[Split] Step reset and entry reset

Thread in 'Discussion' started by DIGITAL, 23 Aug 2008.

  1. It's not really though. Not under Shirase conditions where the ARS players are being pushed pretty much constantly to their limits while the SRS players have the option for extra thinking time if they so wish.
     
  2. You guys are crazy...

    DIGITAL seems to be missing my point completely. I said the word "intuitive" exactly 0 times, whereas he used it no less than 7. None of this stuff is truly intuitive, though all of it can become second nature with enough practice. caffeine doesn't get me much better either... I was talking about how entry reset goes against the general design principles of TGM, however this was just in passing comment. My main point is an expression of disgust that you guys could possibly see it as a step forward.

    What I'm trying to say is that entry reset is not better. In no way is it a step forward. You say that step reset is bad because it permits sloppy finesse when placing pieces at the sides. Well, with entry reset, you get a similar problem. Assuming that entry reset is large enough to allow you to efficiently place pieces on the sides, which is a very reasonable assumption, then the game will tolerate sloppy finesse when placing pieces in the middle. So you don't actually gain anything with entry reset, you simply move the "problem" around. I say problem in quotes because, quite frankly, when step reset lock delay is low enough it's just not possible to milk it to slow the game down. And, hey, torikans do a damn fine job of stopping lock reset abuse otherwise. If you want to enforce tight finesse, then entry reset is a step in the wrong direction. Move reset has it's own set of problems too. Step reset is the least abuseable (at high speeds).

    That's why you have nothing to gain with entry reset... But you definitely have something to lose. If you consider 0 < G < 20, entry reset is totally broken. This is a crucial part of the game for developing players. Even with larger entry reset of like 5 seconds, you'll have pieces "randomly" hard dropping into place as players consider the strategic implications of various moves. Not to mention gravity will drop the pieces to the ground, in place, and then they'll just sit there with a very confused novice player wondering how the game is too stupid to realise he wants to place the piece there (or that it's physically impossible at this point to place the piece anywhere else). By contrast, step reset lock delay is the natural extension of gravity.

    But the biggest reason of all that I think you guys are crazy, is that there's no meaningful room for innovation in the lock reset mechanism. You guys seem to expect lock reset to be a driving force to push the player all the way to perfection. The reality of the situation is that it only pushes the player 75% of the way, tops, and that a desire to time attack takes the player the rest. There's simply no need to push the player more tightly, and honestly just about any reset mechanism, with low enough values, will push the player enough to get them playing the real game. DIGITAL and caffeine are all moistening their panties with the thought of entry reset, but there isn't even any point in changing the system.
     
  3. Edo

    Edo a.k.a. FSY

    Sure, it might sound more predictable in theory - "you have exactly x frames of lock delay" - but in practice, this doesn't matter.

    Quick quiz: A falling tetromino hits the ground. How long does the player have to move it into position? (Hint for the step reset people: lock delay is 17 frames). Entry reset people: place your bets now! Oh wait, it's not fair you say? You need to know what happened previously you say? Supposing this is the move in question. Obviously, it's quite a complicated move to pull off, so for this move to even be possible with entry reset, you'd need at least 24 frames or so. But still, how can you tell exactly how many frames you used to do the first part of the manoeuvre? All you can do is estimate, which means you have to do the remaining part of the manoeuvre with an unknown amount of remaining lock delay.

    Another point: I would be very against any lock delay system which made this move identical to this move in terms of ease of performance. Judging from how well ACE-ARS (specifically ACE-ARS 2) was received by fans, I'm pretty sure both move reset and entry reset would be an unwelcome dumbing down of the TGM series.
     
  4. Just an FYI edo that your examples don't apply with TGM3 wallkicks.
     
  5. tepples

    tepples Lockjaw developer

    "Rhythm" means that a falling piece will hard drop when a counter expires. "Entry reset", at least as implemented in Lockjaw Engine, doesn't mean that. It means that the lock delay counter will decrement while a piece is in landed state, and only a new piece can reset this counter. (In >10G, this is pretty much equivalent to rhythm.) For instance, Tetris DS multiplayer appears to use a form of "Finity" that involves both entry reset and move reset at once. While a piece is in landed state, both lock delay counters go down. Moves reset the ordinary lock delay counter, while only a new piece resets the other. When either counter reaches 0, the piece locks. It's similar in effect to Ti World's limited move reset, except it counts frames and not moves.

