Most important things in a TGM or Tetris game?

Thread in 'Discussion' started by Kasumi, 5 Jul 2011.

  1. My seemingly endless quest to create a Tetris game that clones ARS somewhat, but not any specific modes brings me to these questions:

    What are some things that would really draw you to a new Tetris game?
    What are some things that would make you never want to touch a specific Tetris game?
    What things are most important to you in a "TGM" game?
    What things could you let go if you absolutely had to? (i.e. the guideline forces a lot of changes, and you can only fight against some of it, not all of it)
    Things you barely care about?

    Feel free to answer just a few, or discuss or whatever.

    So here are my answers to get started:

    1. A lack of infinite spin.
    2. Flat Side Up spawn orientations
    3. Any kind of 20G. If I had not been introduced to it, I would not be playing Tetris at all today. Bonus points if I can choose to start at 20G through some option. Extra bonus points if there's a mode that starts at a slow 20G and increases speed.
    4. 20G Multiplayer. It has its own number because it's worth more than just bonus points.
    5. ARS. Its wallkicks, orientations and colors. (Not including firmdrop/locking soft drop behavior. I could live without that) Bonus points if it's the only thing available. Another Sega rotation derivative might be fine as well.
    6. A Doubles mode.

    1. Slow shift rate after DAS kicks in. Game Boy Tetris (1989) is especially horrid. Tetris DS is really, really bad. Tetris DX is fast enough.
    2. Infinite spin. I really hate it. Most of the time, I have to touch Tetris games with it anyway... :facepalm:
    3. A lack of wall kicks. New games have no excuse for this.
    4. No lock delay.
    5. Only one rotate direction. Sorry Sega's Tetris :(
    (in no particular order)
    1. Flat Side Up spawn orientations
    2. 60hz DAS
    3. ARE+IRS
    4. Step Reset+lock delay
    5. Modes that end, with at least one fixed at an increasing 20G speed.
    6. The grading system? I realize it's a huge part of TGM, but that's never what's kept me in them.
    7. ARS wallkicks.
    In order of what I'd give up first to what I'd fight for:
    1. Modes that end/Grading System.
    2. Flat Side Up Spawn Orienations
    3. Step Reset
    4.ARE
    5. Standard ARS
    6. 60hz DAS+a 20G mode+IRS. I'd probably give up 20G before fast DAS, since 20G is sorta pointless without it anyway. Even though move reset would make it playable, it would just feel clumsy and slow like it does on Tetris DS.

    1. Hold piece, I suppose.
    2. Floor kicks.
    3. Enter above Ceiling or not.
    4. Soft drop speed, as long as its faster than the the falling speed assuming the falling speed is <1G
    5. Top out behavior

    Things related to my specific game:

    I want it to be a way for SRS players/new players to get into TGM style gameplay without the difficulty scaring them away. Have some easy modes that reach really, really slow 20G (maybe 60 frame lock delay?), not have a million options etc.

    I've also been battling with how to handle hard/firm/locking soft/nonlocking soft drop controls. I have always hated TGM's required up AND down to lock if a piece isn't already touching the stack. Possibly losing fractions of a second on every single piece plus the extra motion... :twisted: It's part of why I never play Master modes. I also think if one plays well, a soft drop is needed very rarely.

    Having a harddrop lends itself well to having a non locking soft drop, but lots of people expect soft drop to lock.

    My preferred setup would be Up: Sonic drop, Down: Harddrop like blockbox multiplayer. But then people might miss soft drop.

    Even though I want as few options as possible I'll probably break down and allow all four to be used in any game mode, with the default being Up: Hard Drop, Down: non locking soft drop, since Tetris DS seems to get away with it.

    Would changing small parts of ARS throw you off? Things like fixing Mihara's Conspiracy, making floor kicks possible when a piece is not grounded, and wall kicks possible when the I has no part of it adjacent to a wall in the vertical state?

    Fixing Mihara's Conspiracy would probably throw me off, but it being solved just makes more sense to me and probably to new players. The other changes wouldn't even bother me and would make more sense to me.

