Amit Varma is a writer based in Mumbai. He worked in journalism for over a decade, and won the Bastiat Prize for Journalism in 2007. His bestselling novel, My Friend Sancho, was published in 2009. He is best known for his blog, India Uncut. His current project is a non-fiction book about the lack of personal and economic freedoms in post-Independence India.
A couple of weeks ago, I took part in the Match IPL, playing for Goa Kings. The IPL here stands for Indian Poker League, and it follows a similar franchisee model as cricket’s IPL. I don’t play poker these days, having retired a couple of years ago, but a good friend, who was the mentor of this team, asked me to join, and I thought it would be fun.
Also, one important reason I joined was because this was a new format of poker, and I wanted to set myself the intellectual challenge of understanding it and optimising for it. When I was active, I was mainly a live cash game player, with decent tournament results in the Asian circuit. But this format of poker was different in key ways from both cash games and tournament poker.
Match Poker is a format played by teams. It’s explained here, but I’ll sum it up briefly. Let’s say there are seven teams with seven players each. They play each other on seven tables, with one player from each team on every table. Also, there’s one player from each team on every seat. So if you are on seat 1 on table 1, all the other tables with have players from other teams on seat 1. So every team will have a player on every table and every seat.
The idea is that the same hand is then dealt across tables. So all teams play the same hand from every position. At the end of every hand, a team’s chips across positions are added. The team with the best chip count gets 7 points, the next team gets 6, and so on down to 1. The chip count is reset, and all teams start the next hand equal in chips. At the end of a certain number of hands – 200 in the case of Match IPL – the team with the highest points (not chips) wins.
The Question of Luck
According to the guys who thought this up, this format ensures that “the luck element in conventional poker via the ‘random draw of cards’ has been removed.” In fact, the Match Poker guys claim that because this removes the element of luck, that makes poker a true sport. They are using this rationale to get Match Poker into the Olympics.
This claim is both moot and false. It is moot because of two reasons, one small and one big. The small reason is that all sports do have an element of luck, and that’s doesn’t make them less of a sport. The big reason is that even though poker has a greater quantum of luck than other sports, it is still a game of skill in the long run. What happens in any one hand is largely luck, but given a large enough sample size, skill will make the difference.
Also, the format doesn’t eliminate luck entirely because there is still much variance in the game. Let’s say you are the best player in the best team. You hit a set, and maximise pot size to get your opponent with top pair all in. No other team manages this. But then your opponent hits a runner-runner full house, as he will 2% of the time. Your team played the best here – but you will come last, and the team will get just one point. This is luck, and it doesn’t matter in the long run because it all evens out. But you need a decent sample size of hands for the skill to show. 200 hands – or even 2000, or perhaps 20,000 – is not enough.
How Match Poker is Different from Poker
Although Match Poker is set up like a deep-stack cash game, it is different in two fundamental ways. One, the unit of measurement here is not chips, but hands won. Two, you are not playing against your table, but against all the other players sitting on your seat (and dealt the same hand) at the other tables.
Let’s start with point one: chips don’t matter. Teams are not ranked according to how many chips they win in a session, but how many hands they win. This is the opposite of regular poker. A study on online sites showed years ago that the players who win the most hands lose the most money. A good cash game player will lose more hands than he wins, but will win more when he wins than when he loses, and be overall profitable.
An illustration of this is set-mining. I will always play 44 preflop, if there is just one raise, and I will hit my set only one in eight times. Seven times I don’t hit the set – but the one time I do, I make enough money to compensate for the times I folded. But in Match Poker, that doesn’t matter. Point two explains why.
Point two: You are not playing against the table, but against other players on your seat. Let me illustrate this with the set-mining example. Let’s say you get 44. You fold preflop, while you know all the other players on your seat will call. Seven out of eight times, they will fold on the flop, and because you saved that preflop call, you are first on your seat. One time you are last. Assuming ceteris paribus (all other teams and players get equal results in other seats), your team gets seven points seven times and 1 point once, for a total of 50 points in eight hands. All other teams get 24.3, splitting the remaining points. Thus, while set-mining with small pairs is profitable in regular poker, folding them preflop is profitable in Match Poker. It is +EV in this format, or as I’d call it, +MPEV.
The same logic holds for speculative hands like suited connectors. If other teams are likely to play those hands, and they lose more than 50% of the time, your profitable move is to fold. Ditto for chasing flush draws on a flop. Remember, pot odds and chip EV don’t matter, because this is not traditional poker. So a lot of moves that +EV in regular poker are -MPEV.
With this thinking in mind, I formulated three thoughts that I wanted my teammates to think before every hand.
1. I am not playing this hand against the table. I am playing it against other players on this seat.
2. What are the players on this seat likely to do with this hand?
3. Will I win this 50% of the time?
If you have a speculative hand that your opponents (the players on your seat) are likely to call, and that hand will lose more than 50% of the time, then it is +MPEV to fold it right away.
So here’s what the points system means. Teams get from seven to one points for every hand. The average is four. It’s all zero-sum, so teams win what other teams lose, and the amount won is equal to the amount lost. You might have one team winning chips on a given hand and six teams losing, in which case the team that lost the least gets six valuable points. You might have one team losing and six winning. But generally, if you fold every hand, you should get around 4 points per hand. (Simulations validate this, FWIW, with the limited data I had from a previous event.) The team that won Match IPL won with 821 points from 200 hands, or 4.1 per hand. Three out of seven teams finished above the mean (800).
