Poker: limping in early position and other lousy strategies
Every player who sits down to a game of poker executes their chosen strategy, like a kid with a lemonade stand. Some strategies (like some lemonade stands) will make money, others will not. The strategy of a pro would take thousands of pages to describe in full. The strategy of a recreational player could be written on an A4 piece of paper, with room to spare.
The important elements of a poker strategy include how many hands the player contests, how passively or aggressively they play, whether they bluff, and whether they are sticky when calling down. Strategies are guided by heuristics such as: ‘always check to the pre-flop aggressor’; ‘3–5 big blinds is a good amount to raise’; and ‘play more tightly when out of position’.
One heuristic one might advise to a rank beginner is, ‘don’t limp in early position. If you’re first to act, either fold or raise, but don’t just call the big blind bet.’ Yet I know players who have played thousands of hours of poker who still just call the 1 big blind when first to act. I don’t say this disdainfully; there are good reasons why weak poker strategies aren’t forced out of the strategy ecosystem.
Why is limping in early a weak strategy?
A beginner might just accept their teacher’s advice not to limp in, and will note that top players and computer solvers never do it.
The problem is that good seldom comes from this strategy. Sometimes an early call invites later players to limp in behind and it doesn’t drive out the blinds. Our hero ends up playing a multi-way pot out of position. Other times they get raised, and face the unenviable choice of either folding or calling the raise and playing a large pot out of position.
Why poker is not like chess or tennis
Let’s say limping in early position is the poker equivalent of playing chess by moving the pawn in front of your king and marching your king up the board over the first 10 moves. Now you could tell a beginner that this is a bad strategy. If they don’t take this message to heart and play this way anyway they’ll lose their first 10 games of chess. The error of their ways will be viscerally clear to them.
Poker isn’t like this. Unless a player is wont to reveal their cards to the table, or to fold a royal flush on the river, no strategy is really THAT bad in the short run. Limping in the lojack in a 6-max game is worth about -120BB per 100 hands (more on this statistic below), a terrible win rate by poker standards. But it means that if they’re playing the very low-stakes poker I play, that mistake is barely costing 50c a hand.
Most of the time they’ll just lose 1 BB, if they get raised preflop or whiff the flop and fold. Sometimes they’ll even win a very big pot, such as the 170BB win that one player made in one pot calling from the lojack in the database I analysed (see below). The feedback loop is very weak.
Why do some players continue to limp in early position? Perhaps they’ve never watched a poker youtube video and don’t know it’s a weak strategy. Fair enough.
Perhaps they’ve heard it’s a weak strategy against perfect opponents (such as a solver), but believe that it might work against their player pool. It is true that many recreational player pools are less likely to ‘squeeze’ (raise) them preflop. So at least they get to see the flop with their medium strength hole cards. On the other hand, their opponents are also more likely to limp behind, so they go VERY multi-way to the flop. These pots are hard to win.
Perhaps they know it’s a poor strategy but they just don’t care. Remember this is low-stakes poker. The financial implications are negligible. This is an interesting perspective for me. It would be like me going to my weekly tennis game and seeing everyone serve over-arm. I announce that I have no injury preventing me from serving over-arm. I’m able to serve over-arm. But I nevertheless intend to only serve under-arm. ‘But that’s a poor strategy’, my tennis mates will tell me. ‘But I just don’t care’, I say.
The difference is that tennis is equally fun to play serving over-arm or under-arm. Folding or limping in early position is not equally fun in poker. Folding is boring. Poker played properly is arguably a pretty boring game.
So just how bad is limping in early position?
If you’re read some of my other medium blogs, I purchased a few million micro stakes poker hand records in December 2023 and did some data mining on this data set, using python. I’ve returned to this data set to calculate the win/loss of limping in first to act in a 6 handed game of poker.
The hand records were of 1.35 million hands played in Nov/Dec 2023 on poker stars, at $0.05/$0.10 NL 6-max. They were set out in 23 folders and 4151 text files, and took some hours of python plus chatgpt to extract what I needed! (As an aside, you learn about ‘edge cases’ when you analyse a data set this big. Players named ‘errorFLOP’ or ‘John_!.:’ play havoc with your expression recognition code …)
Of these 1.35m hands, I eliminated hands that didn’t have 6 starting players, or where players were sitting out, or got disconnected, where there was a straddle, etc, leaving 0.74m hands.