    And... could you envision yourself enjoying a game that was otherwise very much like a TGM game if ARS/step reset wasn't even an option anymore? I used to be against the idea, but... these days I wouldn't mind it. I like 20G in SRS flavor too if it has ARE, IRS and fast DAS. And move reset/SRS kicks would allow for ARE to drop to 0 and still be playable.

    Well that's all. Hopefully that's not to much to answer.
     
  2. Muf

    Muf

    Awesome centralised cheat-proof ranked ARS online versus
    Online doubles that doesn't suck

    Move reset
    Non-TGM randomiser (bag, memoryless, piece bias, SZ starts, etc)
    Lag or framerate issues

    ARS
    Step reset
    Proper colours
    High speed
    GRS

    Hold
    Floorkicks
    Multiple previews
    Proper colours

    Guideline, legality, official Henk Rogers approval, SRS players, etc.
     
  3. Altimor

    Altimor a.k.a. Ghett0

    1. What are some things that would really draw you to a new Tetris game?
    ARS netplay that doesn't devour my input

    What are some things that would make you never want to touch a specific Tetris game?
    Move reset, sluggish DAS, the game not being able to kick my ass or give me a hard time winning.

    What things are most important to you in a "TGM" game?
    ARS, IRS, IHS, quick DAS, grades for e-penis

    What things could you let go if you absolutely had to?
    nothing

    Things you barely care about?
    top out behavior
     
  4. Zaphod77

    Zaphod77 Resident Misinformer

    1) challenge, combined with fairness. This is the essence of any good modern tetris game. The TGM series does a very good job of that
    2) good play control, and frame rate. 60Hz action, and solid control are a must.
    3) New modes. A new tetris game MUST bring something to the table that wasn't already there. WHy should I get THIS game instead of playing the one I have?
    1) Infinity. I've already got enough games with that.
    2) broken scoring (15 level marathon, I'm looking at YOU! i refuse to pay for a game that has that in it.)
    3) bad control
    4) lag
    5) Fake Difficulty (a very good example of it is in master mode of tetris party deluxe, where the lock delay after reset is greater then the delay before moving or rotating. The SOLE purpose of this is to cause misdrops when you think you are reacting fast enough.)
    In no particular order

    at least 2 CCW and 1 CW rotation button
    step reset
    ARS
    IRS
    sega colors
    IHS, if hold is present
    GRS
    tgm style levelling
    20G mode
    Death mode (or equivalent)
    pieces spawning flat side up

    In other word,s pretty much everything. :)

    Colors, though i'd hate it
    SPawn orientation, though it would take getting used to.
    spawning inside the field (yeah it makes it a bit easier. won't make much difference, though)
     
  5. Thanks for the answers, guys.

    Muf: Yeah. I hadn't even thought about the randomizer/SZ starts, but that's important to me as well.

    Zaphod:
    What's your complaint there? Too much reward makes maxing out the score happen too easily?

    I also have at least two modes that haven't been seen before.
     
  6. Edo

    Edo a.k.a. FSY

    Kasumi, you mean to say that you've never heard how guideline scoring works? Watch this and all will be revealed: http://www.youtube.com/watch?v=ISANNTEf3J0
    Incidentally, DumbledorsArmy is actually a pretty good all-round player and a decent guy, so even though the strategy he's using is broken as fuck, don't take it out on him!
     
  7. Zaphod77

    Zaphod77 Resident Misinformer

    My complaint is that you get maximum points for doing repeated single single combos, and finishing each level with a TST.

    The goal system is stupid. The intent of it seems to be that each single line worth of points (for the section) is one goal. But it rounds down, so getting two singles in a row, followed by breaking the combo gives you 2.5 goal, which rounds down to 2 goal, squeezing in many more points.

    Then, overkill on the goal doesn't carry over to the next round, so you finish with a single high scoring clear when you have one goal left (tack it onto the end of a combo if you can).

    Yes, it really is that dumb.

    The other thing is that if you soft drop a piece, then hold, you can then hard or soft drop the piece and double up on drop points.

    Here's how I'd fix Guideline marathon to play as presumably intended.

    1) track fractional goal
    2) make it so extra goal carries over
    3) Only award the soft drop points after the piece locks, so they will go away when you put the piece in hold.

