CORDIAL MINUET ENSEMBLE

Having won the first tournament and, before that, being #1 on the leaderboards, I thought it'd be a good time to share some ideas on strategy. If you've been playing for a while, you may already know all of this. Because Cordial Minuet does the math for you, the learning curve appears to be pretty shallow. Which is great news for beginners! This is not a how-to-play guide. For that, check out Zach Fleeman's video.

CM is largely about mind games. It's about avoiding cognitive biases, avoiding fallacies, and exploiting the same in your opponent. These cognitive biases are how we normally think, so it takes some practice to move beyond them.

The Four Virtues of Cordial Minuet

Humility
Start out with $0.01 games and practice on those for longer than you think necessary. The higher the stakes, the more skilled the players. And you won't be very good to start. There are sharks out there eagerly waiting to take your money. When you're ready, try out a few $0.10 games to see if you can stomach higher losses.

Discipline
The best advice I can give you is Asminthe's Bankroll Management. To summarize: once you're ready to play for reals, stick to stakes equal to 1/20th of your current account value. The neat part is that its literally IMPOSSIBLE for you to lose all your money if you do stakes as a percentage of your bankroll. I followed this advice and grinded my way from $10 to $40. Low stakes gives you the confidence to play aggressively when need be and when you lose it won't be the end of the world.

The discipline part comes into play when you see a higher game and you think "just this once." You lose that game and your opponent starts a game even higher (probably their goal all along). You join that game, go on tilt, and lose a bunch of money. Don't do it. Ignore this advice at your peril. I've already witnessed the fallout of not following it.

Patience
Players tend to get impatient after several rounds and begin playing recklessly. This was compounded significantly in the time-limited format of the tournament! Good play involves folding a lot, even though that seems boring. Unless you're clairvoyant, most of your picks will be mediocre. Accept that.

Sometimes you'll get bullied by someone who bets 20 coins on every opening pick. No worries. You only need to win 5% of rounds to beat them.

Never convince yourself that your opponent "can't have a good hand again." This is gambler's fallacy.

Greed
Well, duh. Praise be to Minoson.

Column Picking

The goal: predict your opponent's moves and make yourself unpredictable.

The first tip here is that, at least some of the time, you should be picking randomly. I use a dice roller to make this easy. The dice roller has the added benefit of recording which columns I picked in the first round (which I would otherwise have to write down).

The most advanced thing you can do is to record those picks, so in the last couple rounds you can learn something about what your opponent gives themselves. This is very difficult to do, but will up your game if you can manage. Ideally, you'd review every single pick but that's beyond me.

Any sort of column picking strategy can be countered. Even random play can be exploited (just pick the column with 3 high numbers and you've got a 50-50 shot of having a great opening).

Here are some common strategies. Remember, these can apply to later rounds as well.

Given to you by opponent: lowest high number
The opponent finds the row with the lowest maximum number and gives it to you. This one is damn common and some players will use it nonstop. If you can predict this, you will have the advantage in most rounds. However, it can still be challenging to play against. It's a smart play for your opponent because it compresses your possible scores. Many times, it's less important to win the scoreboard then to APPEAR like you've won the scoreboard.

Given to you: a single high number
The opponent gives you something like a row with a single 33 or 34 and no other high numbers. This means on most rounds that this is awarded, you will have a mediocre starting pick. I use this a lot myself.

Opponent gives themselves: highest sum of high numbers
They might consistently give themselves a row that has two or three high numbers in it. For them, it appears smart because they've got a 33-50% chance of a great pick and the rest of the time, they can fold. But if you find them doing this, it is easily countered.

Ordering bias
Some players seem to be biased to a particular side of the board such as awarding you the last row frequently or only giving you one of the first 3 rows.

Another strategy you can try: give yourself a column that, if you get a high number in it, cancels out the highest score in the column you give them. It only works occasionally, but when it does you have a clear cut advantage.

Seeing What Your Opponent Sees and Crossing Things Out

This is something I don't think mid level players are doing enough.

Take the last round for example. You see 8 numbers (4 rows of 2). Often, you might think, "My opponent will surely give me this 1, so I have to avoid picking the column with the 1." Sometimes it feels like you always get that lowest number, but remember that your opponent is not seeing the same 8 numbers that you are. Two of those rows you see are already gone, so that 1 has a 50% of not even being available anymore. Furthermore, the greened out columns that you picked are still around for your opponent. You have to pay attention to what's in those even though it's initially tempting to think they're not important. Even if the 1 is still around, it may be less than significant than a 2 and a 3 in another row that you happen to know are gone.