    Or how about "step recover", which is what TOD uses since milestone 3? Whenever lock delay is not full, the piece falls at least 1G, and lock delay recovers by the inverse of gravity (e.g. 4 frames in 1/4G).

    Garbage enters during ARE. If the player can prevent lockdown, such as by abusing floor kicks like CPU opponent "Nachanca" from The New Tetris, the player can prevent ARE. This exploit, which works in SRS without step recover, entry reset, or Ti Classic-style floor kick limits, is hauntingly similar to Tetris DX's exploit:
    Code:
    |     |  |     |  |     |  |     |
    | T    | -> |     | -> |T    | -> | T    |
    |TTT   |  | T    |  |TT    |  |TTT   |
    |     |  |TTT   |  |T    |  |     |
    `----------'  `----------'  `----------'  `----------'
     Start     Gravity   Rotate right  Rotate left
    
    Such situations are part of why I developed step recover for TOD M3 before I learned here how lock delay actually works in Tetris products.
     
  6. First off, my panties only moisten for you, thiefy. You can abduct my colors any day. [​IMG]

    Step reset is a great system. I'm satisfied with playing step reset, really. I'm not saying get rid of step and replace with entry, and you're right that there's probably no point in changing the system. TC is all about different ways of approaching Tetris, and there's nothing wrong that. That's why there's no problem with me saying entry reset is a cleaner and more logical system.

    On 1G 7DAS, it's possible to move a piece to the sides in under ten frames. I don't know what you mean by efficiently, since all you do is DAS the piece to the side. It's important to note that this isn't about needing enough entry reset delay to get to the sides. What is important is how many frames should you need to place a piece anywhere. For fancy maneuvers that take many moves and rotations, wasn't it you who said that one should abandon these placements in the fast pace of death mode? Also something important: you say that entry reset encourages sloppy finesse in the middle, but there's no such thing as finesse in the middle (think about it). Entry reset does not "move the problem around." It just fixes it. Entry reset is a step in the right direction for tightening finesse.

    Does "0 < G < 20" mean the gravity between zero and twenty? If so, why would pieces randomly hard drop in to place? Do you mean once the piece has landed already? Why would the novice be confused by gravity dropping pieces to the ground? How is step reset a natural extension of gravity in this sense? Could you please explain this better?

    I agree totally. There's no reason to push the player more tightly. What's good about entry reset is that it pushes the player to play more consistently. There's no reason, with fast DAS, a move on the other side of the screen should need more time than a move in the middle when both require the same amount of input. It's unwarranted.
     
  7. jujube

    jujube Unregistered

    that sounds fair enough to me. so whether the player allows the piece to fall slowly or firm drops, they'll have the same amount of time to place it once it lands. this would seem more than fair to a novice player whose only "tetris" experience is with a clone that features lock delay = rate of gravity.

    [​IMG]
     
  8. DIGITAL

    DIGITAL Unregistered

    The word visceral implies intuition.

    With the problem you stated, step reset in no way fixes it. If you notice the trend in Ti, players stack very evenly given hold, 3 previews, and the I floorkick. What's the purpose of step reset there? It's basically crippled and acts VERY much like entry reset. But wait, it'll reset lock delay if a piece falls a row? How arbitrary and random would that be? Do you have to design your stack as a pyramid to abuse it? That's why I believe entry reset IS a step forward because it takes away this unnessary bloat.

    Now that you mentioned the uneven distribution of finesse flex among the columns though, a solution would be to reset lock delay every column the tetromino moves outward from it's default entry orientation. Delay cannot be reset if you move towards the default entry position nor can it be reset by crossing the default entry position to the opposite side. But like caffeine just mentioned, there's not much finessing the closer you get to the middle so is there truly a need for this system?

    Yes, I do expect lock delay to be a dominant driving force but I do not expect it to push the player to perfection. Don't where you got that from. The main point as caffeine pointed out is to push consistency.
     