    After that, the only thing that won't increase goal is soft and hard dropping, so the faster player will score more.

    But wait, that's still broken, because then it will still be to the players best interest to score only singles so they get more soft drop points.

    Simpler fix. Points are meaningless, and make it a time attack mode, but use the existing goal system.
     
  8. TOJ Challenge is pretty neat: enforce a time limit, use variable Goal, award time extensions on level up, and go all the way to 20G. I think there is an ending level, but I can't remember if there is a time remaining bonus or not.
     
  9. That's exactly what I'm saying. I don't play guideline games often, and I play them for score even less.

    I'm with you both that that's a stupid system, I was just unaware that's how it worked. Thanks for explaining it.

    Edit: Is there a point to this bloated system besides making maxout possible in a mode that ends? Would anyone have a problem with classic style endless where each single is worth X, double is worth Y, score is (scoreforelinescleared*level) with possibly some reward for combos? I've never agreed with rewarding soft drop since it encourages playing really low in the well, and I can't place pieces as well low because of the distance from the spawn chamber. Selfish, I know.
     
  10. Zaphod77

    Zaphod77 Resident Misinformer

    Standard 10 lines per level marathon ending when you clear level 20 is a better system than the current guideline standard. Simple, effective, and rewards higher scoring clears more than soft dropping.

    And endless is perfectly fine too. For a great example of what can be done with endless under the Guideline, check out Tetris: New Century if you can get ahold of it.
     
    • linux support
    • grade system
    • achievements (with the condition that they aren't hidden or dumb like Cultris's, example being like "manlock every piece from 0-100". I guess I'm thinking of something analogous to TGM's current medals system, but with it permanently recording your best medals to your account)
    • some kind of interesting unique modifications to the flow of the game, like how Shirase introduces rising garbage for a bit, then ancestor blocks, big roll, etc
    • linux support, repeated for emphasis
    • a cool 20G mode like Death or Shirase to go with the traditional survival mode

    • single rotate
    • inability to remap controls
    • slow auto-shift after DAS
    • lack of linux support (I have had almost zero success playing Windows-only games with wine and virtualbox)
    • a rotation system that isn't floor-aligned, if I'm going to have to use it in high gravity

    the guideline

    learning to play ARS when all you've played is SRS is an extremely frustrating experience, and would be with pretty much any lock delay you can imagine: if you make one mistake, you're liable to cascade fail a bunch of pieces afterwards while you figure out what happened and how you can fix it. this isn't as big a problem in SRS games due to infinite spin/move reset. for this reason, I think the best way to help people learn the basics is to have infinite-ARE puzzles, like getting a bravo with ARS and some specific field/piece sequence.

    i think the way to go here would be to just have the different drops all available as buttons in the config, and let the player remap their keys however they like, leaving drops they don't care about unassigned. perhaps someone will find a way to play super-efficiently with the 4 drops all on different keys.

    if your intent is to make a new Tetris game with only a TGM kind of feel, there's no need to limit yourself to pure ARS. fixing the conspiracy, and perhaps making it more symmetrical (imagine how SRS deals with the S and Z pieces) are both worth considering. I believe DTET handles the conspiracy by changing the first kick-test direction according to which way you rotate, which could be useful to think about.

    personally, I think step reset is too strict, though that's not to say I like move reset either. I had an idea a while ago of a sort of in-between reset where the lock delay resets on step, on the first rotate on each step, and each tap until you tap in the opposite direction. in theory, this would let you stall across platforms in the same way you can currently stall down pyramids, without being too easily abusable.

    while we're thinking about creating a new Tetris game, I was thinking about making a TGM-feel game a while ago and half-assed some ideas for it. since I'm probably never going to go and code it, here are some of them:
    • instead of ARE, have a pause before gravity takes effect during which you can move/rotate the piece
    • instead of a hold slot, maybe a skip piece button usable every x lines cleared or what have you, or that costs grade points
    • no step reset except when this causes a piece to freefloat (this would be kind of the opposite of the step/move-reset compromise I mentioned above, where you can't stall down pyramids like you can now)
    • synchro if a direction is held during rotation, even before DAS kicks in
    • for the Master mode, have the randomizer/preview count/etc change for each level section (for example, 0-100 might be bag, 100-200 might be expdist, 200-300 might be hist-4 6-roll, getting progressively more random until the 900-999 section which is memoryless)

    note, I'm not saying a game -should- have any of the above, they're just ideas I had that you may like
     
