Blazephlozard
He/Him
Banned User
Joined: 2/27/2013
Posts: 175
Location: Ohio
Hmm, alright, perhaps I'll make a copy of the hole 10 R+0 that either adds an I at the start, or if first piece is I, replace that with two L's. (Automatically through code) Taking the I-spin into consideration (as well as L-1's that begin with an I on column 5), I think I'd just want to include the 5th line in L-1 too.
Skilled player (1456)
Joined: 11/26/2011
Posts: 656
Location: RU
Blazephlozard wrote:
Hmm, alright, perhaps I'll make a copy of the hole 10 R+0 that either adds an I at the start, or if first piece is I, replace that with two L's. (Automatically through code) Taking the I-spin into consideration (as well as L-1's that begin with an I on column 5), I think I'd just want to include the 5th line in L-1 too.
Sounds workable, looking forward to see true power of brute force magic! Btw only two J's at right branch can be used, not two L. Foundation with two L's is possible only if top of tree has 4 blocks which is extremely rare. Normal tree with 3 blocks at top can not alow use two L's.
I show you how deep the rabbit hole goes. Current projects: NES: Tetris "fastest 999999" (improvement, with r57shell) Genesis: Adventures of Batman & Robin (with Truncated); Pocahontas; Comix Zone (improvement); Mickey Mania (improvement); RoboCop versus The Terminator (improvement); Gargoyles (with feos)
Blazephlozard
He/Him
Banned User
Joined: 2/27/2013
Posts: 175
Location: Ohio
Looking into the fastest way to drop pieces for trees, on Level 18 (drop speed = 3 frames), holding Down (which increases drop speed to 2 frames) is equal to doing 3 Down, 1 blank (which alternates 3 frames and 1 frame). Though you want to start that Down hold one frame after you gain control of the piece, to save 1 frame. Since 3 on 1 off is equal speed to holding Down, horizontal movement can be done in the off frame (as well as the very first frame) without losing time. But 5th-8th knee will be getting cleared on Level 19, right? So if you want maximum potential frames, they have more limited hole options. (In reality, the extreme difficulty of manipulating the I-pieces would almost definitely lead to RNG delay frames, during which you can do horizontal movement freely, but I figure I should bring it up) Bottom-most hole is column 5/6/7 for early input end (probably 6 for most options) Next highest hole I believe has to be 3-9. And above that, it'd have to be 4-8. Hmm, yeah, it'd have to be two J's... but what about this? I wonder if you could look into some larger 5-line options like this... (Just realized I didn't test that its actually possible to get the pieces there, especially that T on the far right, but I'm sure there's some similar shapes possible) (If we get too into this, using Discord for real-time chatting could be more useful than swapping replies on the forum here)
Skilled player (1456)
Joined: 11/26/2011
Posts: 656
Location: RU
Yeah 5th-8th knees will be will be cleared at level 19 speed. For ending input earlier 6th column is perfect - just single rotation, no extra movements needed. And yes multiple "I" pieces at end for sure will be troubleful to find RNG. But now we very very far from endgame. So too early to discuss it in detals. So far we need to choose debut :) What starting RNG is best for first knee is most actual now.
Blazephlozard wrote:
Hmm, yeah, it'd have to be two J's... but what about this?
That "T"-piece at very rightmost corner is impossible to place like this. --- Small update: Found 3 new cases for L+1 subgroup for hole#10 (and few other holes):
I-J-S-T-T-I

I-J-S-S-J-I

I-J-S-L-I-I
And respectively 3 their twins with "J" at end for L+0 subgroup for cental hole (and few other holes):
I show you how deep the rabbit hole goes. Current projects: NES: Tetris "fastest 999999" (improvement, with r57shell) Genesis: Adventures of Batman & Robin (with Truncated); Pocahontas; Comix Zone (improvement); Mickey Mania (improvement); RoboCop versus The Terminator (improvement); Gargoyles (with feos)
Skilled player (1456)
Joined: 11/26/2011
Posts: 656
Location: RU
Due to unexpected circumstances with recent ban of Blazephlozard at TASVideos, our discussion here is no longer possible. --- But project is not canceled. We will continue to work via email conversation. Fun fact, a month ago i wrote:
Archanfel wrote:
In current TAS Acmlm played tetris like god. In the new TAS we will play tetris like devil.
Seems i was a prophet - to make good Tetris TAS need to make deal with devil.
