There is a fantastic blog post by Ross Rheingans-Yoo about the shortcomings of poker as a tool to teach trading and why Figgie is a better one. https://blog.rossry.net/figgie/
He also wrote an equally brilliant eulogy about Max Chiswick, one of his colleagues, who helped him develop it. https://blog.rossry.net/chisness/
1. If I have an obvious skew in the distribution of cards I've been dealt, I can assume the common suit and hence the goal suit and try to buy the cards of that suit.
2. If the distribution of cards I have is more or less even, then I just save that round and try to make back the base cost.
The hard part has been trying to understand how to update my beliefs based on the trades being made. If you assume everyone is making the rational choice, you might be able to come up with some strategy, but if its against humans who might be trying to bluff, I have no idea. Although, they say its a win-win game, so maybe theres a way there too.
This reminds me of how I got my first job in trading.
They had all the candidates going around doing various gimmicks like mental arithmetic. But one of the things we had to do was a trading game where you had to get a set of some commodity by trading your cards with other candidates. So just a screaming pit of kids trying to declare what they wanted to swap for.
I got dealt a ridiculously good hand, and all I had to do was wait for a couple of people to give me the missing two or three cards.
Once I started, poker was the game the bosses had us play. It seemed like a 10 quid subsidy from the head trader every day, I just had to wait around until he decided he'd played enough and went all in on what was always a crappy hand.
That essay about why poker probably isn't the best game for traders is pretty good btw. But there was a golden age of online poker going on at the time, and you could sit there with 5 tables open if you wanted.
I like the design of the game, having looked at it superficially. It's basically a market for cards, with what seems like a goal for collecting the right suit.
I downloaded the app to try it. It's interesting, appears that you can join a game with other players? I see two tabs: "4 starting" and "12 active" But clicking on these had no effect until suddenly the overflow broke rendering.
I am fascinated by this app mostly because it has a recruiting message at the bottom from Jane Street. I've never seen that before.
But, it's also interesting because it is one of the more poorly designed apps I've seen. It's react native when you check out the logs using adb. I bet this was a side project from a Jane Street employee. They did their best.
This game seems like it shouldn't work playing one-off games with no money on the line. How do you win in that context - by ending up with the most money at the end of the game? That would change the gameplay very significantly - a strategy that has you end with $1000 one time in three and 0 the rest should be favoured over one where you end up with $300 every time.
Poker solves this with poker tournaments: everyone starts with say $10,000 in play money and plays until one player wins it all. Backgammon solves it with match play: two players play first to say 7 points. In both cases, additional complexity is introduced to the game as optimal strategy changes depending on the size of the bets relative to the players' stacks or the number of points remaining.
Backgammon is just about playable as a one-off without a cube, though not by serious players. A single hand of poker without wagering makes no sense at all. Figgie seems closer to poker than backgammon, based on my reading. But it's being presented as a game to play one-off with bots or strangers. What am I missing?
Or let me practice against bots before going live.
Also the notice that my username was taken wasn’t clear (was in white) so i thought the app didn’t work until i learned my username was already taken by scrolling up and seeing the notice that didn’t stand out
Love this. I’d really like to play with real cards, but the uneven suit setup makes that tough as a party game unless the host sits out. I guess you could prepare a bunch of decks ahead of time. At 4 minutes a round you’d need maybe 10 decks for an hour of playing though. To even it out I guess you’d have 12? But that means you’d have more info about what’s likely in later rounds..
Hmm. Maybe I can nerd snipe some into making a deck sorter..
I looked at this. I don’t fully understand the game yet. One thing I do not like is the emphasis of speed. Humans should not be making fast financial decisions, trading or otherwise. Sure you could improve with this game. But you don’t have to be able to make fast decisions to be successful in finance.
For in-person play with ordinary playing cards and an arbitrary number of players, keep the rules the same except:
* Start with 1 deck per each 2 players. Shuffle the decks together.
* Deal 10 cards to each player, & set ante at a constant 50 chips per player.
