20G ARS Strategy - Chapter 0 - Introduction/Discussion

[Edo]

Seeing that 20G is the play mechanic that gives people new to TGM the most trouble (heck, even veterans find it hard once the speed ramps up...), I thought some sort of strategy guide would be very useful. Now, I know some of you are going to say that it's not really possible to put your 20G knowledge down in words, or even diagrams, but J.O. has done a pretty good job of it (http://www13.plala.or.jp/TETRiS_TGM/index.html), and I'm sure we can do even better! We have an entire community of TGM players here, and some of us are actually not bad (DIGITAL, I'm looking at you, kid http://www.tetrisconcept.net/forum/images/smilies/icon_wink.gif ). Regardless of our experience level, we can all make a contribution; either by volunteering our wisdom, or by asking the right questions.


So. Fellas? Are ya with me?


Good, I knew I could count on you! I've given this some thought over Christmas, and I think the biggest problem we'll face is actually keeping it stuctured, and preventing it from becoming a messy free-for-all. What I propose we do is break down the concept of 20G into its component aspects - i.e. break down 20G into 5 distinct, easy to learn "chapters".


Here are my suggestions:


Chapter 1 - Mobility

Explaining simply how the pieces move, rather than going into detail of what constitutes a good or bad move. This will probably include a systematic coverage of wallkicks and synchros, IRS and the like. The importance of pyramid structure, with the peak at the 5th column, should also be covered.


Chapter 2 - Perfect Stability

This is where we cover what makes a good move, and how to play so as to maintain mobility indefinitely. I expect this will cover a lot of common patterns and themes, so we'll probably need to use either the fumen website (http://fumen.zui.jp/default.htm), or jago's The TOOL (http://www.tetrisconcept.com/forum/viewtopic.php?t=24&), or some other alternative.


Chapter 3 - Compromise

This is all about knowing when to create a small hole that's easy to recover from, to save yourself from creating a much larger hole later on, that's impossible to recover from. Also knowing how to minimize damage, and take the lesser of two evils.


Chapter 4 - Recovery

How to regain stability after making a massive mis-drop, or playing a sub-optimal move that creates a big hole.


Chapter 5 - Style, Finesse, and Hard Core Time Attack Techniques

Everything you need to know about playing faster.


These are only preliminary, so please suggest better Chapters if you can think of any. Perhaps Compromise and Recovery could be rolled into one, and maybe Stability needs more than just one chapter.


Anyways, once we've agreed on the skeleton of the project, we'll then create a seperate thread for each Chapter, and start filling in the detail. I expect that despite our best efforts, things will inevitably get a little messy, so after we finish, the whole thing will get wiki-fied, and neatened up in the process. This could even become part of Petit's excellent TGM Guide (http://www.tetrisconcept.com/wiki/index.php?title=TGM_Guide), if he so wishes.


We'll want to plan things as best we can, and I expect we'll also want to assess our progress along the way, so, we'll reserve this Chapter 0 thread for any questions/suggestions/messy discussion we want to have.


So, does anybody want to start the ball rolling by asking a question or proposing different Chapters?

[Red_Star]

okay, I have few questions.


say you have stacked up and you find that you have a few lines with holes in them and the holes are disperse( like one hole near the right side then one clear on the other side). what is the best way to approach getting rid of these lines?


another problem i have had with the holes is after getting rid of the holes i have found that i have wasted a good amount of time and all of the single and double line clears have given me very little points, would it be better to just keep stacking for tetrises on top of the lines with holes?


Finally, say you do something dumb like me and this happens to you


http://i199.photobucket.com/albums/aa73/darkheroaja/20gquestions.jpg

is there a twist to get over that misplaced I piece, not just for the L piece but for any piece?


I'll have more questions for everyone soon.

[DIGITAL]

okay, I have few questions.

say you have stacked up and you find that you have a few lines with holes in them and the holes are disperse( like one hole near the right side then one clear on the other side). what is the best way to approach getting rid of these lines?

another problem i have had with the holes is after getting rid of the holes i have found that i have wasted a good amount of time and all of the single and double line clears have given me very little points, would it be better to just keep stacking for tetrises on top of the lines with holes?
My advice is to stack over the holes if they are very dispersed. You want to clear them with triples and tetrises. Take as much time as you need but just avoid getting tangled in a mess of singles and doubles. Try as best you can to not overstack when building the triples and tetrises.
Finally, say you do something dumb like me and this happens to you

*picture*

is there a twist to get over that misplaced I piece, not just for the L piece but for any piece?

