January 16, 2011

No Bad Beats? No Thanks.

"If it weren't for the river, there wouldn't be any fish."

—Ironman Santa Claus

Today is the planned launch date for a new Everleaf poker network skin, No Bad Beats Poker (NBB). NBB's gimmick is to offer No Limit Hold 'Em (and a few other games) free from those annoying "bad beats, suck outs, and bad luck." NBB promises: "Out play your opponents and build your bankroll! You’ll get your money back if you take a Bad Beat."

So how does NBB pull off this poker miracle? The basic rules are:

  • You must be playing at a No Bad Beats Poker table or tournament
  • At least one player in the pot must make an All-In wager
  • At the time of the All-In wager, your win probability (for that pot) must be greater than the Bad Beat Cutoff % posted in the table description (i.e., 70%).
  • You cannot fold at any point in the hand
  • You must lose the contested pot (i.e., cannot hold the high hand at showdown)

As an aside, the last two rules are really pretty silly to list at all. If you either fold or win the hand, you haven't been the victim of a "bad beat" in any commonly accepted definition of the term. (NOTE: The NBB FAQ page has more detailed explanations of the NBB game procedures).

Now it's important to note the effect of the bad beat protection. If two or more players get it all-in before the river, and the hand with the higher probability of winning loses (and the probability of winning is high enough to exceed the table bad beat threshold), the favored hand gets all of their wagers (not just the all-in amount) refunded to them from the pot. The winning hand (the one that sucked out) wins the remainder of the pot.

The entire NBB concept is horribly flawed. In terms of execution, it appears that NBB regards a "bad beat" as any time a favored hand loses to an underdog, even when the underdog has a substantial amount of equity in a pot. The scenarios used by NBB to explain the game play generally use a bad beat threshold of 60%, 65%, or 70%. In other words, if two players get it all-in on the flop with an overpair against a flush or straight draw, the overpair will typically find itself in a "protected" status, even if the drawing hand had proper odds to make the call. Or, in a preflop all-in between two pocket pairs (say the ever-deadly Aces vs. Kings), the higher pair is roughly an 80% favorite. The higher pair essentially gets a freeroll, winning the entire pot if the lower pair does not improve, while getting a full refund if the lower pair sucks out.

The NBB concept also impacts proper play. If the bad beat threshold is 65% or less, hands like overpairs and top pair might be well-advised to push against obvious drawing hands, to gain the NBB protection. By contrast, pure drawing hands generally shouldn't be played aggressively out of fear of generating a negative freeroll situation. Pushing preflop with Aces should be an almost standard play, while hands like Kings, Queens, and Ace-King might be auto-folded against tight players who push all-in preflop (which immediately suggests an interesting new way for tight players to steal preflop).

One interesting scenario which occurred to me is that a monster draw on the flop might well be able to gain NBB protection, and have a freeroll at hitting their draw. Imagine JhTh against 6s6d on a board of 9h8h2d. According to Cardplayer.com's odds calculator, the monster draw is more than a 70% favorite. So if the monster draw pushes (or calls) an all-in the draw wins the entire pot if it hits, and loses nothing if the draw misses. That's right, miss your draw, keep your money!

Much of the egregious idiocy of the NBB effect on game play could be remedied with a simple requirement that the bad beat threshold be a minimum of 85% or 90%, high enough that "standard" plays like pair vs. pair preflop, or pair vs. draw postflop, aren't eligible for NBB protection. Setting the threshold at 90% would essentially require a two-outer (e.g., set vs. overpair) or runner-runner on the flop, or a four-outer on the turn (e.g., two pair vs. straight or flush) before NBB protection kicked in.

Even if the NBB protection threshold were set at 90% or above, the entire NBB concept is flawed. Looking back at the Aces vs. Kings all-in preflop matchup, the Aces have roughly 80% equity in the pot. When the Aces win, however, they get 100% of the pot. The extra 20% of the pot above the Aces' equity point is properly regarded as money that is "on loan" to the Aces until the hand arrives where the Kings suck out and win the pot. The problem with the NBB concept is that it essentially steals equity from underdog hands and awards it to favored hands.

Finally, the NBB concept is objectionable merely as a matter of game philosophy. In football, one doesn't let a team drive to the one-yard line, then simply award them a touchdown because statistically teams score touchdowns from the one-yard line with a 90% or better rate. Likewise, the point of poker is not to wait until a highly favored hand is dealt to you by chance, and then simply declare victory. Poker is supposed to be a game of skill. Maneuvering opponents to call bets with bad odds is a key skill for winning at the game, if not the key skill. But opponents will quickly learn not to make such poor plays if they know that the favored hand is essentially insured against losing.

Bad poker players continue to play badly because their bad bets are occasionally rewarded. Similarly, casino pit game players keep playing because they occasionally take the house for a big score. But if the swanky casinos of Vegas were built on small house gaming edges (often 5% or less), why do poker players need bad beat protection when they can repeatedly get their money in the pot with a 30% or better edge? Bad beats and suckouts can be frustrating, but they are the bait needed to land the biggest fish. NBB is the epitome of killing the goose that lays golden eggs.

"Las Vegas is a city built with money won from people who aren't very good in math."

—Penn Jillette

1 comment:

  1. Interesting concept, but it sounds to gimmicky for me to try it.