* At the end-of-round reveal, the common suit is whichever suit has the most cards. If there's a tie, flip cards from the remaining deck until the tie is broken and the common suit is determined.
Reading articles like this I feel like I have an intellectual disability. I can't make heads or tails of what the point of these kinds of trading-simulacra card games are, or even how they work.
While this is a fun intellectually stimulating nerd toy for engineers, trading just doesn’t make sense to me.
Making massive sums of money from bad policies of governments, shorting companies or even entire countries and making money from wars by shorting and longing commodities is a repugnant way of making money and is immoral.
This is no better than crypto trading and gambling.
Jane Street's Figgie card game
(figgie.com)473 points by eamag 15 February 2025 | 117 comments
Comments
He also wrote an equally brilliant eulogy about Max Chiswick, one of his colleagues, who helped him develop it. https://blog.rossry.net/chisness/
The basic strategy that works for me:
1. If I have an obvious skew in the distribution of cards I've been dealt, I can assume the common suit and hence the goal suit and try to buy the cards of that suit.
2. If the distribution of cards I have is more or less even, then I just save that round and try to make back the base cost.
The hard part has been trying to understand how to update my beliefs based on the trades being made. If you assume everyone is making the rational choice, you might be able to come up with some strategy, but if its against humans who might be trying to bluff, I have no idea. Although, they say its a win-win game, so maybe theres a way there too.
They had all the candidates going around doing various gimmicks like mental arithmetic. But one of the things we had to do was a trading game where you had to get a set of some commodity by trading your cards with other candidates. So just a screaming pit of kids trying to declare what they wanted to swap for.
I got dealt a ridiculously good hand, and all I had to do was wait for a couple of people to give me the missing two or three cards.
Once I started, poker was the game the bosses had us play. It seemed like a 10 quid subsidy from the head trader every day, I just had to wait around until he decided he'd played enough and went all in on what was always a crappy hand.
That essay about why poker probably isn't the best game for traders is pretty good btw. But there was a golden age of online poker going on at the time, and you could sit there with 5 tables open if you wanted.
I like the design of the game, having looked at it superficially. It's basically a market for cards, with what seems like a goal for collecting the right suit.
I am fascinated by this app mostly because it has a recruiting message at the bottom from Jane Street. I've never seen that before.
But, it's also interesting because it is one of the more poorly designed apps I've seen. It's react native when you check out the logs using adb. I bet this was a side project from a Jane Street employee. They did their best.
Poker solves this with poker tournaments: everyone starts with say $10,000 in play money and plays until one player wins it all. Backgammon solves it with match play: two players play first to say 7 points. In both cases, additional complexity is introduced to the game as optimal strategy changes depending on the size of the bets relative to the players' stacks or the number of points remaining.
Backgammon is just about playable as a one-off without a cube, though not by serious players. A single hand of poker without wagering makes no sense at all. Figgie seems closer to poker than backgammon, based on my reading. But it's being presented as a game to play one-off with bots or strangers. What am I missing?
Give me one button and throw me into a live game.
Or let me practice against bots before going live.
Also the notice that my username was taken wasn’t clear (was in white) so i thought the app didn’t work until i learned my username was already taken by scrolling up and seeing the notice that didn’t stand out
Hmm. Maybe I can nerd snipe some into making a deck sorter..
* Start with 1 deck per each 2 players. Shuffle the decks together.
* Deal 10 cards to each player, & set ante at a constant 50 chips per player.
* At the end-of-round reveal, the common suit is whichever suit has the most cards. If there's a tie, flip cards from the remaining deck until the tie is broken and the common suit is determined.
Would be way more fun having some mascots to trade instead of common cards
is this the same game played by FTX sam in that Michael Lewis crypto biopic ??
Making massive sums of money from bad policies of governments, shorting companies or even entire countries and making money from wars by shorting and longing commodities is a repugnant way of making money and is immoral.
This is no better than crypto trading and gambling.