  9. DIGITAL

    DIGITAL Unregistered

    In practice, you don't count the frames. You time the rhythm and entry reset is always consistent. As for complicated maneuvers, that's why you don't perform them. I said this in another thread but I'll say it here again. Not all 20G finesse is optimal. Compensation with smart stacking and problem solving is the key here. That's how I see all the top Ti players play.
     
  10. jujube

    jujube Unregistered

    i gotta agree with DIGITAL and caffeine here. entry reset really forces the player to stack in a way that doesn't allow for lock delay abuse. there's nothing really wrong with step reset, and it's less forgiving than move reset of course, but like DIGITAL said, if you're stacking flat (which leads to more tetrises) then step reset isn't much different from entry reset anyway.

    i'd also like to point out that in Lockjaw using hold does not reset lock delay when playing with entry reset. this is a good thing because it discourages abuse and encourages initial hold.
     
  11. jujube

    jujube Unregistered

    8 people voted for step reset + SRS...

    would 1 of them like to modify Heboris to make it work correctly?
     
  12. Ah, we're more on the same page then. But DIGITAL was definitely talking about it as if it were objectively better.

    Right, I think slow placements should be avoided. But placements in the middle will always be faster than placements on the sides, and you have no choice but to fill the entire screen. Finesse with such simple placements boils down to how fast you lock the piece. All this is really obvious if you're TAS'ing the game, but the effect is still there for realtime human play. Now with 10G DAS you'd have a point, but I'm talking about 1G DAS.

    Err, I guess I didn't explain myself very well but I was interpreting entry reset to be like lockjaw's rhythm mode, where a piece has x frames until it's locked.

    Ah, but consistency isn't desireable because not all piece placements are created equal. If it takes you the same amount of time to move a piece to the wall and the rotate and then lock, as it does to simply lock a piece dead center, then you are wasting frames with sloppy placement with the center pieces.
     
  13. Edo

    Edo a.k.a. FSY

    As I'm predominantly a TAP player, I gave those as the most well-known and easy to understand examples, but there are plenty of other cases involving other pieces that are still applicable to Ti.
    Without I floor-kicks, hold, and 3 previews, you occasionally have to rely on more difficult moves (this also means that it's impossible to play with perfect rhythm). If TAP had entry reset, but with the same figures for lock delay, I'm pretty sure that even the best players would find it unplayable. Watch some of kan's death mode replays, and count the number of moves he does that would fail with entry reset. Admittedly, there's not that many, but they're significant enough in number to make things extremely awkward.

    Because TAP players rely on a larger repertoire of moves, some of which clearly require more time to execute than others; in order for entry reset to be playable, the figures for lock delay would have to be increased slightly. So, even though the player only needs that extra bit of lock delay for the occasional move, it has to be given for all moves.

    The added features in Ti serve to make things simpler. (Seriously, you can't possibly argue that they don't). Floor-kicks mean that moves that were previously more difficult and took longer to execute, are now quite playable at high speed; the extra previews allow you to plan further ahead in order to avoid situations that rely on other difficult moves; and even if you don't plan adequately, hold opens up more options that can help avoid difficult situations. Players can now focus on a smaller repertoire of simple moves. It's only possible to play rhythmically (hence making entry reset more feasible) because of these simplifications.

    Even so, entry reset isn't necessary, and ultimately, can never push the player as much as step reset. Although the simplified Ti moves will have very similar execution times, there will always be some that require more lock delay than others: a simple DAS and lock can be done in under 6 frames, but a DAS and tap back might take 7 or 8. The move reset lock delay has to be long enough to allow these slower moves. On the other hand, with step reset, the player might think: "I only have 6 frames of lock delay, but need to do a move that requires 8. To facilitate this move, I need a step." Not only does the step reset player have to deal with fewer frames for the majority of placements, but they're also forced to think more.
     
  14. This is where you differ from caffeine. What's wrong with a little bloat if the lock delay is only meant to push 75% of your speed? Either (with entry reset) There's bloat with simple placements in the middle, or (with step reset) there's bloat with very efficient but more time consuming placements to the sides. The only way out of that is some variation of move reset lock delay.

