How Much Luck and Skill/Edge Exists in a Game of Open-Face Chinese Poker? Part II
Open-face Chinese (OFC) has become one of the most popular games in poker. Top pros like Jason Mercier, Barry Greenstein and Shaun Deeb have become unofficial ambassadors for the game, while dozens of other pros can be found playing the game either late into the night or on their tablets. One of those players is Nikolai Yakovenko, who helped create the ABC Open-Face Chinese Poker iPhone app.
Yakovenko, also known as "Googles," is originally from Moscow, Russia and is now a poker player and software developer that resides in Brooklyn, New York. After several years at Google New York working on ranking algorithms, he's been developing independent software projects ever since while making both World Series of Poker and World Poker Tour final tables. Here Yakovenko, who has $313,598 in WSOP earnings, examines how much luck (variance) and skill (edge) there is in a game of OFC.
*This is Part II in a two-part series. Please be sure to check out Part I by clicking here.
It’s fun to go for Fantasyland, and being skillful at getting to Fantasyland will be enough to beat a non-expert player in the long run, if you can take the swings. But are my successful big hands paying off the fouling needed to get there?
This summer at the World Series of Poker, the most successful OFC player I played with was Atlantic City mixed-limit pro Ryan Miller. He’d play three-handed and four-handed against us aggressive players, setting his hands a lot more conservatively that we did. It seemed like he wouldn’t foul a hand for hours at a time. Ryan rarely went to Fantasyland (I estimate at least 3-4 times less often than aggressive players like myself, Ashton Griffin and the Russians), but he never fouled, rarely got scooped, and made a big hand every once in a while. He ended up with all the money.
I’ve been wondering what it must be like to play like Ryan. Could I have more success, and lower my volatility, by taking advantage of my computer opponent’s weakness for dubious draws? I wrote the computer AI, so I know its weaknesses. Most of its gambles for Fantasyland are sound, but not all. The computer struggles with properly accounting for live card combinations, and it doesn’t do a great job accounting for its opponent’s chance of fouling, when deciding whether to give up on its own hand, when not to gamble, because its opponent is likely fouling already. But I can account for all of these thing in my play.
Defending Against the Maniac
Thus, I followed up my 223-hand aggressive-play session with a session where I tried to play like Ryan. Since I knew my computer opponent will foul 1/3 of its hands, I took care to avoid foul risk. Instead of pairing my in the middle with a pair and live kicker on the bottom, I’d split aces and put one up top. I made a lot of one-pair (bottom), ace-high (middle), K-high (top) hands — which were usually good enough to avoid the scoop — and could win an easy +6 scoop for me, when the computer fouled, as it often did.
The results were remarkable. Over 191 hands , I won +167 points, for an average of +0.88 points/per hand. That was a half point better per hand than when I was gambling aggressively in the previous match. I only made Fantasyland once (0.5% of the time), as compared to 7.6% previously, when I was taking much more risk to do so. However, I reduced my fouling rate to 19% of hands, which more than made up for not getting to The Land.
As with the aggressive-play breakdown, I’d take that +0.88 points/per hand figure with a grain of salt. The computer ran bad, going to Fantasyland only three times (1.6% of the hands), despite fouling 30% if the time.
In a longer match, I’d expect the computer to make Fantasyland on something like 2% of its hands, and thus pay off that high foul rate somewhat.
In any case, the “play like Ryan” strategy worked. It’s like the advice often given in poker books for playing against a maniac: “don’t build the pot, let him build it for you. Call down his bluffs, and don’t value-bet.” Similarly against an aggressive OFC opponent who likes to gamble for big hands, let him set the pace by being the first to lock into a draw, make sure that you don’t foul, and try to beat him in a single row, avoiding the scoop in case your opponent completes his risky draw after all.
Your opponent will be concerned with making his big bonus hand, so he won’t be careful to out-kick your hand in every other spot. Someone who plays aces in the middle won’t care what he makes on the bottom, as long as it’s at least two pair. You can give up in the middle (since you can’t beat aces), but make sure you make a better two pair to beat him on the bottom, with the extra bonus of scooping the hand if he fouls.
Similarly on top, when a player is going for Fantasyland, he will keep that queen-high or king-high open in order to pair the face card later. You can play a small pair in the middle and ace-high or a king/jack-high on top. When your opponent fails to pair his face card, your top is the winner and you won’t get scooped.
Solid but Not Passive!
None of this is to say that the solid player can’t ever make a big hand himself!
Even playing the role of a cautious player here, I don’t play the queens on the bottom to fill up my flush, instead holding out for an open-ended straight flush draw. I’m immediately rewarded with the and an (almost) risk-free sweat for Fantasyland.
Two cards later, my straight flush is still live, and I pair my ace.
Even if I wanted to back off of the pair of aces in the middle now, I couldn’t do it. It’s easier to foul with on top than with in the middle, given the rest of my hand.
