Do Not Find the Fox - Redesigned

[@The-EJ-Factor]

Time to turn it into a deck builder rogue like

[@MewtwoStruckBack]

Fox Joker gets added to Balatro in 2025?

[@JoaoVitorBarg]

​@@MewtwoStruckBack LOL Would be funny

[@lu_ck]

roguelike deck builders are the gaming equivalent of everything evolves into crabs

[@freescape08]

I fully agree; that was not a game, and you found a very constructive way of making it one. Love it!

[@AlexPBenton]

The 2 player version is in fact a lot like tic tac toe, but the difference is that you’re trying to *avoid* making the pattern, and force your *opponent* into making the pattern. Playing tic tac toe like this is actually dramatically more fun than regular tic tac toe, try playing where you and your opponent each try to avoid getting three in a row, and you’ll quickly see what I mean.

[@vispreni]

Yeah that's a fun spin on it, reminds me of the board game Quarto - here you're trying to avoid getting four in a row, but across four different attributes (color, height, shape, and consistency). Another spin here is that your opponent chooses the piece for you, and you choose its position. One of my faves: boardgamegeek.com/boardgame/681/quarto

[@joshuasims5421]

I had fun reimplementing find the fox in python, and reproducing his stats. Didn’t do much with the online version. The only problem was calling it a game, it’s more of a stats demo

[@AlexCheddarUK]

Really interesting video! I love the idea of revealing a few tiles face up at the start of the game! Very intreresting that your final version 2.0 is essentially what I landed on for my second challenge I'm currently attempting, although I also added some annoying Os on the board to make it more difficult than the first challenge (I may have ended up making it too difficult)

[@TheOne_6]

hello there fox guy

[@Pomodorosan]

Fox

[@vispreni]

Yeah I saw your FOOOOX attempts, I was so happy you went that route! Putting O's in the corners makes it easier though, doesn't it? They can't ever start or finish a F-O-X string in either direction, making surrounding positions safer. Does that leave only two Os in the face-down pile, or did you rebalance the letter count?

[@AlexCheddarUK]

@@vispreni I only have two Os in my pile. I picked the four Os on the board to 1) Make the game harder (the centre Os more than compensate for the "freebie" corner squares that can't be used) and 2) Make the game shorter since it's harder to fit in a one minute short when I have to chat about where I'm placing things! and 3) I thought a big imbalance in my stack of tiles would make the decisions more interesting

[@vispreni]

The logistics of making it fit into a Short make sense, I'd have to contemplate some more on points 1 and 3 - I always prefer emergent difficulty to one that's preset. I'm currently experimenting with having six of each letter for increased randomness, the leftover tiles keep you guessing. Have you considered creating a Print n' Play pdf that people can donate/buy and download off your website? Dunno if you're more passionate about selling books or making the game version.

[@WYXkk]

It is still identical to the original game, because they are both equivlent to 'randomly arrange the tiles first, then reveal them in some order', and the order obviously doesn't matter. It can only give more sense of strategy, but there is actually no.

[@kikivoorburg]

I’m not a statistician, but given that there are a fixed number of each letter tile, you should be able to use the tiles currently on the board as information that you can use to update your strategy right? I understand the equivalence to a totally random draw if the probability remains uniform throughout the process, but a fixed tile set means knowing the already-drawn tiles provides additional information on the probabilities of the next few tiles. It would be unintuitive to me if there’s no way to use that information to create better board set-ups. But of course mathematics doesn’t have to be intuitive, so I could be wrong!

[@thehexagon_yt]

@@kikivoorburg if you know probabilities, you can change length of the game, but not probability of winning. It's not very intuitive, but all probabilities are connected and order doesn't matter. If you need a further explanation, why it is so, I can provide it, but I'd rather not repeating all of that for 100th time - there are good explanations here in comments and replies, including mine.

[@gavin5410]

THANK YOU.... put my thought perfectly into words... it's not a game!!!

[@gavin5410]

Btw your game and your video are great, thank you for that too. Sorry to leave such a negative comment initially

[@johnjohn6763]

Placing facedown tiles wherever you want only gives the illusion of choice. At the end of the day it’s still a randomizer 4x4 grid

[@hisho2255]

Not quite, if you can place them down wherever you can know the probability of getting a certain letter based on what has been placed before it. Therefore you can make meaningful choices on what the most statistically likely piece is and place it according to that.

[@johnjohn6763]

@@hisho2255 I haven’t proved it rigorously but I feel strongly that it is not the case that there is a meaningful distinction here. Consider a particular pile of letters and a particular square. No matter what I do, the probability of that square being any of F O or X is solely dependent on what is currently in my pile. No matter at what point I decide to place my piece, no matter what information I gain in the future, there’s no escaping that that tile will have to be one of the tiles in my current pile chosen at random. If I stall placing the piece, I will simply know with slightly more certainty what it will be, but I am still ultimately placing a random letter at that position.