    There are efficient moves though where you must DAS, go neutral, then rotate, then lock. A very fast operation once you get the timing down, but still more complicated than simply DAS'ing to the wall. If you're looking to aggressively push the player, the system should take into account the complexity and theoretical perfection of the move. I'd say entry reset, but with every action frame not counting towards the lock delay timer. So, rotation and movement wouldn't count. And to prevent infinity style anti finesse, if a piece enters a location it was previously in, it would count towards the timer in spite of the action. So mash-rotating the O would do nothing.

    And I say the game is inconsistent and if lock delay is going to push you at all it should take that into consideration. Step reset and move reset scale the lock delay to the player's actions, albeit only roughly. But with small enough values it works extremely well. Short entry reset lock delay will put much more strain on the player to execute placements on the sides. And if you relax the sides to more reasonable expectations, then you've got a ton of fat for the center placements.
     
  15. DIGITAL

    DIGITAL Unregistered

    I have to admit, I was more or less debating with regards to where the series is heading (Ti's added mechanics). For the earlier entries in the series, it's debatable depending on how you approach it.

    The advanced moves in TGM/TAP were born from step reset. You cannot apply them directly to entry reset. If TGM had started with entry reset, these maneuvers might not have been preferential. Advanced TAP players utilize them because they are advantageous. Step reset encouraged this type of play and that's how the metagame evolved. If entry reset makes it so that these techniques are disadvantageous, then the metagame will adapt.

    You do have a point here. In the context of Ti though, this is a moot point as it's advantageous to stack flat, where step reset provides little to no advantage over entry reset. As for TGM/TAP, the metagame evolved around pyramid stacking so I see why you'd "require" that extra bit of lock delay if you're doing some complex finesse that requires row dropping. I used quotes because you don't absolutely have to stack in a pyramid form just to play.

    Aside from that, you don't have to have step reset to perform complex maneuvers without making it too easy to perform simple maneuvers. Actually, it makes no sense to me. Why reset lock delay by row dropping? Why not by column? I briefed over this idea on the last page.

    Entry reset isn't truly necessary and would barely make a difference (with adjusted lock delay values) in Ti because it'll play very much the same to a player with strong finesse and who stacks flat. Entry reset would be most noticeable to those who try to apply pyramid stacking finesse, but that's not how the metagame in Ti is evolving.

    As for thinking more, that may be the case but consider this. SRS's wallkicks are much more complex than those of ARS and requires you to think more. I see two types of difficulty. One is hard because it's hard to understand (SRS). The other is hard because it's limiting (ARS). Step reset is the former, but to a much lesser degree. Entry reset is the latter. They're both hard in their own way. It's just like how TAP is hard because it's limiting, but Ti is hard because the speed + utilization of hold, 3 previews, and floorkicks is harder to grasp.
     
  16. jujube

    jujube Unregistered

    i don't want to interrupt this discussion again and repeat myself, but i think these poll options need to be updated. the "both ARS and SRS use step reset" option won't work because there is a bug that nobody except me and Zaphod seems to care about. it needs to be fixed if people want step reset with SRS in this tournament. this would probably involve adding code that would recognize when the lowest segment of a piece has stepped down, and only then would the lock delay reset. if we try to fix it by changing the basic rotations then we don't have SRS anymore.

    an easier fix would be to remove the bit of code that makes it possible to reset the lock delay manually. then we would have entry reset. if nobody wants to do this, we don't need the "both ARS and SRS use entry reset" option.

    another way to go would be to add a 4th contest with shirase 500 speed and everyone uses move reset. just an idea.

    i just hope we're not voting for our favorite lock delay system here. we need to do something practical for the tournament.
     
  17. Zaphod77

    Zaphod77 Resident Misinformer

    Here is how to fix step reset for SRS.

    Have it be a combination of step rest, and entry reset.

    When a rotation that causes a floorkick is performed, the y coord of the lowest block of the piece before rotation is stored, and the game switches from step reset to entry reset until the lowest part of the piece drops below tht line. The entry reset delay will be twice the step reset delay.
     

Share This Page