With three cards left to play after I place the , I have ten outs to make a bottom hand to beat aces, which makes me a 2-1 favorite to do so (ten outs, three draws left, heads up ~68% chance to make it). If you won’t take a 1/3 chance of fouling for a chance at either a straight flush or a live Fantasyland draw you’re playing the wrong game!
I pull the next, and at this point I have to play it on the bottom. I can’t foul with a bottom straight, it’s unlikely that my hand will get scooped, and I’ve got three outs twice for Fantasyland.
I miss the my live queen the first time, and so that last card Fantasyland free-roll is a perfect time to do a full-screen 3D squeeze, which you can check out on my ABC Open-Face Chinese Poker app, where you can play your friends and squeeze any card.
Pulling back the corner, once I see that tip of his little sword, I know that this is the King of Hearts. I won’t be to Fantasyland, but I did have a shot at it, without ever taking much foul risk, and while keeping the possibility of a 15-point straight-flush bonus on the bottom as long as possible.
As it were, I lost -1 point on the hand. If I had played the for the flush earlier, I’d have won +3 points instead, and would have had a chance to scoop for +6 going into the last card.
Taking even a little bit of risk to make a monster hand has its downsides, but any player would have taken this gamble. It’s like a hold’em hand where a tight player tank-ships his straight flush draw on the flop. Yes, he’s taking some risk, and yes he wants his opponents to fold, but come on, that’s not what you call gambling!
Weak Hand, Aggressive Opponent
I mentioned before how to play against an aggressive player: the same way you would play against a maniac in hold’em and let him take the lead by committing to a draw. Don’t take significant foul risk, unless you have a chance at a really big hand. Instead, keep your hand flexible, so you can turn it into a hand that needs to just win a single spot to avoid a scoop, or completely shut it down and get a free +6 points when your opponent’s risky play turns into a hand that’s sure to foul.
Let’s take a look at an example.
This hand wasn’t much to start, and now I’ve pulled a , for which I have no obvious spot. It’s tempting to put the card up top, or in the middle to try to beat my opponent’s likely pair of 8’s or 9’s. But do I really want to pair that ten in the middle or on top? It’s not getting me to Fantasyland or to a full house, so why take the risk? My opponent is already taking a serious foul risk early in the hand by playing an ace on top. His hand has way more upside than mine, but he can’t qualify his hand without beating the ace in the middle somehow. That involves a good chance to foul. I’m going to sit back, wait for his hand to go south, and maybe make something like ace-high in the middle or two pair on the bottom. I’m taking no risk to foul, since my upside is not high, and who knows, I could always back into something.
My opponent immediately takes his foul risk from “moderate” to “code red,” while my path toward a relaxing low-upside hand is going as planned.
To pair the in the middle here would be as bad as hitting a 16 in blackjack against the dealer’s six. Stay pat and let him bust!
There is no need to improve my hand on top or in the middle. I just need to fill those spots in with whatever, and let my opponent bust. Improving the bottom could help me save a scoop in case he qualifies. If anything, pairing the six only hurts my odds of making the bottom, since I’d be taking some chance to foul the middle.
A couple of cards later, my opponent is drawing almost dead. Maybe 5% to make it.
… and he fouls. As it were, I pull a third and would have qualified even with the in the middle, but why take that chance?
My hand wasn’t much, but it was a hand that’s easy to play without fouling, and in this case, that was worth 6 points!
I learned a few things about OFC in putting this piece together. Most notably, I learned, much to my disappointment, that against an aggressive player, it pays to play solid, to not take too much foul risk, and to pick up those +6 point opportunities round after round.
The “correct” style of play, the Platonic Ideal, is somewhere between the very aggressive player and the cautious player who waits for his opponents to foul. But against specific opponents, it might pay to play it safe.
The computer was an opponent that I could beat in a number of ways, although my edge against it was smaller than I would have thought. In the first case, I beat it with aggressive play by fouling at the same rate, but making bigger hands more often. In the second case, I made a lot fewer big hands, but by having so many +6 rounds as to more than make up for it.
I also reduced variance by playing more cautiously, although the difference wasn’t so large . In any case, my advantage, even against an intermediate player, was swamped in a short match by the game’s variance.
Which leads me to my final point: no one is a big favorite against any decent player, at least not on a per-hand basis. Therefore, the most important play is probably to not misread your hand, or otherwise commit careless blunders. If you make a careless play every hour or two that gives your opponent a few free points, that is probably worth more than the edge that you may have on that opponent . Then again, the same is true of chess and backgammon and it doesn’t seem to hurt these games much. Poker is mostly luck, and yet we all still love it.
Good luck, happy squeezing, and I’ll see you in Fantasyland.
*Lead photo courtesy of mgorodinsky.tumblr.com.