reasonable” by the analysis, especially if Gavry was perceived as a tight player.
Flop: Building the Pot with a Set Over Set
The flop came Q-3-2, giving both players a set.
Gavry, with bottom set, made a standard continuation bet of about one-third of the pot. This was a clear range bet with 99% frequency, as the board texture is highly favorable for the opener’s range.
Adams, holding the top set, responded with a check-raise to about 4x Gavry’s bet. This was considered a “very good size,” and the analysis revealed a crucial point about deep-stacked play: Adams should be check-raising
at full frequency with his sets and two pairs. When this deep, extracting maximum value becomes more important than protecting a checking and calling range.
Gavry correctly called with his Pocket Deuces, as the solver indicates he shouldn’t be three-betting much in this spot.
Turn and River: The Cooler Completes
The turn and river brought running fours, completing the board as
Q-3-2-4-4. This unlucky runout gave both players a full house:
- Brandon Adams: Queens full of Fours (the top full house).
- Justin Gavry: Deuces full of Fours (a lower full house).
On the turn, Adams made a full-pot bet, a slightly larger but acceptable sizing compared to the solver’s preferred 3/4 pot bet. Gavry called with his set, which was the right play against Adams’ polarized range.
On the river, Adams led out with another large bet, a little below the pot size, which the solver considered “a very good choice”. This is where Gavry faced the ultimate dilemma. He decided to jam all-in. According to the GTO analysis, this was the correct play;
all of Justin’s full houses are supposed to jam the river every single time to get value from weaker hands like straights and trips.
The $933k Question: Could Gavry Have Found a Fold?
The short answer is probably not. The analysis concludes that it is “really really difficult” for Gavry to avoid getting stacked in this scenario.
Folding a hand as strong as a full house is generally considered unreasonable. Gavry’s jam is the standard, theoretically sound play. The strategy relies on the assumption that Adams will call with enough worse hands (like straights) to make the jam profitable.
If Gavry were to deviate and fold his lower full house, it would create a massive exploit in his own strategy. If Adams knew Gavry would overfold in this spot, Gavry would then have a huge incentive to start bluff-jamming his entire range, knowing Adams was folding too often.
Ultimately, both players took lines that were considered “pretty solid throughout the hand”. Gavry was simply the victim of a brutal cooler, a situation where even near-perfect play leads to losing a massive pot.