Stack Management in Limit Deuce-to-Seven Triple Draw Tournaments

Deuce-to-seven triple draw is a notoriously volatile game. Bets are usually made on every street and pots frequently see the river, making for high-variance tournament affairs. In limit hold’em or stud, players are more likely to fold on the flop or fifth street, only losing .29L and .25L (L — the amount of chips you invest in a pot if there were a preflop raise and bet on every street). However, in triple draw, players expect to sink almost a full L when they enter a pot.

This summer, I was excited to make day 2 of the $2,500 World Series of Poker deuce-to-seven triple draw tournament because of how scarcely the game is found in the tournament landscape and how abruptly a run in them can end. I brainstormed a new tactic to navigate these unstable tournaments. I dubbed it — anti-gamble technology.

Conventional Strategy

I was apprehensive to try this approach as it was contrary to basic principles I learned about the game. When drawing fewer cards than their opponent, good triple-draw players usually bet the next non-river street unimproved. After the first draw, if Tifa pitches one, and her opponent draws two, the correct play is for Tifa to bet even if she bricks, as she rates to be ahead even when missing. Poor players can often disguise themselves as competent ones for some time by following these predetermined betting patterns.

On the other side of this equation sits the person drawing more cards. Triple draw has large fluctuations partially because it is often correct for the person behind to continue when missing. This is especially true between the first and second draws as the wager is still a small bet, with two further draws remaining, leaving hand equities relatively close.

If I raise in the hijack, call a reraise and draw two cards, I will typically only have the better draw two holding (three cards to a 7 containing a deuce, as well as 8-3-2, 8-4-2, and 8-5-2). A majority of times the three-bettor will draw one. With these strong hands, I can call their bet intending to draw two again if needed. Even though my opponent could have improved on the draw, I can infer that she is betting unimproved and am facing 8.5-to-1 pot odds with the possibility of making a strong hand. With these holdings, I still have a reasonable chance at making one of the five best hands.

If I had a worse holding, such as 8-6-3 or 8-7-2, even the best draws leave me with mediocre 8-6s and 8-7s. This drop in hand strength is often the difference between it being correct to continue or fold, and a reason it is undesirable to play these hands from early position.

Let us change our perspective to the three-bettor’s now. Ben opens in the hijack, we have 8-6-4-2 on the button and reraise him. He calls and draws two, indicating that he likely has the tight range we modeled above. We miss our draw. If we bet the next street, he never folds. Forty-seven percent of the time Ben will improve to draw one, another nine percent leaves him pat, totaling 56 percent of the time he improves to a hand at least as good as ours. However, he will also call even if he misses, with reasonable equity.

Deviating From Standard

We have another choice. We can check. Forty-four percent of the time we will have given Ben a free card. However, there are situations in tournaments where it is acceptable to do this, as otherwise we are adding chips to the pot with no fold equity. In situations where villains never fold and a bet shows a small edge, a check in the name of stack preservation is defensible. This is a trivially fundamental tournament principle, but can be missed in triple draw because so much of the game revolves around robotically betting anytime you draw fewer cards than your opponent.

In a cash game, it is critical to always punish them for the times they miss. However, on the tournament bubble (less than 15 percent of the field left when 10 percent cash) or at the final table, it may make sense to forsake this edge, to avoid the tacked-on pot inflation.

Checking in these spots diminishes our contribution by one bet immediately, but also shrinks the pot size by two bets, disincentivizing all players from bluffing, snowing, bluff-catching, and value-betting later. We are not only decreasing our variance because we are putting less money in the pot, we are also decreasing our variance because there is less reason for to fight for the pot later. Anti-gamble technology!

When Is As Important As How

Knowing when to utilize these ideas is just as critical as actually applying them. It is imperative to measure variance avoidance fluidly — this issue is not black and white. Many players enter a tournament under delusions that they can gamble early and tighten up late. I’m not trying to gamble or avoid gambling at any point. I’m always in this business of mitigating variance; the questions are “to what degree?” and “what edges will I tradeoff in the name of it?” With thirty percent of the field left, variance is not something I embrace, but I am still going to charge my opponents when I am drawing one to their two.

Each decision in a tournament is optimally viewed as making the play that wins the most chips and minimizes variance. Oftentimes, the two forces are at odds, and you have to do your best to resolve the tension between them and make the best overall play. I tend to assume that people err too cautiously once they have decided to implement these tactics, given how weak-tight play frequently rears it’s head as early as the mid-stages of a tournaments.

Thinking Ahead

Not betting in these situations also creates some interesting side effects. Anytime you deviate from standard lines, future decisions, image, and metagame will also change. Next month, I will examine a hand from day two of the $2,500 WSOP event where I chose not to bet and was left to deal with some unconventional spots. ♠

Ben Yu attended Stanford University but knew even before finishing that he wanted to embark on a journey to become a one of the finest professional mixed-game players. He made his debut onto the tournament scene in 2010 with a second-place finish in the World Series of Poker $1,500 limit hold’em shootout and followed it up in 2011 by leading the WSOP with seven cashes across six different games.  In 2012, he moved to Rosarito, Mexico in order to continue playing online and was enthralled to perform well at the World Championship of Online Poker, including a final table appearance at the $10,300 poker 8-Game High Roller, and a cash in the main event.