I show you how deep the rabbit hole goes. Current projects: NES: Tetris "fastest 999999" (improvement, with r57shell) Genesis: Adventures of Batman & Robin (with Truncated); Pocahontas; Comix Zone (improvement); Mickey Mania (improvement); RoboCop versus The Terminator (improvement); Gargoyles (with feos)
Challenger
He/Him
Skilled player (1686)
Joined: 2/23/2016
Posts: 1061
LOL Good luck improving the current run ;)
My homepage --Currently not much motived for TASing as before...-- But I'm still working.
Post subject: Why can't this piece be tucked?
Joined: 8/27/2020
Posts: 5
Why doesn't this piece in this image tuck? https://imgur.com/BbvCVYu I believe I tried every frame even though I only had to test two. This is the inputs on that specific frame: https://imgur.com/g57ZZZ2 Here's the input file: https://filebin.net/qg5w1a06cry81k4a
Skilled player (1456)
Joined: 11/26/2011
Posts: 656
Location: RU
ChallengingRiptide wrote:
Why doesn't this piece in this image tuck?
Not sure your question is serious or joke (or i misunderstood it). But seems you simply press L too early. Try to press it at frame 1092 and it will tuck. --- Some news about TAS: 1)Discovered a few new useful tricks: a)Pausing trick allow not only "I piece" at left but also two "L pieces" (and some other). It expand 3-tetris packing possibilities at left side. b)With help of LR input is possible to make two consecutive inputs! It alow to move pieces further to right. It significantly expand possibilities packing at right side. 2)Turn out that placement of Tree is sigificant. It can affect playfienld. Invisivle blocks can appear above top that can ruin good attempt. The playfield starts at 0x0400 in RAM. That's the topmost visible row, let's say row 0. Row 1 starts at 0x040A (10 further in hex) and so on. When pieces are moving in rows -1 and -2, it'll check at 0x0400 - 10 and - 20, but that underflows to being 0x0400 + 246 or + 236. That's 0x04EC and 0x04F6. Normally these values would only get filled by blocks you place, but the glitchy triple-tetris pushes the bottom most row. Due to it being "+246" instead of "+250", the invisibles are on the right, offset by 4 spots from where the tree is. So needed to be very careful with tree placement to avoid problems with invisible blocks. 3)I finished complete guides for 3-tetris packing for all 10 holes: Hole#10 Hole#09 Hole#08 Hole#07 Hole#06 Hole#05 Hole#04 Hole#03 Hole#02 Hole#01 This data need to be load load into lua script(s) to be able use bruteforce. As soon as loading be ready actual TAS can start recording. 4)Also i found a lot better RNG start for first tetris. Now current best start for first knee is 68 frames ahead of acmlm.
I show you how deep the rabbit hole goes. Current projects: NES: Tetris "fastest 999999" (improvement, with r57shell) Genesis: Adventures of Batman & Robin (with Truncated); Pocahontas; Comix Zone (improvement); Mickey Mania (improvement); RoboCop versus The Terminator (improvement); Gargoyles (with feos)
Skilled player (1456)
Joined: 11/26/2011
Posts: 656
Location: RU
Recenly r57shell got interested in Tetris project. He is very skilled coder. Insead of re-creating old Acmlm approach, he proposed to try completely new approach. More advanced. He made generator that can search not Left and Right cases separately, but all possible 3-tetris packing sequences as one whole. Later bot with brute force can autocomplete any case from this data base. That generator found even more cases than i was able to find by hands. Like in all other "Man vs. Machine" competitions the man came last. So case lists in my previous post now incomplete and outdated. New aproach has great potential. With such powerfull tools will be possible execute new TAS even better. It will not play Tetris like god. It will not even play Tetris like devil. Nope. In new TAS it will be play of Tetris beyond good and evil.
I show you how deep the rabbit hole goes. Current projects: NES: Tetris "fastest 999999" (improvement, with r57shell) Genesis: Adventures of Batman & Robin (with Truncated); Pocahontas; Comix Zone (improvement); Mickey Mania (improvement); RoboCop versus The Terminator (improvement); Gargoyles (with feos)
Personman
Other
Joined: 4/20/2008
Posts: 465
Heya, just wanted to say I'm excited for this TAS and thoroughly enjoy your updates :)
A warb degombs the brangy. Your gitch zanks and leils the warb.
