Threes/2048 - Example of intrinsic difficulty curve from game design

edited in General
[edit: changed the title to show Threes, which was the original game... just feel bad about putting the clone up without knowing about the original]

[Edit: link to Threes http://asherv.com/threes/]
http://gabrielecirulli.github.io/2048/

Love how the rules of the game intrinsically ramp up the difficulty as you progress. Essentially the further along you get, the more high number tiles you have to deal with. No need to add in any difficulty curve with additional rules, it's right there in the design itself. That said I couldn't break past 512 after getting so close for friggin' ages, so maybe the inbuilt difficulty curve is too hard :p

Hope it brings you entertainment fora bit. Otherwise anyone else have good examples of game rules that naturally/intrinsically create the difficulty curve? Most, if not all, of the style of game where you have similar things combining to create more powerful things in a limited space have this sort of effect. E.g. that gem game I can't remember the name of right now [on reflection that sentence wasn't that helpful, anyone know what I'm talking about?]. I remember watching an extra credits video about it (if I find it I'll post it here), anyone seen any other interesting examples of this sort of effect?

Comments

  • Got to 256, what a cool game! It's like triple town lite :) I don't know how well they "control" the difficulty cos basically that depends on the tile spawned - if it's something that is smart enough to always "fit" the situation you're in then dannng well done, my maths can't even fathom how (take the 4 blocks next to the blank space found and use the lowest value there, best I got :P)

    Triple Town kinda fits the model you're talking about, then there's 868-HACK, then something like Crimsonland is kinda like that - the difficulty is just in the increasing rate of enemies spawned, which is arguably the same as the start of the game (it's a constant rate of increase). Dead Run, does that count? The game gets harder by the same increment by pushing the player forward in the screen without making the game ACTUALLY different in any other way.

    Would any game with a cumulative burden/obstacle fit that bill?
  • 7284 points, 512 max XD lost to rookie mistake XD
  • edited
    Gotta wonder if this came out before or after Threes! on iOS. I kinda feel like it's a bit of a clone, especially given the art direction... Controls are slightly different though and Threes! has better tactical allowances.

    -edit- Yeah, checked out the game on iOS that this links to. Fuck that noise.

    P.S. Yes, I get that you "can't do much about this" but it's still predatory and bullshit and shows an underlying character that's both entitled and gormless beyond belief.
  • Oh? Is it a clone? I haven't played Threes, is it basically the same thing... in multiples of 3s?! Seriously??
  • I love this game...
    and I kind of dont mind if its like the other games mentioned cause I played this first so the others will be the "clones" to me.
  • Firstly, the 1024! game linked to is complete bullshit (as pointed out). They even say "No need to pay for ThreesGames." I can't actually try the game, so I have no idea how much is the same as Threes!, but it leaves a bad taste in my mouth to say the least.

    Concerning 2048: tried it and it was quite fun, but there is quite a large luck element as far as I could tell. Haven't played Threes!, but the similarities are definitely quite striking. I see that 2048 lists Threes! at the bottom of the page, and they aren't trying to compete in the same space as Threes! (I believe Threes! is iOS only), so I have less of a problem with 2048.

    Considering the links and description of 1024!, it would seem like the order of release was: Threes!, 1024! and finally 2048.
  • edited
    [edit: please note I've not played Threes, I don't have an iPhone to try it on, so I'm just going on the website's gif demo]

    The only difference in the rules is the start condition, the rest seems all the same, no?

    Moving the blocks around are the exact same rules between the two (2048 and Threes), the combination possibilities in the beginning are the only difference (assuming they both do block placement randomly - the algorithm to place blocks might be different but they are probably both randomly placed)

    for Threes: the blocks are
    1,2,3,6,12,24 etc

    for 2048/1024 the blocks are
    2,4,8,16,32 etc

    replacing this with a,b,c,d,e,f the rules become, all blocks move in a direction and combine with the below rules, a,b is randomly spawned on a blank cell

    Combine rules for Threes
    a+b = c
    c+c = d
    d+d = e
    etc (cannot do a+a or b+b)

    Combine rules for 2048
    a+a = b
    b+b = c
    c+c = d
    etc

    This makes 2048 slightly simpler and easier as there is more chance to get a+a and b+b whereas Threes you have to find 2 different blocks... actually on reflection since a and b are randomly spawned it actually is the same difficulty since you have as much chance of a+a, b+b, a+b or b+a, appearing in a row ... actually Threes might be easier

    ... soo err... it's a blatant rip that has a subtle difference that isn't even a difference at all... *cough cough* ... what was I doing here again?