    • Like Zaphod77, I'd like a challenging and fair game. I'd like to die knowing that I suck and that a better solution could be found, not because the game decide to crush me.
    • A sense of progression (achievement, rankings, statistics, "have you noticed ? three months ago, you were at that level, now you're there ! congratulations !"). Mastering Tetris (any style) is a long way to go; it's a good thing to know I play better from time to time.
    • Single rotation button
    • Micro-transaction for gameplay elements
    • Move reset/infinity
    • Lack of predictability regarding the piece placement (I don't want to mash buttons to get a piece where I want, I want to place it with a precise sequence of input)

    It's coherent game design. It had four or five iterations (Sega Tetris, Shimizu Tetris, TGM1, TGM2, TGM3) of the same gameplay, all of them played to death and feedback were given to the developers, so very few things feels broken.

    If you want me to enumerate specifics gameplay elements, well, that hard ^^.
    Sonic drop, what Zaphpod said

    Maybe difficulty shouldn't be something to be scared of. Super Meat Boy is both popular and quite difficult. There should be a way to make a top out less costly (lives ? restrictive Torikan and an indication of how long you still need to shave off ?)

    I also like Blockbox control scheme (i.e. up sonic drop, down hard drop).

    One alternative I imagined (although the original idea was colour_thief's (or Jago's ?)) was the use of an "accelerator" button that would replace or be added next to the hold button (D).

    Up and down are non-locking. Left and right are non-DAS.
    Press D and up/down become locking and left-right are DASed without charge.
    (if you still want the hold functionality, I suggest either that double press or that D+C (second rotation) will hold )

    I think you're describing Tetris Kiwamemichi (TTC answer to TAP at the time, I think). I could only play it for 5 minutes or so, so I cannot say if it was rubbish or not.
     
  11. Zaphod77

    Zaphod77 Resident Misinformer

    post entry delay.

    The following games i know of have it. This is not an exhaustive list. :)

    1) GB tetris, for every piece, every level, but it's really only small enough to get in a couple of taps.
    2) NES tetris, for the first piece only. Significantly longer than the GB version.
    3) Tetris Kiwamemichi, for all pieces, even under 20g, and the delay is fairly long. This, combined with 1G auto shift, significantly lowers the difficulty of the game.

    As Guideline games go, Kiwamemichi isn't THAT bad. The infinity can be turned off, and it does have a 4 player simultaneous 40x20 wide well co-op, the control is solid, and the DAS is at least as good as New Century (one of the other decent Guideline games).
     
  12. Tetris & Dr. Mario also had gravity delay, and it made Level 29+ playable... This was instead of the standard ARE that NES Tetris (its original version, in a sense) had.
     
  13. It's not instead of; it's in addition to. T&Dr.M has longer ARE/Clear than NES.

    Add Atari/Tengen to the list of games with pre-gravity delay. Also, I don't remember GB having it at all.
     
  14. I am long winded. I am sure you all knew this.

    DrPete's Post

    SRS and ARS are essentially equivalent to me in 0G, and slow 20G. But I guess that's because I started with ARS. All the smart placements that you can do in ARS (except some LJ kicks that come to mind), SRS can do. So I place pieces in SRS where I would in ARS, and the game feels the same.

    (If I can avoid it) I absolutely won't allow people to continue to rely on SRS climbing bailing them out. In SRS, you can do much of what you could in 0G in 20G. It destroys the limited placement options that make 20G what it is.

    I will make it so while you're training you'll never top out, the lock delay is infinite, good placements will be shown. But if I can avoid it, I won't support climbing 20G. There will be slower gravity modes as well, however.