I'll have more questions for everyone soon.

Yes there is for the J.

http://www13.plala.or.jp/TETRiS_TGM/kouza/9pics/02.gif

[DIGITAL]

Here's a rough outline of the three areas I think are most important and most basic (subtopics are in no way final or even close to it). Is there any topic that doesn't fall in any of these?


What can be done? (Mobility)

- Basic rotational states

- Spawn position

- Spawn orientation

- IRS

- Wallkicks

- Twists(with & without wallkicks)

- Synchro

- Block synchro

- Double rotation techniques


What should be done? (Stability)

- Stacking patterns

--- Hole Transfer

- Pyramid

--- High center

--- Column specific

- Overhang

--- Edge

--- Middle

--- Line clear

- Compromise

--- Skimming

--- Intentional platform/fill

- Recovery


How should it be done? (Efficiency)

- "Freefall" finesse
- Wall pushoff (wallkick) finesse
- Block synchro
- DAS finesse
- Rotational state finesse

--- Overhang "tuck"

- Double rotation finesse
- Level stop strategies

[Poochy]

is there a twist to get over that misplaced I piece, not just for the L piece but for any piece?I think the best you could do is to place the L on the left, and hope for a J or I. (If the next piece is a T, Z, or O, place it in the divot and hope for a J or I after that.) The placement for a I should be obvious. For a J, DAS to the right, and while holding the DAS, rotate counterclockwise, clockwise, then counterclockwise again. That'll move the J over and into the valley on the right.


This is actually another advantage of building with the empty column(s) on the right - due to the asymmetric wall kicks in ARS, the J block can actually wall-kick its way over a 1x1-sized bump when moving to the right.


Here's an illustration using the TOOL:

http://webpages.charter.net/poochyexe/TAP/JTwist.gif
And some alternate placements:

http://webpages.charter.net/poochyexe/TAP/JTwist2.png
(First one is for if you get the I next, or if you don't think you can do that maneuver with the L fast enough. The others are for if you pull it off but the next piece is a T, Z, or O.)

[Amnesia]

Here's an illustration using the TOOL:




On the first GIF, there is an useless "right" with the "L".

You can directely perform a 180° and move it on the left..Then the end of the movement is correct..

[Edo]

On the first GIF, there is an useless "right" with the "L".
You can directely perform a 180° and move it on the left..Then the end of the movement is correct..
Yes Amnesia, this is partly true, although there is a real danger of the L getting stuck in the 2 cell wide gap, due to not performing the 180 fast enough. Some players find that a slight "feint" in the opposite direction helps with the timing. I believe this technique was also used in jin8's Ti Master GM video, at level 345. Obviously, this technique would be useless at much higher speed 20G.

Here's a rough outline of the three areas I think are most important and most basic...

So you're suggesting more of a "topic and sub-topic" approach, rather than having 5 or so big chapters. I think this is a good idea, and here's my suggestion for making it work: people volunteer to cover a sub-topic, and then create a thread for themselves, we'll try to keep each thread reserved for the author alone, in order to avoid the problem of "too many cooks spoiling the broth". If after they've finished, someone else feels more needs to be added, they can carefully add on more info to the thread, or if needs be, create another thread to provide an entirely different take on the sub-topic. I realize this may result in a lot of separate threads, but if we make a cenral, stickied index or contents, that links to all the separate threads, we should be able to keep it nicely organized and structured.


What does everyone else think, does anyone want to suggest an improvement, or an entirely different way of doing things?