    @Tuism r.e. your first reply, I'm not sure that a constant ramp up in enemy spawn rate is what I was thinking (not that you were wrong in saying they might be in the same 'class', I'm just struggling to differentiate them). Is it fair to say that increasing the spawn rate constitutes as a game rule specifically (or directly) for making the game harder, whereas in Threes/2048 the game rules exists and the difficulty is indirectly (almost deviously) increased as you get further in the game?
  • Threes started like 2048. It has evolved and turned into the shape Threes is in now as a response to fixing gameplay issues that Asher noticed... This 2048 thing would be fine if it didn't so obviously bite Threes' style and reference the game so hard, it's a shit implementation of a better game just for recognition because the other thing is good. Without Threes, 2048 wouldn't exist.

    Anyway, actual relevant gameplay differences, stuff that Threes does to allow for strategy:

    A move only slides blocks 1 space, they don't move as far possible in a direction. This means you can be a lot more tactical with your board space and how you arrange/manipulate matches.

    New blocks only spawn on rows that actually moved last move. Think about it, you'll see why this is brilliant combined with the preview of what's going to spawn next (1, 2, 3 or 6+) again, this creates a ton of strategy and meaningful choice in how you make moves.

    Threes can spawn things larger than 3. The exact rules aren't really known yet, Asher isn't talking, but there does seem to be some sort of pattern. Personally I think it has to do with combo merges, but yeah. Threes clutters the board with the split 1-2 thing as well as forcing you to deal with higher tiles in edge positions sometimes.

    Everything in the game is a swipe. Everything. It's genius consistency.

    The blocks talk to you. At first this doesn't make a difference until they *taunt* you for not making a move in a while, it's remarkably effective at messing with your consideration and forcing fast moves in reaction, which tend to screw you over more often.
    Thanked by 1raxter
  • Played threes on day one. Played 2048 today. Threes is definitely the better execution. The sheer charm dripping from everything pixel! But I think there's is enough deviation in 2048 to warrant its existence even only as a game design study. That description of 1024 is just pure bullshit though...
  • Just learned that the music for Threes was done by a friend of ours, Jimmy Hinson. Haven't heard it, but Jimmy's great, so the music is probably too. :D
  • Just learned that the music for Threes was done by a friend of ours, Jimmy Hinson. Haven't heard it, but Jimmy's great, so the music is probably too. :D
    The music gets stuck in your head and never leaves... I totally bitched at Jimmy about that ;)
  • So now there's another one called LHC which collides particles to level up to eventually get to a BOSON HIGGS :D

    http://milrivel.github.io/LHC/

    I love it, the ripoff became a meme all on its own :)

    Looking at this, I really wonder - is 2048 - the original ripoff - "better" than Threes for being mechanically simpler? Or more accessible?

    I know that Threes is a "better game", but what did 2048 do right that Threes didn't do, that it got much more traction than Threes?

    And should traction be the goal, or "making a good game" be the goal?
  • edited
    Tuism said:
    I know that Threes is a "better game", but what did 2048 do right that Threes didn't do, that it got much more traction than Threes?

    And should traction be the goal, or "making a good game" be the goal?
    Wat? Did 2048 top the App Store for over a week or something while I wasn't looking? AFAIK 2048 got noticed here because it's a PC game and Threes isn't out on PC, that's it.
  • I'm just saying 2048's become a bit of a meme, it's the outward face of that game mechanic. If one considers Flappy as successful, then one's gotta consider 2048 as successful.