    You're right about this, but I will probably do it anyway. S and Z pieces are especially strange to me in SRS, and the I piece is only slightly less strange. I like the consistency of the asymmetry. If an I block is horizontal, and I press rotate right, which of its vertical states do I get? In ARS, it's always the same vertical state. In SRS, it could be either of the two. (Note, I'm referring to a horizontal state in the middle of a game. From spawn, it would always be the same in both) I think ARS is less confusing for people who say... don't use both rotate buttons, or rotate a lot while they're thinking and the piece is falling. In ARS they can count on one consistent action from the visual state of a piece. To me, that's success in design.

    SRS actually has some very good design choices as well. Symmetrical rotation and flat side down spawn orientations are actually excellent design choices for a different sort of Tetris game than what I want to create.

    DTET did do some really cool things, and is a good bridge between SRS and ARS. I'll think about it, but ARS has sort of an old school appeal. :wub:

    ARS is also VERY consistent, and well designed. When I was programming ARS into my game, I was actually looking into the aspects of it rather than blindly following what the wiki said. Like why L and J won't rotate when certain block combinations surround them. They thought of so much, and solved so many problems. I doubt I could do better.

    SRS... is... very much the opposite. It seems what ruled its design was to make rotating a block possible in nearly EVERY case. So the kick rules are inconsistent even in fairly similar (but not exactly the same) cases.

    Part of this is perhaps not specifically SRS' fault. Older versions of Tetris didn't rotate in a way where the lowest block is always at the same vertical position. This meant that if one wanted to keep blocks rotating this way visually without breaking in games where you can move the piece along the ground, you had to write some crazy kick rules. An SRS rotation on a flat surface can send that piece into the ground. So floor kicks are required. This is barely true in ARS.

    In fact, the exception is the I piece. And I think this is an old consequence that comes from Sega's Tetris where it was actually a better choice to have it like that.

    In Sega's Tetris, you couldn't rotate a piece, such that some of the blocks would be above the top of the well. So if an I piece had its vertical state have the same vertical position as its horizontal state, it would have been impossible for the piece to rotate until it fell three whole cells from spawn. With what they chose, it only has to fall one cell like every over piece in the game. (Consistency! It's important in design. ;) )

    In modern Tetris games, this isn't so, and it's something I could "fix" so that I wouldn't even need to floor kick, but I like the challenge it presents.

    (One day, I'll write a long article about what I think was the actual thought process behind lots of different aspects of Tetris behavior. That day is not today :$)

    I believe Tepples did some good work on unbreaking SRS in his TOD rotation system, and others have tried as well.

    But I'm not in this to add broken systems just because they're popular. I'm not in this to add design choices I don't understand. I won't blindly copy a TGM torikan for example, because I don't know how they got the number.

    I like extreme consistency, I like not overwhelming people with options. If I can create a rotation system like SRS that doesn't destroy all the other systems the game has in place I will, but I will not write special game rules for it just because it's there. SRS breaks step reset and 20G. To some, that would be a reason to not have step reset/20G. For me, that's a reason to not have SRS.

    I didn't mean to go all game design geek on you, but my life is kind of ruled by it. So ends my "meanie" response.
    Let's say I'm a new player. I move a piece right a lot, and the piece doesn't seem to lock. Then I move it left, and it locks. I would be absolutely puzzled. Step reset makes sense to me from a design standpoint, and it's not strict by itself. With a long lock delay, it's not needed at all, and it just makes a short lock delay LESS strict.

    Assuming rotation and movement are possible: It makes 20G like 0G. Assuming just rotation is possible: It allows one to rotate pieces to the state one could not get to with IRS alone. I may not have a problem with this. I'll think about it.

    I've seen this under lots of different names, in different games though.

    I may have seemed dismissive and mean, but you have brought a lot of really great ideas and things for me to think about to the table.

    PetitPrince's Post

    I've always found it super strange that I like Tetris The Grand Master. Because I don't like games like Super Meat Boy. Or I Wanna Be The Guy. I agree that difficulty shouldn't be scary, but lots of people look at TGM like, "Wow. I could never do that." I want to take that feeling away with my game.
    It sounds like you may not have seen my Endless Tetris video: http://dl.dropbox.com/u/17232562/Endless Tetris Sample.avi (10 megabytes, it's worth watching)

    No need for lives. You'll never top out. I think Henk Rogers once said he introduced infinity because it was frustrating to lose from one small mistake. I believe there are other ways to solve that problem. If people want to play Tetris Endlessly, they should be able to. Endless would just be for training. I also plan to have extremely slow 20G modes to get used to it. Torikans that actually reveal how much faster you have to be in the main game modes are a good idea too.