Skilled player (1456)
Joined: 11/26/2011
Posts: 656
Location: RU
Personman wrote:
Heya, just wanted to say I'm excited for this TAS and thoroughly enjoy your updates :)
A few days ago we finished the first full test-TAS: Link to video It is about 11.7 seconds faster than currently-published TAS by Acmlm. 400+ frames were saved before transition to level 19 and also 300+ frames saved after transition. Keep in mind that it is only rough test-TAS. It is far from perfect. A lot of work still wait ahead. With some changes in strategy and in tactics yet a couple of seconds will be possible to squeeze. Further improvements that we want to apply are too complex to explain and even more complex to implement. The only one thing can be said for sure - final form of new Tetris TAS will blow your mind.
I show you how deep the rabbit hole goes. Current projects: NES: Tetris "fastest 999999" (improvement, with r57shell) Genesis: Adventures of Batman & Robin (with Truncated); Pocahontas; Comix Zone (improvement); Mickey Mania (improvement); RoboCop versus The Terminator (improvement); Gargoyles (with feos)
Skilled player (1020)
Joined: 1/9/2011
Posts: 230
Fantastic! I'm looking forward to the final form, as well.
Challenger
He/Him
Skilled player (1686)
Joined: 2/23/2016
Posts: 1061
Very cool strategy! Working on this game is so, so laborious... and I never expected to see more than 10 seconds of improvement. Looking forward for the lastest version :)
My homepage --Currently not much motived for TASing as before...-- But I'm still working.
Skilled player (1456)
Joined: 11/26/2011
Posts: 656
Location: RU
Challenger wrote:
Very cool strategy! Working on this game is so, so laborious... and I never expected to see more than 10 seconds of improvement. Looking forward for the lastest version :)
The strategy that used in test-TAS will look like a naive childish plan compared to the true beauty of Machiavelli's plan which we dare to put into final form. If everything will goes perfectly may be even 15 seconds of improvement can be achieved. Holy Grail of Tetris TAS is to get 4 "I"-tetromino in row at very end. It is really big timesaver, at least another 100+ frames can be saved this way. However to fulfill the conditions that allow to get such flawless ending need to do absolutely insane things. Besides getting grail itself, there are also multiple other smaller treasures here and there each of them require non trivial changes to get these improvements works. Hell a lot of preparation work is still needed before it be possible to start recording final version. Tetris is like abyss deeper you go the darker it become. Fortunately comrade r57shell is crazy enough to jump into deepest parts of Tetris abyss and illuminate it with using bright flash of his programming skills.
I show you how deep the rabbit hole goes. Current projects: NES: Tetris "fastest 999999" (improvement, with r57shell) Genesis: Adventures of Batman & Robin (with Truncated); Pocahontas; Comix Zone (improvement); Mickey Mania (improvement); RoboCop versus The Terminator (improvement); Gargoyles (with feos)
Joined: 8/14/2007
Posts: 21
So some interesting developments have occurred within Tetris that have given rise to two possible categories: 1) Fastest to kill screen 2) Fastest to loop Tetris What's going on is that there is a minefield of game ending bugs after a certain point, you hit one and it halts the game (thus a true kill screen). Where TASing becomes interesting here is that not only can the TAS avoid all of the myriad of kill screens... but also give a rough idea of how far you can push it. I mean, it's only a matter of time until a human does this... Link to video
Joined: 10/8/2023
Posts: 4
Location: Edmonton
It would be cool to see fastest kill screen with everything new that's happened, opposed to loop anyway, but honestly both would be beat to see.
DrD2k9
He/Him
Editor, Judge, Expert player (2210)
Joined: 8/21/2016
Posts: 1086
Location: US
Almagnus1 wrote:
So some interesting developments have occurred within Tetris that have given rise to two possible categories: 1) Fastest to kill screen
I think option 1 would be better termed “Fastest Crash” because of the various game crashing bugs. In my opinion, for something to be considered a kill screen, it needs to be: 1) consistent across all play throughs of the game and 2) not have when that endpoint/kill screen occurs be impacted by gameplay (at least from a human standpoint). For example, the Duck Hunt kill screen of Level 0 is not beatable but a human (though it is by a TAS). Because that level will always been effectively unbeatable by a human in normal gameplay, it’s a kill screen. Likewise, Pac-man’s kill screen is also always on the same level. But with Tetris, because the crash can happen from various bugs, it’s not a consistent kill screen that would impact all players equally who reached a certain point in the game. Instead, when it occurs depends in the player’s actions.
Joined: 8/14/2007
Posts: 21
Here's the raw footage for the first human crash of Tetris, which took about 38 minutes of playtime from level start to crash. Link to video Edit: For those interested in the technique they're using: Link to video