My game is broken I think, each time I reload the game I'm having to play through the tutorial, and a lot of the times so far through the whole thing freezes, and the text goes, if I'm lucky enough to have music then it's goes jumpy like a broken record, all I can do is close the game down. I'm gonna try one last install but if it happens again I'm kinda scared to play in case I end up on a good roll and it bloody dies on me, I don't understand why it's started doing this as up until now the on,y issue I had was the sound occasionally cutting out. I'm glad my data is still on gamecentre as my husband doesn't believe my score but then he's crap and jealous lol.
I had something similar happen to me, and I have had a couple of goes where I have slid two threes together but it has stayed as a 3, not changed to a six. Nothing that stopped me playing though Take it you have done a reboot.
Thanks, it took me at least 300 games to reach it so I need to play a lot more to reach 1536, I think.
Ok game seems to behaving for now, I've even got my old data back, I restarted my iPad, dunno if it's coincidence or whatever it I'm relieved to see my his ire list back. @roger congrats.
Not odd at all,at 7.30 pm my husband and I sat down for a few games on our iPads, little did I know he was also playing threes next time we looked at the clock it was 11pm, it honestly only felt like we'd playing half an hour lol.
phew! that's good to hear. Sorry about that hiccup there, cloud. seems like it was some major iCloud weirdness. we haven't heard of anything like this go down other than with your case and i'm just glad it was retrievable. please keep us in the loop if further issues arise, email here: http://asherv.com/threes/support/
My current understanding of the card draw algorithm Different explanations for the various sequences I have graciously been given by volunteers have been fighting it out in my brain for the last couple of days and I now need to write something down before everything devolves into chaos up there. Don't worry, I'll spoiler everything out for those of you who don't want to know. Spoiler 1. Systematic problems Recording games without error is hard. Really hard. You are looking at a board of numbers, you are thinking about numbers and how to combine them and what numbers they will form and as the game progresses you get more and more excited for your progress and on top of all of that cognitive effort you are tasked with writing down the correct numbers of cards that you draw. There are just bound to be mistakes because we are humans and we get confused. The very difficult part is now identifying these mistakes and either correcting them or dropping parts of the sequence if correction is impossible. This is a chicken-and-egg-problem because in order to notice mistakes one would first have to understand how the sequences are supposed to look (i. e. understand the algorithm that created them) which is precisely what we are trying to do here. And here comes the next problem: even if you think you understand the algorithm and the first 80 % of a sequence conforms to your predictions before things start to look weird that doesn't mean that the error occurred at (or even near) the point where your algorithm stops to apply. The error could have happened way before that and just by chance, nothing looked suspicious. The result of all that is that we have a significant amount of noise in our data. Stuff that we can't get rid of and somehow have look past. The bad news is that getting a script to look past that noise is hard - doing it as a human is easier but less reliable. This brings us to the next and probably more important part. 2. Current concept Following are the observations that I and others have made that appear to be true even though you can find counter-examples in the sequences for some of them. • The board always starts with 9 cards. There is no argument here, this is the one thing on which all sequences agree. As a corollary from below, these 9 cards always contain at least one 3 and at most 4 of either color. • The highest number you can draw is 3 ranks below your high card. If your high card is 24 or less, you only draw 3s. With a high card of 48, 6s start to show up. At 96, 12s come in and at 192, 24s appear. There are several sub-theorems to the bonus cards: Spoiler • The probability for each possible bonus card value seems to be equal. This means that as your high card increases it is more and more unlikely that you will see the highest possible bonus card. On the flip side, the value of bonus cards (even if they are announced in the coming update) becomes less and less predictable as the high card increases. • I am not sure when a new high card takes effect. It either does so immediately (during a running stack) or at the start of the next stack (see below on the topic of stacks). Figuring this out is particularly nasty due to the problems listed in section 1 and the very low probability that you see an appropriate bonus card after the high card increases but before a new stack begins. • It would appear that bonus cards happen in every other stack and rather than appear instead of a 3, they are just added to the stack. There are examples of 2 bonus card in one stack but again, I am not yet sure that this is possible in the algorithm or a result of errors. • It could just as well be possible that bonus cards are completely unrelated to stacks. Instead, there is just a constant 5 % chance on every move that your next card is a bonus card. • It is entirely possible that 3s can also be drawn as bonus cards. Proving or disproving this will probably not be possible without a large number of error-free sequences (i. e. never). • The game uses an undrawn stack of 12 cards (four 1s, four 2s and four 3s). A stack is created at the beginning of a game and cards are drawn individually both to fill the board (see above) and to be added with every move. When a stack is depleted, a new one is created. I know I posited this theory before and then retracted it but it just fits to many beginnings of sequences (where I suspect the error-rate is lower) and I still believe in it. This has some consequences: Spoiler • At most you will have 4 1s or 2s on the board without their respective counterpart present. If you have 4 1s or 2s on the board, you will draw the corresponding other 4 very soon. If someone completes my challenge this theory will be disproved. I am however very confident that this holds true. Preventing a flood of 1s or 2s is exactly what the card draw algorithm is supposed to be designed for. • In theory, this allows you to anticipate more than your next card as more and more of a stack is drawn. For example, if you draw three 2s and four 3s (in some order) you can infer that the next 5 cards contain one 2 and 4 1s (in some order). However, in order to do that you must never miss a card in your counting and precisely know where a stack begins and ends. Especially in a later game with bonus cards showing up, this becomes more and more challenging (see section 1). I stated earlier in the thread that one of the merits on which the validity of any scientific/academic theory should be judged is the ability to make correct predictions. To this end, I have arranged to Skype with another ambitious Threes-academic (or "Threesearcher" ba dum tsss!). He'll be playing and telling me the card draws while I concentrate on logging and predicting cards. I will post results tomorrow. Sorry for the black wall of text, but I feel better now.