The win / loss of these 0.74m hands was as follows:
In a computer World, the lojack (LJ), hijack (HJ), cut-off (CO) and button (B) would only open hands pre-flop that have a positive expectation (EV). In the long run (i.e. 0.74m hands) this positive expectation is borne out. In that sense, the LJ loss of -$317 in the above table is an aberration (albeit a small one). As expected, the win rate improves the later the position.
The small blind (SB) and big blind (BB) both make losses, as expected since they are forced to enter the pot. The worst starting position in this empirical view is the BB, which lost $41K in aggregate, though a computer may say that it should be the SB. The small-blind ante is no compensation for not being able to close the action pre-flop and having to play every later street out of position to everyone. It may be that low-stakes players feel the need to defend the big blind and hence play this position particularly badly.
Needless to say, the most profitable position to be in is the casino! Pokerstars took $49K in rake putting on this game. Note that in solver land the rake gets extracted from the SB and BB. The other positions adjust their opening ranges (compared to no rake) so that they still only open hands with >0 EV after rake.
Breaking down the LJ position further:
In 0.57m hands, or 77% of cases, the LJ simply folds, for zero profit or loss. This is admirably restrained, much more conservative than my player pool! A solver would fold even more often, 83% of the time. It’s very disadvantageous to play in early position in poker.
Players that raised pre-flop made a good profit of $4.5K, perhaps reflecting that these were stronger players. Finally, ‘re-raises’ represent players who initially called but then 3-bet a subsequent pre-flop raise. This is a well-known ruse for players holding premium hands (QQ, KK, AA, AK, etc). It is very profitable, 193BB per hundred, though these holdings were always going to print money. Once the player re-raises, the ruse is up: the player field would (or should) then fold.
Finally, we have the 42,509 hands that called in the LJ, and made no aggressive action. This action lost $5K, or -120 BB per hundred hands, the statistic I quoted earlier. Of these hands 11,554 lost exactly $0.10. Either they faced a raise pre-flop and folded, or they whiffed the flop and gave up then.
Strategies that survive the strategy ecosystem
Theoretically and empirically, limping in early position is a money-losing proposition, yet it survives in the strategy ecosystem of recreational poker.
In some activities (e.g. golf) it is immediately apparent that a stroke was played poorly (the ball was sliced) and also how to fix it. In others (e.g. tennis) it is obvious that your forehand could be better (no top-spin, say), but executing a revised strategy to play a top-spin is not easy. In poker, it is sometimes not even obvious that the play was bad (e.g. there’s a thin line between the stupid call and the hero call!). As noted earlier the feedback loop in poker is weak.
If players who called from early position were directly ‘exploited’ (in poker terminology) by the opposition by being 3-bet, they would feel a bit more pain. They’d often lose that 1BB without even seeing a flop. Even so, the pain is small.
Turning to other strategies, it’s a truism that (most) recreational players bluff too seldom and call too often on the river. If players were exploiting the bluffing imbalance, they would start to call less often. This in turn would incentivise players to bluff more. Then players would call more. And so on, until a computer-esque balance is reached. And yet the prevailing strategy imbalance lives on.
Though we shouldn’t be surprised. Consider insurance. Insurance practitioners speak of the ‘anti-selection death spiral’. This is where the best risks leave an insurance pool if they are over-charged relative to their likelihood of claiming, prompting the insurance company to increase prices, prompting the 2nd best risks to exit too, and so on until the insurance pool collapses. Yet insurance customers don’t always act rationally, through lack of interest & knowledge, and even community-rated insurance systems still manage to survive (e.g. Australian private health insurance).
Similar to the failure to exploit weak strategies in poker, I’m sure I could select against my insurance company a bit harder by shopping around. But I couldn’t be bothered. And this is to potentially save $100s of dollars in premium, not 50 cents!
Conclusion
In one sense, poker is an easy game. I could probably teach any newcomer to chess how to beat the median player on chess.com. But it would take a year of chess coaching. In less than 30 minutes I could teach a newcomer to poker how to beat my player pool: ‘Only play the best 50% of starting hands pre-flop. Don’t limp from early position; either raise or fold.’ Actually, that only took 10 seconds to explain!
But I know what that newcomer will say: ‘Sounds boring. You’re probably right, but I don’t care!’