[@johnjohn6763]

@@hisho2255 a rough attempt to prove this in a more rigorous manner, I am not a mathematician so I could be wrong here. The claim I am attempting to disprove is: For any given bag and any given number of tiles, there is some strategy to place the tiles which impacts the probability distribution of what tiles go where in some way. The 1 tile case is trivial. The 2 tile case has you place either of your bagged tiles at random, then reduces to the 1 tile case. No matter which you choose, there is a 50/50 chance for either outcome. The 3 tile case is trickier. Let’s consider the bag FOO. I want to not place F in my first available square. All other outcomes irrelevant, does some strategy exist that can let me in any way whatsoever change it from a 2/3 success rate. Placing it right away accepts the 2/3 success rate, so we can’t do that. If we place on another tile, we have a 1 outcome that guarantees a success and 2 outcomes that change our odds to 1/2. Summing out all possible outcomes, we get three outcomes, one that’s a 100% chance of winning and two that are a 50% chance of winning, which adds up to… 2/3 odds. Since case 3 reduces to the trivial case 2, and no matter where the letter is placed it’s still 1/3 odds for any letter to go anywhere, case 3 is also trivial. This extends to any number of tiles or bag configuration. If you would like to disprove this, I’d consider the simpler “Don’t Find the Ox” game. You have 14 blank tiles and simply one O and one X. Find some strategy to avoid making the OX which is superior to random chance. I believe no strategy exists.

[@hisho2255]

@@johnjohn6763 oh wait you said face down tiles, nvm yeah it doesn't matter. I thought you flipped each tile after placing each one. If you only flip them after placing all of them then there is no meaningful distinction.

[@johnjohn6763]

@@hisho2255 I do mean if you flip them up.

[@memesalldayjack3267]

cool to see someone's thought process, just showing the conclusion is pretty good and preferable most times, but sometimes it's cool to see "what about this? nah it's not that good"

[@DanielLCarrier]

That last method means you can put it off for longer, but the probability of filling up the board without having "fox" is exactly the same. After all, no matter which blank tile you put in a cell, it has a 1/3 chance of being each letter. I suppose that's not quite true if there's a fixed number of each letter, but I still don't think strategy will make a big difference.

[@chaoticsilver8442]

There IS a fixed number of each letter, though. That's one of the core rules of the main game.

[@thehexagon_yt]

That's quite true even with a fixed number of each letter. There is an illusion of choice.

[@chaoticsilver8442]

@@thehexagon_yt Not really. Because, with a fixed number, you know the probabilities. The game gives you Six O's, with Five F's and X's. As stated in the game's videos, You want to have your O's, in the corners if at all possible- As it is IMPOSSIBLE to get a corner O to spell a 'fox' -Meaning... Playing Saving your corners for when you have a majority of O's. Say, you've, by random chance, already placed about half of your F's and X's, (Say, placed three F's and two X's), leaving you with two F's, Three X's, and you only placed one O... Going for the corners NOW would mean there's over a fifty percent chance you'll get an O there. (With five O's, and five F's and X's combined) And if not, and you end up placing an F or an X in that corner, then your chance is even higher to hit an O for one of the three other corners. Granted, it's still highly random chance. You'd be relying on the luck of getting to having a majority of O's remaining before you can actively TRY tactically place these O's... But, still, with an awareness of which pieces are left on the board, you can try to work with the odds, and place your pieces based on which type of piece has the majority, and you're most likely to get. So this alone does open up some level of strategy.

[@thehexagon_yt]

@@chaoticsilver8442 I explained this a lot of times already. But this doesn't work. You can get longer game but not better chance of winning. There is a fixed amount of ways to arrange 16 letters (2,018,016) and only part of it (255,304) don't have a fox. Order of you placing tiles doesn't matter, because you're still getting random tiles and still must get one of these 255,304 arrangements. Here's another explanation: place all 16 tiles without opening any. Now, open 5 of them, you get 3 F's and 2 X's. Will going for the corners increase your chance of winning?

[@chaoticsilver8442]

But you're NOT placing all sixteen tiles at once. You're placing them one at a time, and choosing where to place the remaining ones based off of what the odds of getting any particular piece currently is. Yes, there's a fixed amount of ways to arrange the sixteen tiles, but that amount decreases with each tile already set. With the knowledge of how much of each piece of each type remains, you can place your tile judging on where you would want the tile it's most likely to be, at, and where the tile it's least likely to be, to not be at. Does that mean you're guaranteed the tile you're most likely to get? Obviously not. It's still random. But the 'randomness' is set, and specific. And changes in specific ways judging by which tiles are removed, and which are still in the bag. And a player can actively decide where to place a tile based on what those chances currently are.