P.S. I call dibs on edge overhangs, DIGI, you can have middle and line clears http://www.tetrisconcept.net/forum/images/smilies/icon_wink.gif

[Caithness]

So you're suggesting more of a "topic and sub-topic" approach, rather than having 5 or so big chapters. I think this is a good idea, and here's my suggestion for making it work: people volunteer to cover a sub-topic, and then create a thread for themselves, we'll try to keep each thread reserved for the author alone, in order to avoid the problem of "too many cooks spoiling the broth". If after they've finished, someone else feels more needs to be added, they can carefully add on more info to the thread, or if needs be, create another thread to provide an entirely different take on the sub-topic. I realize this may result in a lot of separate threads, but if we make a cenral, stickied index or contents, that links to all the separate threads, we should be able to keep it nicely organized and structured.


If we're going to have that many threads, I think it would be best to petition for a third subforum to keep this all in.

[caffeine]

sorry, go directly to jail. do not pass go.


(no)

[Edo]

I agree with caffeine on this one. Making a separate subforum devoted to ARS will most likely give people the impression that we're strictly an ARS/TGM community, which we're not. I'd say we have equal proportions of ARS and SRS lovers here, and we also have our fair share of folks that prefer the classic games. We should try not to lose this diversity.

[Poochy]

On the first GIF, there is an useless "right" with the "L".
You can directely perform a 180° and move it on the left..Then the end of the movement is correct..Actually, that was a strategy I saw on a Japanese fansite (can't remember which one) as well as in Jin8's video. It allows you to load DAS while doing the 180°, saving a few precious frames without having to increase the risk of hitting the second rotate a frame or two too late and getting the L stuck in the divot.

[Rosti LFC]

Would it not be better to simply stick the project directly onto the wiki, marked that it's under construction?


Any discussion could go here (split into several threads if needed) or on the discussion pages on the wiki.

[Red_Star]

it would seem a lot easier to view the progress if it was in the forum, at the same time it would seem like it could congest other topics in the forum. Maybe have one thread that outlines the updates and progress being made on the wiki so people can see what has been added.


I like the idea of the sub-topics but could it be done in a chapter fashion as edo first put. (ex/ Book 1 mobility, Ch.1 Basic rotational states) have it so new players would know what to work on first, that way they can get the basics down then progress to the more difficult stuff. It could give a way for new players to approach practicing 20g. just my suggestion.

[PetitPrince]

I like Digital's approach. It has "reference document" feeling that should reassure any 20G newbie.


I like the idea of the sub-topics but could it be done in a chapter fashion as edo first put. (ex/ Book 1 mobility, Ch.1 Basic rotational states) have it so new players would know what to work on first, that way they can get the basics down then progress to the more difficult stuff. It could give a way for new players to approach practicing 20G. just my suggestion.

Red Star pointed out an important thing. We should include some sort of "learning path" that would guide them. Or at least something that prevent them from learning a non-obvious wallkick such as this one (http://jagorochi.free.fr/absolute/images/Wall-kick_tech_L8.gif) before even knowing how to use overhangs.

[Rosti LFC]

Or at least something that prevent them from learning a non-obvious wallkick such as this one (http://jagorochi.free.fr/absolute/images/Wall-kick_tech_L8.gif) before even knowing how to use overhangs.


But for me, those sorts of wallkicks are part of knowing how to use overhangs.

[Zednom]

Yes. YES. http://www.tetrisconcept.net/forum/images/smilies/icon_twisted.gif

You don't know how badly I need this guide right now.


Could I request something in particular? For gifs that show animations on what to do in certain situations and how to rotate to get the right wallkicks and such, can those be posted in 1 "real time" speed of how fast you really need to do it, and 1 "slow motion" to really observe what's happening?


If that's too much effort then no worries. Thanks for starting this up though, 20G is really killing me.

[Red_Star]

Zednom:

Could I request something in particular? For gifs that show animations on what to do in certain situations and how to rotate to get the right wallkicks and such, can those be posted in 1 "real time" speed of how fast you really need to do it, and 1 "slow motion" to really observe what's happening?



heh, i was having a problem telling what was going on in some of the gifs especially the ones with real quick movements and quick consecutive rotations.


like he said, if it takes more effort to do then no worries. i'm more than willing to look at a gif a couple times to get it.


Also, again with the chapter with sub topics, since this seems like a big project it might take some time. So, if it was split into chapters based on basics stuff progressing to hardcore stuff, then it would give a structure on what part of the guide to work on first. That way the guide doesn't have the parts about finesse and time saving strategy done before it has what rotations are possible and how to approach stacking. just my suggestion.

