Tetris DS. Standard Marathon Mode (200 line clear) cleared in 1 minute and 18.7 seconds, complete with must-watch credits. Enjoy!
This movie was made in DeSMuMe 0.9.4+ r2881. It works also on 0.9.4+ r2872, and probably, the newest SVN version.
Fix ARM Cycle Count must be on. Fast Fetch-Execute should be left off.
Aims for fastest completion of Standard Marathon.
About the game
Tetris DS is well-known for its intense Nintendo theme. It is also known for three other things that make it appealing to TAS:
The hold piece, which allows a player more freedom.
Sonic or Hard drop; where pressing up causes the blocks to fall immediately. This isn't new; it existed in some PC Tetris versions.
No delay between pieces, except during line clears. Similar to PC versions. In Tetris terminology, ARE=0.
Sliding pieces along a surface. If a piece comes to rest on a surface, or is nearby, rotating it will cause it to slither along the surface, sometimes in unexpected ways. This is particularly noticeable in the late stages of this run. It is possible to create "T-Spins" with the T-Block.
About the run
This run took longer than expected. Every time I thought I had figured out something, the game threw in something new. I redid the last few sections three times because I kept running into I-Block trouble (I need 50 Tetrises). The last 8 lines took a long time to optimize. Due to the unlucky piece sequence, I cannot complete it in 500 pieces, needing 501. To complete it in 500 pieces requires either an even number of T-Blocks in the first 500 pieces, or an I-Block as the 501st piece, neither of which happened.
I initially made the run without using Fix ARM Cycle Count. After completing the run, I figured out that Fix ARM Cycle Count is required to avoid a graphical glitching of the Level 4 transition (when the run reaches 40 lines). So I edited the DSM by a menu redo and gameplay copypaste. In doing this, I saved 38 frames on menu transitions and 1 frame in the gameplay itself.
Fix ARM Cycle Count is also required for Catch Mode to work correctly, although Catch Mode is not relevant to this run.
Gameplay is 4722 frames (1m 18.7s) long, from the first "ghost piece" appearance to when the line count reaches 200. A TAS I found on YouTube goes from frame 807 to 5571 (4765 frames). However, it is not exactly known what the beginning and ending points were.
Thanks
Thanks to Baxter for various discussions about a possible TAS showing off different modes. Since selecting what to show off is subjective, and potentially long, I decided to do a fastest run on the showcase mode because it is short and objective.
Thanks also to an anonymous TASer (probably a Japanese speaker) who made the TAS linked above so I can compare.
mmbossman: Fast, technically sound (the youtube version that was posted is tough to verify, if someone can find the source for it, I reserve the right to review this decision), people like it, and the goal is easy to understand. Accepting for publication.
mmbossman: I would like to apologize to FractalFusion for prematurely accepting this. I thought I had followed along with the thread well, but I somehow managed to miss this post, which provides a [dead link removed] to a .dsm input file that beats this same mode with same goals in 1:37.53, or just shy of 3 seconds faster than this submitted movie. And, as I stated above, and as is stated in the rules, since this movie does not beat all existing records when aiming for the same goals, I am now setting this to "Rejected". I apologize again to FractalFusion for this oversight.
There are many different Tetris engines with different rotation rules, drops, acceleration, wallkicks, latencies, etc.. You'd be surprised how technical a seemingly simple game like this can be.
That was fast! All I hear was almost constantly (almost every second) the sound of completing a Tetris.
Entertaining to watch!
To sum it up: it's short and sweet!
Voting Yes!
Yes vote for obsoleting all the published Tetrises!
<klmz> it reminds me of that people used to keep quoting adelikat's IRC statements in the old good days
<adelikat> no doubt
<adelikat> klmz, they still do
I've held off posting my opinion in this thread since I didn't want it to cloud other people's opinions, and I certainly didn't want this to be the first post of the submissions thread. As Fractal_Fusion mentions in the thread, we had some discussion on the goals, and this post is in no way meant to be against Fractal_Fusion (I really hope he doesn't dislike me for making this post)... I'm just quite critical of this TAS.
Let me start off with the claims that it's more awesome or even that it should obsolete the published TASes:
FODA wrote:
Baxter wrote:
klmz wrote:
Yes vote for obsoleting all the published Tetrises!
That's like a New Super Mario Bros DS TAS obsoleting all published Mario TASes :/
No because it isn't the same game. It would be like that GBA super mario 3 remake obsoleting smb3 on the nes and snes.