For all to enjoy The ultimate 6144 "Volleo" tile card has been discovered! Once again, I believe this is the first time that this tile card is scored in a genuine legit game. Here are the critical screenshot that made this run possible and the winning screenshot of the tile card itself: I completed this game over a number of play sessions. I did not track my time, but it likely took me about 10-15 hours (most of the time spent planning my moves, obviously). Like my discoveries of the 1536 and 3072 tile cards, here are some notes of observations concerning this particular run: 1. Once again, it was a random "high number" tile card (in this case, a 384 tile card) that made this run a success. 2. The most difficult part of the game is to learn how to get out of a potential jam, the most dangerous of which is "staggering". This occurs when a "low number" tile card appears between two very "high number" tile cards. It is very important to remove staggering as early as possible (without replacing it with another staggering). 3. I don't use so-called center or corner strategy. Instead, I make it a priority to keep the "high number" tile card that I want to match against the wall, preferably not in a corner. This way, when a random "high number" tile card appears on the same wall, I can get to that card quickly to match. 4. While the game undoubtedly requires skills to win, the "element" of chance plays a significant role in this game. In fact, I would argue that chance dominates over skills in the later levels. I found it simply too difficult to maintain two separate chains to create two identical "high number" card tiles to merge. Instead, my strategy is to create only one "high number" tile card of each kind, so that whatever the random "high number" tile card appears, you can make use of it to escalate. 5. The game is quite taxing to play at the later levels. Near the end, I was keeping count of the 1's and 2's that were appearing and would frequently change my strategy when I could count on the fact that these tiles might not appear for awhile (assuming the stack theory is correct; see #6). 6. I, too, am convinced that there is some unknown stack from which the tile cards are drawn and this stack gets renewed and reshuffled. 7. I am fairly convinced that, given a number of open rows or columns where a new tile card can appear, their probabilities are NOT equal. More often than not, the new tile card would appear in a "less" favorable row or column instead of a "more" favorable row or column with which I could do an immediate match. I am fully aware of the potential issue of "recall" bias, so I welcome other players' impression of this theory. Anyhow, as the 6144 tile card is the "true last" tile card, I will retire this game at its current state. Alas, doing so also means that I won't get a "final" score for my run. It has been fun! Congrats again to the team at Sirvo for creating this wonderful game!
First of all: wow, congratulations. What a progress! But you are not done yet. Now that you have unlocked "all" cards it is time to face the *lock achievements, starting with Threelock and (I assume) going up to Ninety-Six-Lock. In lieu of an official end-score, I hereby present you with the value of your current board (as far as I can reconstruct it from your screenshots) and minimal score should you end your game: 531,765
Thanks. I am going to make an iTunes backup of my entire device, which should save the current game progress. This way, I can delete the app and restart from scratch. Like I said before, I put too much effort in this current game that I just could not bear to see it end. I also want to congratulate you, kamikaze28, on all of your wonderful work you did on establishing the theory for this game. As I explained in my notes for reaching the 6144 tile card, I used your theory of the "card stack" at times when I really needed to "count" on certain cards to appear to try to get out of a jam.
Did you get the sequence written down for kamikaze? It must be insane. Also, is this the same game that brought your first 1536 and 3072? Do you realize that the last 1 point achievement is actually the next tile, the slightly changed Volleo one?
Great work Kamikaze, again. Spoiler However, your theory fails to incorporate my theory of the non-3 card coming opportunely (i.e., when a 3 would be an issue). Now y2kmp3 has stated that TWICE this allowed him to break the world record. I don't believe in coincidences. Also, from some of the sequences it seems that the new largest card has appeared first, though others disprove that. Finally, to address y2kmp3's idea of the game being biased against you, I find that the new cards often come on the bottom row. If your strategy revolves around there, your setup will get moved around more from cards appearing there. Try making a top focused setup for a game.
Good god. I've only gotten to around 3,000 points. You guys must be deeply invested. Anyway, this game is just so pleasing. It's cute and colorful and the music is great. I normally don't like simple games like this, but this one has proved me wrong.
Well done y2k, As an extra measure you can make a back up the save file using ifun box, you don't need to be jail broken and you should be able to pop the save file back on and continue the game should you wish, it might also be possible to copy the save file, then load one one and finish, you should still have the it her file in tact. I don't think I would want the game to end either so would probably do the same as you. Wtg again.
This game is so maddening, people have obviously become dangerously obsessed with the theory and hidden design. Be careful guys, if you figure too much out you will break the game. I broke Dungeon Raid, sadly realizing too late that it was impossible to die and so my high score was... Infinite... I kept a snapshot of the game for a while and tried various crazy ways to die, but I had simply broken the Matrix.
