With most information hidden, the game Stratego had stumped AI until now
Thread
Unofficial Hacker News client; not affiliated with Y Combinator.
With most information hidden, the game Stratego had stumped AI until now
Unofficial Hacker News client; not affiliated with Y Combinator.
hnedeotes · · focus · HN ↗
Those new additions can invalidate the whole training data by a single new "card" that changes completely the dynamics and would be easy for a player to understand and incorporate but not for an algorithm (perhaps with enough compute to re-train it regularly it could) - that along with the decision trees being orders of magnitude deeper, wider and with more conditionalities than go, chess or stratego - even through the same turn with the same cards available and same table state - would probably pose much harder problems for a compute bound algo.
qsort · · focus · HN ↗
There is nothing that, in principle, makes MTG different from poker or bridge, and we have superhuman engines for both.
hnedeotes · · focus · HN ↗
In my own game you don't have shuffle/draw randomness but the pool of options is statistically tending to infinite (if I would have 500 or 1000 scrolls designed and MtG depending on the format has that depth) when compared to something like chess, or this game. On the other hand in my own game you have to account for much more depth on the possible options your opponent has.
dragontamer · · focus · HN ↗
Ex: you don't really care if the opponent plays Giant Growth or Chastise. The effect is that the opponent is playing a combat trick, and combat has moved from attackers favor into defenders favor.
To defeat an instant speed combat trick requires a combat trick of your own, or a generic counter spell of some kind. Some have interactions (ex: Doom Blade beats Giant Growth but not Chastise), but the overall gist is that opponents can do things after combat is declared. You only need to keep track of how many combat tricks you think the opponent has.
---------
Other situations are card advantage (ex: 2 for 1. If the opponent spends 1 cards to defeat only 2 cards of yours). The traditional card for this is Mindrot, but well placed counterspell can turn a combat trick into. 2-for-1 reversal.
You don't necessarily keep track of how your opponent makes 2-for-1 opportunities. You just have vague gists of them.
---------
Good spells have huge applicability. Doom blade or Murder is high because killing opponent creatures at instant speed handles the vast majority of creature buffed combat tricks, and also serves as a way to stop enemy combos and other such tricks.
In contrast, chastise is very niche. If the opponent were playing like Swords to Plowshares (powerful white instant speed removal), it's pretty much always better than chastise.
If the opponent plays chastise instead, you take that as a win because you know they could have had a deck of better cards. But for whatever reason decided to play with weaker cards...
hnedeotes · · focus · HN ↗
But even then (not saying I'm right) I think the depth of choices, effects and so on, on a format like modern, or legacy, would be very difficult for an AI to top against pros. If you add draft into the mix it gets worse for the AI in my view too.
Because a good play in most situations can easily be a bad play under others. That doesn't happen in chess for instance, given enough decision depth to the algos to see the future game. In my own game I think those situations can occur much easier due to you always having your full deck available. Also, in MtG it's easy to get into table states that are either ahead/behind and then you kinda just have to protect your position (like with denial decks). Then you have the effects that you might remove a creature threat (graveyard) but then that enabling a combo you weren't expecting that needs a creature on the grave, or enabling delve cards or whatever have you. It's much less clear cut for a probabilistic model to make the optimal play at every single interaction. So the more you train the model on all the variations and possible follow ups, the more you dilute its certainty isn't it? In chess, or this game, or RTS such as starcraft, that doesn't really happen in my view.
dragontamer · · focus · HN ↗
No human can possibly keep up with all the possibilities or combinations that are accounted for.
Games of incomplete information have been IMO soft-solved as of.... Maybe 5 years ago? As in, stronger than any human can possibly reach (ie: massive GB-sized matricies accounting for all information iterated over millions of iterations of "he thinks that I think that he thinks that I think that....")
It's not a true Nash Equalibrium, which remains outside of the realm of even computers to compute. But a computer can always reach a closer / better estimate of any Nash Equalibrium, which covers all games of incomplete information.
--------
For Poker, it turns out that a few types of bet sizes (3x pot, 1.5x pot, pot, half pot, quarter pot) covered enough betting patterns to reach superhuman.
And frankly, MtG is simpler than the bluffing game in Poker. Like MtG has bluffs but it's no where close to Pokers level.
There's no crazy deep game for Red Deck Wins vs Control. The game basically plays itself out (Red tries to win before Control comes online. Control tries to stall before Red Deck Wins). There are some games with complex board states but they're largely a game of bluffing + card counting (opponent holds 4 cards, two of which were since the start of game and 2 were top decked in the last two turns. He at best has only planned for 2 responses or got lucky with the other two newest cards. Do I have a play that beats two cards yet?)
syradar · · focus · HN ↗
The possible game state is also much larger than poker. Deck construction alone is 60 cards out of about 30,000 unique cards. Sure, not all cards are viable in all decks, but we can have 1-4 copies of a card in our deck.
So we might not even know if we’re playing against mono-red or multicolored since the decklist is unknown. You can think you’re playing mono-red and then they suddenly play a Plains. Poker at least always has the same 52 cards to reason about.
I do think AI could be great at coming up with decklists though.
hnedeotes · · focus · HN ↗
hnedeotes · · focus · HN ↗
There's also the decision points, in poker it's way less. You're dealt cards, the table reveals cards, you bet/ante, move to the next, bet/ante. It's a very finite sequence of moves until disclosure. Plus it's 52 cards divided by the table players while on MtG it's 60 per players that you can't know (even lands can interact beyond being a resource and in competitive lists usually they do, specially in older formats) - although with enough/infinite time/memory you can probably generate a table for all possibilities, you still have to contend with the interactions other than the card types. A 9 Diamond is a 9 diamond. A Skull Clamp is a Skull Clamp but the way it interacts, its value/threat is highly dependent on context that can/might be hidden.
While I think that MtG is indeed "poker" like underneath the keywords, it's many more levels and I think bluffing can be way more "complex" but simply isn't because even at the pro-tour level prize pools are insignificant when compared to serious poker tables. Some MtG players are known to also dabble/play poker regularly.
I also think that the structure of poker game-play is more prone to be exploited by a competent bot - if you have a "budget" and you assign a bot to a table where the antes are "in-line" with the "budget" it has, it can mathematically (within a very high degree of probability) always turn a profit - ultimately humans fail in part because they enter "bluff" kingdom against a bot as the bots can just rely on mathematical probabilities. Made up numbers but the idea being, you have $200 to play. Choose a table where this allows you to play X games at least, say antes of cents, it should be able to make money most of the time at some point.
Yes, but as the other reply mentioned, the thing is you don't know if it's red deck wins, or a RDW with a tweak for the metagame and building the "he thinks that I think that he thinks" tables would probably require for practical terms what could amount to infinite storage and any of these chains, if followed through, can land the bot in a losing position hard to come back from. Now, to be honest, most MtG players aren't that good either, they play it more like a hobby/fun game, rather than approach it as poker/probabilities.