[@PrScandium]

Your final version doesn't change anything to the odds of winning. The finished grid has the exact same probability distribution since you never know which tile you place on a given turn. The player still has no strategy that is better or worse than placing them all randomly.

[@vispreni]

The finished grid does have the same probability (since the number of tiles remains the same), but only if you're placing at random. If you infer (or calculate) the probability of each subsequent tile being F, O or X and you place the tile accordingly, it's bound to affect the odds vs random.

[@vytah]

​@@vispreniSince you have to fill that file sooner or later anyway, the only thing you can do is change when you lose, not if you lose. You still have to fill it with one of the tiles in the bank. Consider the exact equivalent game: fill the entire board with tiles face down, and reveal them one by one. If you do it in a fixed order, then you have the original "game". If you can do it in any order you want, then you have your "game". But the end result is exactly the same: either the board has had fox from the beginning, or hasn't.

[@johnjohn6763]

@@vytah This is a very elegant demonstration. Good job :3

[@vispreni]

I think your example is exactly where I disagree (though I suck at math so I may be wrong) - taking a tile from the pile AFTER you've revealed one of the tiles changes the % of the remainder. First turn always has the same % - .375 of it being "O" and .3125 each for "X" and "F". Let's say you draw "F". Now the % is - .4 for "O", .3* for "X" and .27 for "F". The crucial thing is that you still haven't placed the tile in a position. Knowing the above odds I'd always put my next tile in a corner, counting on an "O". Whatever comes up I adjust accordingly, trying to nudge probability into a winning constellation. @thehexagon_yt I'm on the verge of making a follow-up video to dissect this, it's breaking my mind trying to understand where y'all are coming from.

[@johnjohn6763]

@@vispreniit may help to imagine the problem not as a 4x4 grid, but as a deck of 16 cards. I think the grid is mystifying the math in a way that is confusing. You have a deck of 16 facedown shuffled cards and want to find a way to “stack the deck” in some way, say you want to get as many aces at the top of the deck as is possible. You may swap facedown cards as you please. This doesn’t do anything since the deck is already shuffled, yeah? You may also reveal any of the facedown cards, but by doing so they’re “locked” in their current square. There’s no point to shuffling the cards, and revealing cards doesn’t give any advantage to what cards are where either. Both decisions lack any meaningful way to actually affect the ultimate outcome of the deck being random.

[@proton..]

a puzzle is a game though

[@thehexagon_yt]

3:57. I can ensure you that this game doesn't change probabilities in any way and only gives an illusion of choice, as some other people pointed out. However, making some of the tiles visible BEFORE game starts (3:01) is a feature that actually can affect chances. However, finding that out would be insanely harder than regular game for one simple reason - you can't just simulate it same way as common game. Yes, it is possible to make some sort of minimax search algorithm or bot like StockFish to play this game, but it's way more complex than regular strategy and regular chances, because involves both. 2:19, When I first discovered this game, I asked the same question. And that's why I made a python program that let you do exactly that, you play against a bot that is trying to make you get the fox. It's not really tic-tac-toe, because involves some luck too, and that's probably the best "game" you can get out of this idea, but I get why you decided to do something else.

[@vispreni]

If I have just F,O and I need to draw the next tile to the right of them (like the original), I have a roughly 36% chance of losing then and there due to drawing X. If I can place the next tile elsewhere, I have a 100% change of not losing. Excuse my lack of knowledge in maths - that's as far as I can claim when it comes to probability. It gets more complicated near the endgame and I can't ever guarantee being able to win, but choosing where you reveal the next tile based on the current layout gives the player agency to set the board up into a winning end state (of which there are a limited number). Take the leftover tiles, check how many of those are no-no's for a grid position, you got the odds of your next move being your last. If there's different options with different % of loss, the probability of the outcome isn't yet fixed. Does your bot know what tile it's playing, or is it going off random picks as well?

[@thehexagon_yt]

@@vispreni while it is true that by choosing where to reveal next tile you cab guarantee not having fox on next move (up to 50% of the board), it only changes length of the game, but not probability of winning. Here's another way to look on it: place all 16 tiles, letters down, before opening any. Now, does opening tiles that you want change probability of finding the fox? There is a limited way to arrange all 16 letters and only part of it doesn't have any foxes (255,304 out of 2,018,016 to be exact). You may argue that my example is different, since you're choosing where to _place_ and not _open_ . But mathematically these actions are equal. Sure, you may play "safely" after FO to guarantee not getting X on 3rd tile, but after some actions you may end up with chance of X higher than 36%, if on other tiles you only got F and O again. Or less than 36%, it's random after all. It's getting random tile and choose where to place it. Both player and bot can see which one they got before placing, but they get random from what's left.