[DIGITAL]

I normally don't like making animations anyway. I can understand the advantage of both freeze frame diagrams and animations but I prefer the former for the ease and clarity (not that an animation isn't clear in every case).

[Caithness]

Something I find very useful for learning 20G ARS is training in Lockjaw. Using the settings for Death on the Simulation Settings wiki page, I just change Lock Delay to Manual lock. Then I have the freedom to not only see which moves and rotations are possible, but feel them in a hands-on sort of way.

[Edo]

Could I request something in particular? For gifs that show animations on what to do in certain situations and how to rotate to get the right wallkicks and such, can those be posted in 1 "real time" speed of how fast you really need to do it, and 1 "slow motion" to really observe what's happening?
This is definitely possible. I wouldn't exactly say I'm very familiar with The TOOL (the program used to make all those fancy gifs), but I believe that it's very easy to simply increase the delay on each frame.

Something I find very useful for learning 20G ARS is training in Lockjaw. Using the settings for Death on the Simulation Settings wiki page, I just change Lock Delay to Manual lock. Then I have the freedom to not only see which moves and rotations are possible, but feel them in a hands-on sort of way.

This is a really great idea. I'm of the opinion that when training, piece placement should always be a cognitive process. Set the lock delay as high as you want; give yourself all the time you need to decide on what you think is the best placement. It doesn't matter if you turn out to be wrong, as long as you're actually engaging your brain. As you get more confident, crank up the speed to force yourself to think faster.


Okay, on with the project. I think no matter what approach we take, it's clear that the first "section" or "chapter" will deal with mobility. If people are in agreement, we could make a Chapter 1 thread now and begin work. And then whilst we're busy with that, we can continue thinking about the best way forward for when we finish the first part.

[DIGITAL]

I'm ready if you are.

[PetitPrince]

This is definitely possible. I wouldn't exactly say I'm very familiar with The TOOL (the program used to make all those fancy gifs), but I believe that it's very easy to simply increase the delay on each frame.

Yup, definitely doable:


http://bsixcentdouze.free.fr/tc/lkick3.gifhttp://bsixcentdouze.free.fr/tc/lkick4.gif

[Red_Star]

so i started playing 20g on tgm and i found that i have trouble with what to do with the I is a real problem. If i daon't have it set up for a tetris yet then i'm stuck either making it flat which blocks off everything or making it vertical which can hinder movement of blocks to the left or right depending on where it is. for the majority of situations it's better to go flat right?


I'm also having trouble with the movement of the L and J blocks. i'm positive they can go over blocks if you rotate them right but i can't seem to pull it off right.

[colour_thief]

Vertical in the middle works pretty good sometimes, especially column 5. That's the "center" column, so you retain all mobility. Vertical against a wall is pretty good too in a lot of cases.

[Edo]

Gosh, sorry about this, I've been delaying starting this project for absolutely ages! I've recently learnt how to use Jago's The TOOL properly, so now it's only my own laziness holding me back. I'll try to get something started tomorrow. Or if DIGITAL or one of the other guys is keen to start without me, feel free by all means.


As for the question about the I piece, personally, I stick it vertically in column 5 a lot. If you watch my M rank video (http://www.youtube.com/watch?v=6PA1YgjNc70) [/shameless plug], I do this move at 157, 236, 270 and 418. The placement at 418 might look like it gives me trouble, but to be honest, I think the I placement here is a sound move; the S at 425 is probably the cause of my subsequent mis-drop with the J.


Sticking the I vertically in column 5 is what I would call a "default move", a move that I fall back on if I can't see a better place to put it. Another one of my default moves that I really like is S IRS'ed and shifted one to the right, waiting for an L to kick under it. Similarly, Z IRS'ed and shifted one to the left, waiting for a J to kick under it, does the same.

[DarthVivi]

Was this ever made? I noticed in the wiki there is a lot of stacking strategy for SRS/low gravity, but almost nothing for ARS 20G. I can't get to level 1000 in TGM1ARS even with infinite lock delay, so I know my strategy is inferior, but it's not intuitive to me how to fix it. A guide like this would be very helpful to me and I'm sure the rest of the western Tetris community as well.