    Yeah, I read that. As way to cancel ARE, perhaps. I may do that, I've definitely seen that suggestion around here.

    That sounds complicated. Initial thought makes me think context sensitive buttons are a bad idea, but I can see this working. Might be complicated for new players. I guess it's not any harder than the run button in Mario games, though. Serves a totally similar purpose too.

    About Grading

    Here's the deal: Those who are already seasoned Grand Masters looking for challenge in a game from me will be disappointed.

    My game's Master level 900 equivalent may get as fast as Shirase 300. This means, it will be faster than TGM/TAP's Master Mode, but probably easier than both because it will have hold.

    As for grading, I'll probably only do a little more than TGM1. Score is the main thing, slow times in certain places disqualify you from highest rank.

    Much like I am unwilling to implement rules I don't understand, I'm not gonna create a mode I can't beat myself. I can play perfectly and get a bunch of Tetrises from Shirase 0-299, then 300 kicks in and I fall apart. Same with Death. I might make another topic so you guys can tell me what's wrong with my playstyle. I've been playing TGM for 3 years and it still seems completely unattainable. Even in my endless game where you never top out, I just keep going up, and up and up at those speeds.

    Other stuff

    I still would like some opinions on fixing Mihara's conspiracy, and changing the I piece's floor kick and wall kick behavior. If people want either to stay, could I get an explanation of what problems the current behavior solves?

    Specifically why you can't wall kick the I to the right when none of its blocks are adjacent to other blocks, and why you can't floor kick an I unless it's grounded. I've tried hard to come up with an explanation and I can't. If no one else can, I'm changing it.

    Edit: For the curious. Yes, this post really is five pages in Microsoft Word.
     
    Last edited: 7 Jul 2011
  15. Zaphod77

    Zaphod77 Resident Misinformer

    The reason you can't floorkick an i if it's not touching ground is to prevent you from stalling in sub 20g play. To prevent you from getting infinite step resets, rotation after a floorkick causes lock delay to become zero for that piece.

    If you remove the I floorkick completely, then allowing a kick left when rotating to vertical would help a bit.
     
  16. I'm not quite sure I understand that, though. If it locks instantly the next time it touches the ground after rotating in a way that would allow it to fall, you can't repeatedly floor kick whether the piece was grounded on the floor kick or not because the lock delay stays 0 whether the piece falls to new rows or not. Could you show the problem this solves with a fumen?

    Even with what most clones do and what I'll probably do(floor kick only works once, but it doesn't change lock delay) it only allows for a single stall.

    The I kicking left suggestion would make too many bizarre situations happen in my eyes. http://harddrop.com/fumen/?m110@MeYigbYimbYigbYimbYigbYimbYigbYitbRSBCeHYU?cJMPGAIAcTAT5AAA There's a reason the I doesn't kick when rotating to its vertical state for sure.
     
    Last edited: 7 Jul 2011
  17. Zaphod77

    Zaphod77 Resident Misinformer

    You rotate I to vertical. it kicks up. as long as you don't rotate the piece again, it behaves normally. but if you rotate it even once after the kick, then it instantly sticks to the stack to prevent you from getting another floorkick. This is what prevents you from doing two floorkicks. there isn't actually any limit coded in TI. I proved this by getting a T to floorkick twice.



    As for the I bit...


    Understand? :)
     
    Last edited: 7 Jul 2011
  18. I read the post where you first made the discovery. I understand how it works. My question was specifically about I's grounded state limitation. Since it's not true for the T. Your first fumen even proves that.

    So look at this fumen. This is what happens in actual Ti (just checked):



    It fell, but still locked instantly. This means lock delay is 0 even though the piece fell. So what you describe would not happen. Because of the frame processing order is gravity after rotation movement this would happen in your fumen:



    So I swear I'm not trying to be obtuse, but I still don't see the problem. It's no worse than the T case.
     

Share This Page