[@humanperson2375]

I was thinking this and got to: you get to look at exactly 1 tile whenever, also place all tiles where ever

[@planespottermerijn]

Basically exactly my idea, and pretty much the same thought process too! Wonderful presentation on it

[@iamsushi1056]

Every time you win the grid size goes up. Every time you lose (optional-playtest!) it either goes down one or resets to 4x4

[@RickyRatte]

How exactly is the version 2.0 different from the initial version? It's still random tile placing without any skill involved

[@BoredYoshi]

the more tiles you place, the more you know about what's left in your "hand". you can use this knowledge and probability to your advantage. so, if there's a lot of Os on the board, you're less likely to place down an O tile, and should place accordingly.

[@thehexagon_yt]

@@BoredYoshi that doesn't change chance of winning though

[@BoredYoshi]

@@thehexagon_yt if you know you have 3 Os, 1 F, and 1 X left, you should place in a spot where an O won't make you lose

[@thehexagon_yt]

@@BoredYoshi doesn't work so. I'm too lazy to explain this for 1000th time, so long story short - multiplying probabilities will give you same results. For morr explanations, read other comments and replies. Some people (including myself) came up with great explanations and examples, someone even ran a simulation. Order doesn't change probability of winning, only length of the game.

[@BoredYoshi]

@@thehexagon_yt interesting

[@plushloler]

Interesting video, though I do not agree that a puzzle isn't a game. Puzzles are games, they're just more about finding the solution once rather than continuously replaying. I'm also not sure if your new version has that much agency for the player? A lot of the time you're still living it to chance, and when you see there's "fox" about to be spelled out, you can take a letter that prevents that from the known ones, or take a risk if you ran out of them (if you run out of them fully, the game is just random from then on). The other strategy bit is placing tiles such that they can't ever be part of "fox", but I'm not sure there's that much depth there. Some other people in the comments said the game is still fully random, which doesn't seem to be the case to me, but it's still very random. The way to test it would be to just play it over and over and over (maybe writing an algorithm that follows a certain strategy) and seeing how high you can get the winrate. I would expect it to be higher than the original game, but I'm not sure how high it should be.

[@harmoen]

I would have just revealed the top letter of each column. You'd only ever have one choice to make.

[@veikkakarvonen831]

Tried the game, I found the fox on both of my two first attempts.

[@Carlang29]

do not find the fox but you have to find the fox

[@sentientarugula2884]

This high quality of a video deserves 100000000 views.

[@everestbocholyan2806]

Fox

[@wilTfrie]

LET ME PLAY!

[@mitchellzemil4890]

My initial thought was like your first attempt, but only revealing the letter after it has been put in place. Kind of like minesweeper where you pick your next position based on the information revealed in past plays. Probably still not a particularly hard game, but with the random reveal each turn you'd have to pivot your strategy or avoid painting yourself into a corner. There could also be limitations on where you're allowed to 'build from' after your first tile is placed - like how Sagrada only lets you play tiles next to ones you've already filled in.

[@brendantw]

So basically just the final redesign? (Which by the way doesn't have any real strategy either, because the order in which you place it doesn't matter in the slightest)

[@mitchellzemil4890]

​@brendantw oop, hahaha yes exactly. Well I feel silly watching it on silent half paying attention to the captions, but yes thank you for keeping me in check lol. Anyways, highly recommend Sagrada - it's very good!

[@midasfury6165]

The only change to the probability is a minor amount of calculation via dependent variables

[@vispreni]

Alas, isn't that what most games are

[@johnjohn6763]

There is no change to probability. There's no way to change the probability of any tile being any letter.

[@nTu4Ka]

That's actually super cool concept!

[@sakshamgupta5022]

I saw the game initially on some short, and I had the thought did I not understand it correctly or how is this a game? So I understood correctly and shared similar thoughts!! Great video tho!

[@1rez378]

What if it was Do Not "Connect FOX"?

[@jay-tbl]

is there a playable version? maybe i should try to make one

[@siosilvar]

This is still not a game, the choices aren't meaningful because no tile or position is actually distinguishable with all the information hidden. Bring back the faceup tiles, that was a good idea.

[@akaidenki5526]

make it a sudoku

[@okram5210]

Dali će napraviti srpsku verziju?

[@vispreni]

Možemo da napravimo srpsku verziju al uzmemo tipa "ЏАК" kao reč :D da ih sve zbunimo

[@okram5210]

@@vispreni može👍

[@PopheadJo-if6ej]

The game should be more theatric, since its most fundamental premise tends to collapse into a dull math formula. A new word should be chosen every day, and the player should be allowed to chose their tiles, with a free space in the middle and with a few tiles already put in place for them. The inherent randomness of seeing the tiles but having a few chosen for you, coupled with a merciful free space, opens interesting decisions instead of math formulas. Winning feels special and personal each time, since the word is a different word every day. I also suggest boards that are not squares, but other shapes too, like circles, rectangles, or triangles. That’ll really keep players on their toes.