    Not that I think cloning is great, I'm discussing the merits of being super visible and how to achieve it.
  • Tuism said:
    I'm just saying 2048's become a bit of a meme, it's the outward face of that game mechanic. If one considers Flappy as successful, then one's gotta consider 2048 as successful.

    Not that I think cloning is great, I'm discussing the merits of being super visible and how to achieve it.
    What are you basing this "visibility of success" on? I can't MOVE without seeing, hearing, tasting or smelling about Threes at the moment on the app store, in indie chats and web circles.

    Actually, TBH, it was mostly Luftrausers for the last few days, but my point is that the only places I've seen 2048 mentioned are here and that XKCD comic.

  • Well here's a question - do you base the success of something on visibility among its creators or its market? Cos I keep hearing about 2048 and its hordes of clones/forks, which is not among indie circles, sure. Threes I haven't heard so much about, so obviously we are exposed to different circles. Interesting.
  • edited
    2048 is huge in tech circles. Almost every day there is a new flavor or variation up on hacker news. Every single time someone tells me about 2048 I point them to threes and they've never heard of it before. So much so jay is games (I think it was them) actually posted an article bemoaning the "2048 clones" which they were corrected on and they rescinded the article.

    Having sunk a lot of time into threes I'm torn as to which one has better mechanics. The deep game is down to a lot of randomness. Every time I try make a strategic play over several tiles and moves I'll get screwed by a high number random tile spawn leading me rage quit a couple of times. On 2048 I often find chain reactions happen and I'm surprised as to how I'm still alive. Which is neat but reduces my feeling of agency.

    Anyway, I think this points towards a general trend about the tenuousness of ownership in the Internet age and reposting of content encroaching with things like 2048 and the flappy clones. Not sure how I feel about that... But if that's the way the internet goes nothing will stop it. Any game based on simple mechanics or premise will be vunerable. Cadence included.
    Thanked by 1Tuism
  • So, basically, people who don't have iOS devices haven't heard of Threes? And even those with iOS stuff now probably won't hear about it because it's off the featured list?

    I'm starting to seriously wish Asher had put it up on PC too.
  • I have iOS devices... And I hadn't heard of it much. But again, that's just me.

    And yes I was basically talking about visibility and measure of success. Not taking away that Threes was an original and a good game. More wondering "what makes good visibility and is visibility (and memeability) a measure of success"

    Not putting Threes on PC - do you think it was a business decision? If they put it on PC, would it have been a paid download or a browser game like 2048 was? Which would be more successful, do you think? Though of course hindsight is 20/20 and we can't predict a future that never happened.
  • Yeah, I just had friend post on facebook an iPhone screenshot of a good looking 2048 version "that he's getting addicted to". Did a search, there are hundreds of clones for 2048 :( Sad to say, unless there's more coverage on threes I don't think most people will ever get the link...
  • Lol I just played the LHC thing and pressed up right down left continuously... And got to Photon + a Gluon... About the same result as if I played consciously XD
  • Everyone that has ever tried to make their own game should read this right now.

    Read it. It's about game design, trailblazing, getting cloned, getting not-cloned, building things, trying really fucking hard, learning, watching other people learn, unlearning, triumph and failure all at the same time.

    It's Asher talking about how they built Threes, responding to 1024 and 2048 (he says that the Numberwang 2048 is his fav so far) and posting selected emails between him and Greg (Wohlwend, who is amazing) to show what the process of discovering Threes was like. If you don't understand why I think Threes is a better game than the rest, perhaps this will help to explain.

    Also, just fucking look at the Threes website and try not to die of jealousy. Then scroll down. Die of jealousy.
  • Fuck the Threes website. I hate it so much. (I don't really. It's lovely. It just makes me jealous and feel inadequate).
  • edited
    I'm about a third of way through reading those emails. HOLY DEV PORN BATMAN. 100% going to be reading the rest this weekend.

    I can especially relate because I'm neck deep in designing my own puzzle game. Trying to get people to understand things is really fucking hard. I'm swearing lots - but in this case I feel strongly enough that swearing is appropriate. I'm also kinda jealous I don't have someone to hash out design problems with.