[Edo]

Was this ever made?
Unfortunately not. I kept on delaying starting, and in the end, other commitments got in the way.

[colour_thief]

For serious though DarthVivi, nothing you read will make you make the right decisions at one piece per second. You've got to play it out, trying stuff like preventing holes in the middle of the screen. If you think hard about that as your goal and also learn how the I piece rotates, you could become quite competent over a month of regular, small play sessions. Whenever you make a hole in the middle, just pause and think about what lead to that hole... Usually it's an unwise placement a few pieces earlier that reduced your stack's stability.

There's a certain value to playing a large volume of pieces (by which I mean not abusing your infinite lock delay too much). On some basic level you've got to condition yourself to automatically avoid the most foolish placements, and again this requires repetition. When you're first starting out, the game is more about learning to avoid bad moves than learning to make especially good ones (like fancy recoveries).

[COL]

I was going to make a separate thread but this one is perfectly suitable for this:
This has happened to me:
http://img31.imageshack.us/img31/9194/0018.png

The pieces came in the obvious order (no t-spins: L,I,T,J http://www.tetrisconcept.net/forum/images/smilies/icon_evil.gif ,S http://www.tetrisconcept.net/forum/images/smilies/icon_evil.gif , then i stopped to play seriously ..)
It seems that after the T, nothing could be really done...any idea?

[Rosti LFC]

You can't do anything about the T other than not have a dip in the middle of the field.

However, if you'd IRS'd the L piece anti-clockwise, you'd have been able to get it down the right hand side which would have given you a lot nicer stack.

Really though, you should have arranged the first few pieces better, so that columns 5 and 6 were the tallest ones. You didn't stack in a pyramid shape.

[colour_thief]

http://img31.imageshack.us/img31/9194/0018.png
The pieces came in the obvious order (no t-spins: L,I,T,J http://www.tetrisconcept.net/forum/images/smilies/icon_evil.gif ,S http://www.tetrisconcept.net/forum/images/smilies/icon_evil.gif , then i stopped to play seriously ..)
It seems that after the T, nothing could be really done...any idea?

Your biggest mistake was the I placement, which cut off the right part of your screen. I would never open with the L like that either. It's already cutting off the right side, though less so. In general you want to keep columns 4 and 5 highest so that you can move pieces to both sides. This prevents holes in the middle.

[DDRKirby(ISQ)]

help my openings suck =D

Here's what I used to do:

S/Z: vertical, with column 5 highest
L/J: vertical, with column 5 highest
I: vertical, with column 5 highest
T: rotate 180, with column 5 highest
O: one to the left

I'm finding having a huge....."tower" in the middle isn't always quite so beneficial.

So I've changed:
L/J : flat, with column 5 highest
I: flat, columns 4-7 (no movement)
O: DAS to left

Obviously openings aren't really THE most important thing out there, but it'd be nice to learn for example, whether people prefer to concentrate on getting column 5 highest, or whether they slide their initial piece over to the left in order to start with a "build high on left, garbage on right" strategy.

The main resaon I put Z/S in the middle is because those are the overhangs I'm most comfortable with fixing.

[colour_thief]

What game are you playing that's 20G and gives you S Z and O as your first piece?

[tepples]

TDS level 20, for one. Or Lockjaw with the randomizer not set to the TGM setting.

[muf]

TDS level 20, for one. Or Lockjaw with the randomizer not set to the TGM setting.
He was asking DDRKirby.

[kotetsu213]

Texmaster does it sometimes if you don't start at 0.

[Zircean]

It could happen in TAP/Ti when the field is cleared after you hit 999 and your next piece is an S, Z or O too.

[colour_thief]

It could happen in TAP/Ti when the field is cleared after you hit 999 and your next piece is an S, Z or O too.

Though if you can get to 999 you don't need to ask about opening moves. "The pieces are invisible, what am I supposed to do with the first Z piece??!?" http://www.tetrisconcept.net/forum/images/smilies/icon_lol.gif

[DDRKirby(ISQ)]

What game are you playing that's 20G and gives you S Z and O as your first piece?

nvm, i was being dumb. ;)