Now, obviously, folding every hand does not win you the whole thing. What I considered the optimal strategy was to fold all speculative hands and medium-strength hands, and push all value hands hard, but to define these value hands tightly. Also, profit in poker comes not just from value hands but value spots. Position matters, and there is much value to be had if you can outplay people in a button-vs-blinds dynamic. I’ll come back to this later: our initial strategy was based on not thinking too hard about spots and focussing on hands.
Instead of playing 20% of hands, as we otherwise might, we decided to play 5%, fold 95% and see how it goes. We would fold all speculative hands, all medium-strength hands (like KJs in early position) and we would also fold strong hands in multiway pots where our chances of winning are less than 50%. (Remember, in this format, pot odds don’t matter. 50% is the magic number.) We even made a hand-chart by position for our players to memorise. This was a new format and we were all beginners here, so that made sense.
We started the tournament disastrously. There were 8 sessions of 25 hands each, and in our first session, we were hit by variance. The problem was Seat 1. Our man in Seat 1 made a series of correct folds, (correct in terms of MPEV), and those hands kept hitting. Sets hit. Random hands hit trips. Connectors hit straights. Hands that would win one in eight or 15 or 25 times kept hitting. And other teams played those hands, and got points for them. We got on the wrong side of variance, which happens. But with just 175 hands left in the session, could we recover?
We remained in the bottom half of teams through that first of two days, though I was topping the individual charts at the end of day 1. In fact, I topped the individual charts at the end of 5 sessions out of the eight, but fell short of winning the MPV at the end of it. And this brings me to the problem of the individual leaderboard.
I assumed that the individual leaderboard would be calculated the same way as the team leaderboard: they’d see how you did against the players on your seat, and assign between 7 to 1 points for each hand. I assumed I led for so long because I was playing optimally. But I later found that this was actually being decided on total chip counts – ie, just like normal poker. (They were assigning differential points based on how you outplayed guys on your seat, but it was still looking at overall chips, not hands won.) Thus, doing well in the individual charts had no correlation to how well you played for your team. One was calculated as per chips, the other was as per hands won.
So given our strategy, how the hell was I leading for so long? Well, this brings me to the issue of value spots. You not only have to play value hands strongly, but also keep your eyes peeled for value spots. When everyone folds to you on the button, that could be a value spot if the blinds are passive. If you are in the small blind against a button open, that could be a value spot if you get him to fold. These are also high-variance spots, but spots you could win more than 50% of the time, so you have to use your judgement. (In this format, btw, I’d define a Value Spot as a spot where you can win the hand more than 50% of the time, regardless of the actual hand you have.)
I took the liberty of searching for such spots against the guy on my left, an excellent cash game reg who would attack my button from his small-blind. I couldn’t simply fold all my buttons, because players on my seat would probably win there a fair bit. So I had to play back at this guy. We had a 3b-4b dynamic going on, and in one hand I stacked him with Q7s against 96. (It went raise-3b-4b-flat, flop came Q96, with 7 on the river. Standard spot.) In another hand on the second day, in another 3b-4b spot, I took an all-in call on the river with A-high, in a spot where he would check-call all showdown hands because I was guaranteed to bluff. His range was polarised, as I thought, but he happened to be bluffing with bottom pair and I lost. The call was correct anyway, because once I have put in enough chips to ensure that I am last on my seat, there is nothing to be lost going all the way. (Ceretis Paribus, again.)
Again, the key rule with value spots is simply whether you’ll win more than 50% of the time, and what your opponents do doesn’t even matter here, because you’ll either get 7 points or 1. You beat the average by winning more than 50%.
So, the optimal strategy is to fold a lot, including all speculative and medium-strength hands, push all value hands hard (if you get coolered, so will everyone on your seat) and use your judgement for value spots. We started off unlucky, some panic set in, and we ended fifth. I slipped off first place in the individual leaderboard, ending 11th, but that was a chip-count thing. (If they calculated that the same way as team points, I’m sure I’d be higher.)
What was worse was that the winning team, after getting lucky on day 1, followed my strategy on day 2 and were thus uncatchable. One of them mentioned that they looked at my hand histories at the end of day 1 as I was leading the individual leaderboard, and their chief strategist is a close friend to whom I had boasted that I had cracked the optimal strategy. They didn’t necessarily get it from me, and this stuff isn’t rocket science to figure out. The fact that they shifted to my strategy on day 2 (one of them folded 24 out of 25 hands in one session, I heard) is intellectual validation that my ideas were correct – though I would have preferred monetary validation. SAD!
(No other team seemed to have figured out the strategy, by the way, with the team that came second playing a LAG style that is perfect for deep-stack cash games but sub-optimal for this format.)
For what it’s worth, I don’t intend to play this again, which is why I am being free with my thoughts here. I expect all the teams to read this, though, thus adding a metagame element, and making their subsequent search for value spots that much more fascinating.
The event was glitzy, with drone cams and so on, so watch it on MTV if you can. I haven’t seen it yet, and I am sure that if I do, I’ll be even more determined to resume my Keto diet.
The archives of Range Rover, my old poker column for the Economic Times.
The Five Commandments of Poker, an episode of the podcast Mera Kaam Poker that features me.
Sita Sings the Blues: The Greatest Break-Up Story Ever Told
Dev.D doesn't flinch from depicting the individual’s downward spiral
9 across: Van Morrison classic from Moondance (7)
6 down: Order beginning with ‘A’ (12)