The same applies to the 2nd round. If your first round picks for yourself eliminated some high or low numbers, remind yourself that the opponent still has to consider them.

Finally, if you're REALLY good at predicting your opponent, you can mentally "cross out" rows that they have awarded themselves. If they bet uncharacteristically high on the first round and there is only one high number in their column (like a 36), chances are they have that row. For the rest of the game, you can pretend like it's not even there. You can safely ignore any low numbers in that row because you've mentally eliminated them. Alternatively, a bet may signal only that they don't have a low number and you can be sure that those rows will stick around for later. This all assumes that you actually guessed correctly though...

Betting

Ideally, you would invest more when you have a higher chance of winning. When you were sure of victory, you would all-in. But Cordial Minuet is about not giving away your position, so betting according to your strength doesn't always work.

Tell your opponent what they want to hear

Your opponent has some concept of what a bluff, value bet (opposite of a buff), and limping looks like. Craft a bet that fits what they expect. For your average player, these might look the following:

Value bet (indicating a strong hand): 50%-100% of the pot. For example, if each player has already bet 10 coins, weaker players will almost always be willing to bet 10 more to continue, even in a weaker position.

Bluff: Something bordering on ludicrous, like double or triple the current pot. Also relevant: spending a lot of time making a decision.

Limping: betting the minimum possible (perhaps 0). Or risking 1 coin on the chance that the opponent has a really bad hand

So figure out what your opponent thinks is likely a bluff or limping and give them that when you actually have a strong hand. And scare them off with a meager "value bet" when you want them to fold.

As players get more advanced and they start second guessing these notions, you will have to modify your approach. The hard part is figuring out what kind of player you're up against.

Keep in mind that some people, especially beginners, can't be bluffed. This is actually good because it also means they can't be scared off.

A note on all-ins
Betting everything you have when you're guaranteed to win can be efficient (especially in a tournament), but a strong player typically won't fall for it. Until you figure out the strength of your opponent, however, it may be worth trying a couple all-ins to test the waters (only when you have a guaranteed win though).

If someone tries to all-in when they might have the highest score, they're almost certainly trying to fleece you. I won't say never because that would be too predictable, but you should very rarely call them on this, ESPECIALLY if a low amount of money is at stake.

I've seen people all-in on the last round when the pot only had 8 coins. Think about that. Why would your opponent risk 90 coins trying to make you fold when only 8 can be gained? This can only mean a strong hand to me.

Common scoreboard situations

Your actual score is the highest score
This is your time to shine! Reveal a low number, but the lowest possible number may be too obvious. Get your opponent to bet as much as possible, but don't scare them off. The absolute best case is that the pot is heavy from your opponent driving the bets (they have a pretty good score and you haven't tipped off your high score). An all-in might work, though it's only really likely if the pot is already huge.

Life doesn't get better than this

Your opponent might have the highest score and you have crap
Fold, son. A bluff would only work in rare situations.

Your opponent might have the highest score and you have the second highest
The most tense moments of Cordial Minuet!! If your opponent has it, they're going to play you for as much as possible. If they don't, they will try to bully you with fears of the former. Maybe you'll get lucky and they're not able to reveal the highest number. For that reason, I will sometimes still reveal a low number even though I'm scared to death (because a high one would just make them fold). There's one other piece of knowledge I've acquired on this:

In the very last round, you are shown two possible scores for your opponent. This seems like a 50-50 shot, but it's not. I came up with this because it looks very similar to the Monty Hall Problem. If you can accept the solution to that problem, the CM version is straightforward. Because your opponent will almost always choose to reveal the highest number (whether they have that number or a lower one), it is like the first door in the Monty Hall Problem. All but 1 of the remaining "doors" gets eliminated. And thus, the probability that they actual have that high pip is 1/6. In general, this means they probably don't have the highest number. Some caveats: a strong player might consistently get great scores. Also, earlier betting make give away their strength. And finally, please don't blow your whole bankroll because jere said you had an 83.3% chance to win (you know, the same odds you get in Russian roulette).

Other pro tips

For humility, watch Rounders. I just saw it the other night for the first time on Amazon Instant; it's a poker cult classic and almost all of the concepts translate to Cordial Minuet. When you're done, watch this gif on loop:

Then to balance out the humility, watch this