Tips for avoiding the castle effect?

I've gotten a handful of GMs now, the most recent of which clocked in at 12:31:45 & felt like a pretty even game for me. I have some amount of room to improve on locking time (I could probably just barely get to sub-12 without changing my stacking on a similar run that just went frenetic), but mostly I need to improve my stacking.

At a certain level, I've just gotten good enough at recovering my mistakes & fast enough at inputting to compensate for my stacking technique. Now there are actual misdrops here & there, but for the most part I put myself in sticky situations that cause minor-to-moderate mistakes. The most common of these is creating a castle-like stack, with a bunch of 1-deep holes across most or all of my stack.

Does anyone have any good tricks for getting out of these situations or any advice on what to avoid when planning piece placements? I'm pretty mindful of trying to leave at least 2-wide gaps whenever I have some choice in the matter. (And of course it's possible I'll just have to learn to feel this one out better.)

In terms of recovery from the castle effect, I'd say I'm overly hesitant about stacking over a single hole in the middle, and now that I'm actually thinking it out, I guess I'm answering my own question. Any good examples or tips about that though? Do you usually try to cover up the middle of the stack, or do you start from the sides if it's the first half of the game?

 

Whether you are happy to create a single hole or not is very situational. It depends on a few things, for example:

1. What column the hole is in - central columns are bad (4-6), if it's nearer the sides then it'll typically be easier to fix as the game mechanics won't pressure you into stacking on top of it so much
   
2. How competent you are at actually digging holes back out - some players are able to correct holes elegantly and with minimal fucking about, others not so much. Obviously if you're the former, then the likely penalty of making a hole is lower because you'll have less trouble fixing it later.
3. Where you are in the run - this is maybe a bit complicated and different players might have a different opinion on this. I'd say for TGM1 where there's no sonic drop, I typically prefer to make a hole early because overhangs cost time (will cover this later). For the middle section of the game (300-700) I'll avoid them if I can, because there's too far for me to go to make holes and not want to fix them, and obviously that costs time. That said I'd consider myself to be pretty good at fixing holes so I don't stress too much over them. Late game if holes have to happen then I'll typically just leave them be, given that there's not long left to go, and I'll hope they don't add up too much to leave me a full stack and wasted time at 999.

For avoiding the castle effect, I'd say specifically to try and keep your central columns high a bit more. Firstly this breaks up your stack a little and means you won't ever get stuck with a completely castellated playing field. Secondly, if your central columns are higher it makes it easier to set up and fuck around with overhangs to avoid holes in the first place. Being good with overhangs takes a bit of practice, but if your middle is high then frankly you can just plonk pieces up and then shove others underneath later if you get how wallkicks work. Case in point:
(gif made by PetitPrince from a sequence in one of my TAP Master runs)

A lot of emphasis gets put on keeping the middle high because it gives you better piece mobility and more lock delay reset, but it also really helps buy you some extra room for overhangs if your stack and piece sequence are particularly awkward. Small one-row overhangs can be quick and elegant to make and fill, but bigger overhangs provide you with a lot more options in terms of possible piece combinations as to how you can actually fill them. For example a 1x1 overhang can typically only be filled by a T, J or L (or an I if you've got space), whilst a larger overhang is a lot less piece dependent.

Again though, overhang versus hole is still situational. In particular the randomiser can completely fuck you if you spend too long waiting for the perfect piece to fill an overhang and it doesn't come, as you'll completely throw your stack out of shape trying to fit things around the overhang while you wait.


If you're focused on time-attack then a hole is also only worth fixing if you can do it quickly and efficiently - otherwise just leave it be and wait for another opportunity to fix it. It's really easy on TGM1 to lose 20-30 seconds of time fucking around trying to unpick a hole from your stack and clearing singles and doubles in the process (as I know from experience). There's a skill to not creating them, but there's also a skill in knowing when to pick your battles once they do happen.

The main thing I would suggest with fixing holes once they do happen is to be patient about it. Don't concentrate too much on trying to sort it, and in particular don't insist on fixing it straight away. Typically if you've had to make a hole your stack isn't in brilliant shape, so sort that out first. Once you've got the stack neat and stable again, then work on digging it out. For TGM1, I'll typically try and stack over it and then skim, like this or this. It's important to not stack over the hole any more than you need to, but at the same time, again, don't focus too much on avoiding it. There's no point wasting time clearing a hole, only to leave yourself with a messed up stack by the end of it that's going to force you to make another one fairly soon.

Also if you've somehow not come across them yet, the TGM guide on the wiki and the translated jp TGM guide have some pointers that can be helpful for things like this.

 

Yeah, now that you mention it, I guess this is usually a problem in the first half of the game, since I stack pretty flat until about 350 or even 400. I suppose even as early as 250 I could starting thinking about overhangs to keep placement options open without a flat stack, and it makes the castle effect a moot point.


In the rare cases where it does happen in 20G, I guess you'd want to avoid covering over the middle, if possible, since it's much harder to recover a hole there.

 

As a GM you're probably experienced in this, but I've noticed in my own games it's something I tend to be bad at - Basically, having a clean surface on your stack is usually more important than having a solid one, especially if you're good at effecient downstacking (which I think is essential in 20G).
What I tend to do is end up with a "castle effect", and then get pieces that are crappy in the situation, such as an O piece. Usually any other piece would have been good, but what I do is place the O in one of the sides to avoid overstacking near the center, and while that is probably the best course of action in the situation, it doesn't improve it at all. In fact, it only makes things even worse, as straightening out the stack could easily take a while (or until you get a T piece), and result in an excessive amount of superflous downstacking before I get back to my "tetris hole".
So the easy solution to this is simply that I need to improve my stacking similar to how Rosti describes so I don't even end up in situations like this, but I figure an effecient half assed solution would be to overstack earlier on, in a way that helps straighten out the stack as fast as possible, so I can dig myself out of the trouble much faster.

 

My best tip for avoiding castles is to place more pieces horizontally, especially S and Z.

 

A high center stack is the best strategy, as eloquently stated by Rosti. In addition to that, you can also try to have a Tetris hole that looks more like a funnel rather than a straight hole i.e 1-wide hole at the bottom, 2-wide or more at the top. That way you can throw anything you don't like (i.e. pieces that make you create) in that funnel. The overall shape of the stack doesn't change, and in the worst case scenario you create a bridge that's easily cleared.

That said, be wary of making too much single; that could slow your game down.

 

crenelations man... i know i dont play the same game as you guys but it nice to see that other people deal with that shit, too. those are my least favorite shape EVER.

 

^ This.
When I play I try to always place an S or Z horizontally if there's a place for it to go. I'll only usually place them vertically if there's no place for a horizontal, or if placing it horizontally would impact the placement of the next piece.
Always make holes toward the sides if the worst happens, and I cannot stress how important this is in 20G. If there are holes in the center then a lot of pieces will just get stuck there as soon as they spawn if you're not careful, more often than not making more holes in the center.

 

I've already made dramatic improvements on this in under a week! Mostly due to stacking high in the center or up to the left in the early game & then center high from 300 on. Technically I haven't improved my GM time yet, but I've essentially tied it again.