    The thing I find most remarkable is that they spent so long barking up the wrong tree, and the that the mechanic that became threes took several months to emerge. That's both reassuring and alarming. Reassurance in that good design take lots of time and iteration - and it's okay that I'm starting on tutorial v4 for cadence. Alarming in that it makes my imposter syndrome kick in hard with the feeling that maybe there's a better game in Cadence I haven't found yet.
  • edited
    I don't think 2048 is a better game. I was asking whether 2048 was a bigger success than Threes, and if we can learn from that success - how to get it, and how/if Threes may have been able to be just as successful, given it was the original one.

    How. Does. Numberwang. Work. (edit: I see it's just the colours XD)

    And thanks for the article, will definitely be reading it when I can :)

    That Threes site is fucking amazing, totally, saw it a while ago and was mesmerised for an hour. The Ridiculous Fishing website was also damn cool like that :)
  • edited
    How. Does. Numberwang. Work.
    That's Numberwang! :D


    [edit: for context for anyone confused by this




    ]
  • Tuism said:
    How. Does. Numberwang. Work. (edit: I see it's just the colours XD)
    Shinty Six.

    Also, I don't think the colours do anything.

  • Dunno if this convo is dead, but I just got this in my email box from Udacity, so I thought I would share it here, seems somewhat relevant to the conversation...

    https://www.udacity.com/course/ud248?utm_source=sendgrid&utm_medium=email&utm_campaign=april-newsletter
  • So having just got a 1536 I think i'm of the opinion that threes is in fact the better game. I find the reason for me getting to this point a curious thing. Reading the threemails (which are totally brilliant and a great investment of your time) led to a curious side effect where I started to feel like I knew the developers. They became real people struggling to make this perfect little game. Moreover hearing them talk about how getting a 768 was a really good effort humanised what "being good at the game is" in a way that leaderboards will never do for me. But reading all the design considerations about how 1536 was a totally batshit achievement lead to me kinda setting it as a goal for myself - sucking me deeper and deeper into the game, and actually changing my perception of enjoyment in the game. Feel like there is definitely something learn in that.

    Also HOLY SHIT I GOT A 1536 IN THREES!!!!
    Thanked by 2Elyaradine raxter
  • http://gabrielecirulli.com/articles/2048-success-and-me

    A blog post demonstrating the cognitive dissonance from the creator of 2048 and how he decided to not make an app version of 2048 (mostly due to the fact it was based on other games) and then eventually caved. He's logic has clearly been through a major post rationalization process, but I do kinda respect the dude a little more. I couldn't even begin to imagine how I would handle that at age 19.

    There is a lot of discussion on hacker news as well (where 2048 first broke). https://news.ycombinator.com/item?id=7704800 It's amazing how hostile some people are towards Threes. And how some oppositely polarized threes sympathisers can come off sounding like obnoxious fan boys in response.

    The whole debacle is a riveting study in ownership and the true value of innovation. Kinda feel like everyone got a raw end of the deal, except the cloners who capitalized on the app store and probably made millions. Fuck those guys.
  • The whole debacle is a riveting study in ownership and the true value of innovation. Kinda feel like everyone got a raw end of the deal, except the cloners who capitalized on the app store and probably made millions. Fuck those guys.
    Yeah. Fuck those guys...

    I find it really hard to sympathise with Cirulli, especially after he's so obviously talked himself into doing something that went against the ethics that were the only positive thing he really had going for him. I mean, if his motivation wasn't greed, then why are there ads and IAP in his mobile 2048? How did he frame this "opportunity" in his mind, if not one that's about money and gains from selling 2048?

    There are many other ways to see what he "went through" (sorry, that doesn't seem like a difficult thing to have happen to you - other people using your work as a conduit to profit off the work of others) as an opportunity: Perhaps it's an opportunity to build a name for yourself as an ethical open source stalwart at age 19; Why not see it as a chance to build a reputation for making good interfaces and being an HTML 5 phenom; Or what about taking the risk and making your own games, using the fact that people will forever say "from the 2048 guy"...