As moozooh mentioned, this isn't remotely the same game either. There are different goals, physics, graphics and music. But let me say why in particularly I think it's not more awesome than any of the three published TASes it resembles the most.
Tetris A mode, 999999 points: needs to take advantage of making a high platform, is constantly on the verge of dying, has a lot of block manipulation. One of the coolest facts that's not really noticed generally though I think is that making all tetrisses only gives you like 999500 points (could be 100 more of less). You need to pressing down a frame when dropping a piece gives you 1 point when it locks, but because you are constantly so high and there is no moving space, I just managed to get the exact amount of points with the final piece dropping. The TAS doesn't get more than 999999 points, it gets exactly 999999 points.
Tetris B mode, all clear 25 lines: Very awesome and inventive objective to get a bravo (all clear) when finishing B mode on height 5. A lot of manipulation had to go in getting a good initial setup, and piece manipulation, and a whole lot of planning.
Tetris DX 40 lines mode at height 5: This is arguably by far the least impressive of these three TASes... but in a TASing perspective, I think it's still a little more impressive than this particular submission. Mostly because the TAS shows the potential of the game better.
===============================================
Fractal_Fusion chose this mode because it was the most because it's the most objective. If by objective is meant comparable, then it's indeed amongst the objective goals but there are many others that are equally well open for TAS competition. Amongst those though, I think this is probably the most boring mode.
It's the standard, complete 200 lines mode, and while it does this quite quickly, there is nothing really to it that makes it special. It are 50 Tetrisses without ever having to get in a tight spot, since you don't have to build high. While the pieces can't be manipulated like in the other games, the amount of freedom with the hold function I think eliminates nearly all planning, if you just build in a steady way, without making big gaps. Even with this relatively boring goal I think the movie could have been made more interesting, which in my opinion is shown by the youtube movie linked in the submission text. It has lots and lots of bravos (all clears), which requires more effort (and my philosophy is "effort = entertainment"). The second part of this particular youtube movie that aims for highscore (starting at level 20... basically the only reason why the first 200 lines are played) also requires a lot of planning to make all the T-spin triples (I personally would have made a highscore attempt at the 25 lines mode at height 5... as 200 lines mode can be set to endless, making a higher score easy on the regular handheld). At any rate, I think the youtube movie is far superior to this submission (even if the first run through 200 lines might be a tiny bit slower).
Does this mean that I would be in favor of publishing that youtube movie (if it were submitted)? Well, not really. This game truly has a lot to offer. This is what I wrote in the Tetris DS thread that I think would be great things to all include in a Tetris DS TAS:
Baxter wrote:
These are the modes, goals, and the order I plan on doing in the TAS:
- Mission time trial mode, level 1-5
- Catch mode, unlocking level 20 as fast as possible
- Regular marathon mode, unlocking level 20 (should this be done as fast as possible?)
- Push mode, (If it works, first game, get a draw) then beat level 5 as fast as possible
- Mission marathon mode, unlocking level 20 while playing around
- Catch mode, get the maximum score of 99.999.999 as fast as possible
- Regular vs mode, beat level 5 as fast as possible
- Regular line clear mode, get highest possible score as fast as possible
- Mission marathon mode, get maximum possible score (as fast as possible)
Making a TAS like this takes quite a bit of time, which amongst emulator differences that don't support my current WIP and other projects didn't make me work on this for quite some time now. No offense at all meant to Fractal_Fusion... but this 200 line submission I could have made months ago, in a matter of one or maybe two days of work.
I don't think we need a published TAS for each of these modes, I am in favor of a single Tetris DS TAS which includes them all... including mostly modes and goals that are quite different than anything else published on the site.
I do recognize why there is such a positive response... the gameplay is really fast, and it looks entertaining. This is however mostly due to how the game is, and not particularly due to the goals or effort put into the TAS. Voting no.
This is probably the best movie I've ever watched on this site.
Or close, at least. Top three. So congratulations, that earns you a big ol' No vote from me!
...
Well,
- It was fast
- It isn't glitched to the point I'm left wondering what in the world is going on or otherwise unwatchable
- It was hilarious (seriously, just try watching how fast those things start spinning)
I still love the end credits.
It definitely should not obsolete existing Tetris runs for the reasons Baxter points out. However I do think it is quite entertaining and should be published. The option to obsolete it with a longer/more varied DS run would still be there.