    I just see his eventual output as a complete lack of confidence in being able to create something new further down the line. So instead he talks himself into being just like every other predatory douchebag.
  • Reading this Wired article about the design of Threes vs that of 2048 makes me want to try for a 6144 again...
  • That article is a nice TLDR for those who hasn't read the epic mail post :)

    I feel a bit irked by the way a game needs a bunch of behind-the-scenes for people to consider it good - not that "people aren't smarter", but that... Well, let me put it this way. I've never liked art in the way that a lot of them are just "things" with their pricetag justified by a piece of paper written by some intellectual labeled "rationale". I always thought art that doesn't speak for itself to be BS - if you have to read about what something means instead of being able to extract it from just your interaction with it, then it's failed.

    Of course you can also argue that the rationale is part of the art itself, but I don't subscribe to that.

    Not that I'm saying Threes is bad. I've played it to death (and only gotten to 729 :( ), and I certainly find that it's a better game than 2048, which was a nice distraction that got disproportionate meteoric fame, but if something needs a rationale to be "good", then... Well, I just think it shouldn't.

    I don't have the solution. Maybe I really just think Threes could have been marketed better. Though I see that the epic mail post was the best piece of marketing it had to this point.

    Maybe I'm just over-philosophising here :/
  • I don't think people needed the behind-the-scenes stuff to realise it's good. Many people (including myself) needed it to pay attention and play the game for a bit longer to see that it's good for themselves. If I did not read through that, I would not have thought Threes is better than 2048 (which I happen to play first), and I would have stopped after a few play sessions. Now I am an addict.

    I agree with you (@Tuism) about the general principle about rationale, but I wondered how many other jewels fly under the radar because we dismiss them too quickly.
  • I have never heard of Threes, but I find the simplicity in 2048 really amazing, and reaching 8192 is a great feeling (got there 3 times so far) - but getting to 2048 is no longer a challenge so it needs to be called 8192 now ;)

    Simplicity in game design, simplicity along with addictiveness is something I have always wanted to achieve yet never ever been able to get right. So while I cant comment on Threes, I can say that 2048 is an amazing game.

    (I seem to be playing a clone of a clone of the close of the original threes game..... no adds, no IAP etc - just free addictive fun).
  • Give Threes a go and you'll see why it's actually much more of a game than 2048 is - 2048 plays itself. For a LONG time. You can't just autopilot in Threes, and the mechanics are effectively just as simple - but gives a much better decision space to play in. Try it :)
  • edited
    Hi,

    has anyone ever tried to make a game of 2048?

    I did.

    https://itunes.apple.com/us/app/4096-rush!/id863378211?mt=8

    Im hoping you guys can give me some credible feedback re the mechanics i have instilled in it.(levels,time pressure, limited lives)

    For my next version i have some new items:
    -the random tile is random in size too, up until 3 away from level tile completion
    -there is a tetris style preview tile at the bottom
    -i have created a jelly like mechanic as seen in candy crush, where you have to match the tiles on the jelly to make it disappear

    Let me know what you think as i find feedback from you guys very valuable!

    image

    Edit:
    Regarding the jelly like mechanic, I'm toying with the idea of having it add to your score as opposed to it having a negative consequence like being unable to complete the level…
    any ideas?
    IMG_2366.PNG
    640 x 960 - 76K
  • edited
    @Usul: While I appreciate the challenge and joy that can be had in building your own version of an existing game, trying to earn money off something that's as emotionally charged as the 2048/Threes cloning issue feels pretty icky :( Your whole "forget Threes" thing in your app store description? That just makes me sad...

    I'm totally down with exploring the design space around 2048 and mixing in what casual matching games have done (the "jelly" mechanic isn't even a Candy Crush thing, they stole it from Big Kahuna Reef when they cloned that) and I recommend people do this sort of thing with Tetris and Arkanoid all the time as their introduction to games. It just feels weird to me to ask money for something that is so much someone else's work, especially when that work itself is actually ripped off of someone else without their permission.

    P.S. I tried downloading your app (you asked honest questions that could do with feedback, despite my misgivings), but I'm not on iOS 7 yet for testing reasons.
  • @dislekcia: Thanks for the feedback!
    You are an emotive person and that makes for good feedback. I am testing the market on various aspects which demands that i remove emotion from the equation ;-)
    Im going to check out Big Kahuna Reef thx!

    When you do download my game, and especially my next version, you will find that it veers away from the "rip-off" tag, and its on that basis that i would enjoy some feedback at seem point!

    Thanks and kind regards,
  • edited
    Usul said:
    I am testing the market on various aspects which demands that i remove emotion from the equation ;-)
    I'm not sure what that means.

    As a developer who's been cloned, I really can't sit idly by and watch someone else do it with a clear conscience. You may indeed not realise what it is that you're doing and simply be trying to learn more about game design, which is a great goal - just don't go about it in ways that are ethically worrying.

    If you're really curious about game design in the Threes space, you should definitely read the Threes letters that Asher and Greg posted, detailing the thinking they put into the game.
  • I spent the day at HP World yesterday, so between sessions I bought Threes and started playing it. Now remember I first played 2048 before going to threes.

    Threes:

    Good
    Unique design - simple and clean
    Shows the next tile to arrive, and a semi meaningful system on how the new tile arrives

    Bad
    Have the designers never played any other Android puzzle games? So many evils in their design.
    - How can a simple game like that take so long to start?
    - Whats with the register with Google+ that it asked for for at least 3 times at the start? I'm on andorid, I paid for it, I have a google id!
    - This is an abstract puzzle game, whats with the popups "Meet Mr Triad!" - I want to play a game not be interupted.
    - Whats with the 4 swipes to finish a game? Why does the back arrow not close the game?
    Random value of the 'large' tiles that spawn.

    What I liked was that you could plan for the new tiles coming in, what I did not like is the randomness of the tiles coming in. If all tiles were 1,2,3 then you could plan for it. I had a lineup of 5 3's ready to be built up and it gives me a 12, or a 24. I found that very frustrating. In the end you cannot plan enough in Threes to build a system that can not suddently be wreaked by the randomness of the tiles thrown at you. [Highest tile I got to was 384 - the equivilent of a 512 in 2048]

    2048 (the version I have downloaded)

    Good
    Siimple, clean, well designed Android interface
    No fluff in the game, it fits perfectly into the Android puzzle design structure

    Bad
    Too much randomness in where the new tile appears.
    2048 should not be the goal, I can reach 2048 in 95% of the games I play, 8192 is a much more worthy goal, as I am usually bad at games and can get to 8192 maybe that is even too low to be a real challenge. (But it is math based, so maybe my logical brain works better in a game like this). [Anyone remember tha game.dev LAN where everyone beat my highscore on my own game but about 5 times?]

    What I like about 2048 is that you are truly in control of your own destiny. Follow the 'rules' of playing the game and you can do well, it just does not feel as random as it looks. People that say do left, up, left, up continually will reach a decent score, but that is not enough to 'win' the game, not enough to get 8192 and higher.

    Personally I think the interface design of Threes absolutly destroys the good in the game. Having read some of the design thread for the game, i can see where they tried to give it a theme, and where they did not remove sections of the theme from their design. In the end, Threes was very dissapointing having read this thread.



  • Threes was built for iOS specifically. The slow initial startup is less of a problem there because people generally just switch between apps anyway, so Threes is usually just ready and waiting for you to play most of the time. Or if you're weird like me and force apps to shut down ;) But yeah, I know what you mean there.

    I'm not sure if the Android version has been updated to match the iOS updates... Stuff like the multiple swiping to end a round are gone now and tiles that are going to be higher than 3 have a different preview icon. The new number character popups only appear when you first see that specific number and they're pretty charming. The voices and music for Threes make it stand out for me, were you playing without sound?
  • Great to hear the small improvements are being made ;)

    I never play games with music. I seriously dislike music and hate games even more if I cannot turn it off. I however cannot remember turning music off for Threes.
Sign In or Register to comment.