Some additional nuances in strategy from TTO2

[larrytetris]

So just a few additional things I noticed from TTO2 matches that I wanted to post here as much to share as for me to remember and browse later. It's a little bit more in depth than appropriate for the stream, or to fit into he 25 seconds between each match, so it wasn't commentated on. I'll add to this thread if more come up in the rest of the playoffs.

From Blink vs. Aitsu -
1) A lot of times the "swiss cheese garbage" would appear seemingly out of nowhere, not after a short downstack combo or anything. It took me a few games to realize it was because whenever Blink spiked Aitsu to about 15ish rows up, Aitsu would often give up b2b in favor of downstacking. He would clear a single to reposition himself, and then set up another TSD, or something along those lines.

This meant that Blink, if he continued stacking upwards to maintain b2b even if it's riskier (which did end up costing him some games), he would be responding to Aitsu's 4-garbage attacks with his 5-garbage attacks. Net: 1 garbage line to Aitsu.

As aitsu started receiving more of these single lines, he would need to cancel b2b more often, and this created more 4 vs 5 garbage scenarios, and it kind of snowballed until one of them misdropped or got a lucky downstack combo.

I don't know yet how to take advantage of this and make a definitive, decisive strategy out of it yet, but I'm chewing on it.

2) There were two instances (that I can remember) that the players used combos. Once it was when aitsu misdropped, he stacked a 2w upwards, and Blink saw this and responded with a 3w above a TSD. The other was a hybrid combo from aitsu. In both instances, the combos were midgame.

The combos made a very, very distinct change in pace in the game. Once one player noticed the other was comboing, they'd generally play considerably more conservatively or more aggressively.

Every other game was decided more or less by who misdropped less; that's why most of the games went into late game. However, I think if either player took the initiative to execute midgame combos (or even openings), the change in pace would be in their favor because they're the ones in the driver seat as far as pacing goes, and your opponent would be the one on their toes.

Midgame non-downstack combos (and other "riskier" setups) work best when both players are at the floor of their matrix. Once you reach the floor of your matrix, you've drastically reduced your output potential. This held true in KoS, and I'm definitely noticing it more in realtime games.

If you're at the bottom of your matrix, 100% of your output is user-generated, meaning you need to stack for the attacks. In the meantime, your enemy will ideally trying to be reach 10%, because the ideal playstyle when you're near the top is to only clear garbage lines, which is 90% enemy-generated (9 minos out of every 10 in a row).

Before my match against arfarf, I did a little training exercise on Nullpo against the AI. I wanted to see how well I could respond to enemy garbage and use it against them. This was my best outcome:
[!--ImageUrlBegin--][a href=\\\"http://i49.tinypic.com/a2tb4p.png\\\" target=\\\"_new\\\"][!--ImageUrlEBegin--][img width=\\\"400\\\" class=\\\"attach\\\" src=\\\"http://i49.tinypic.com/a2tb4p.png\\\" border=\\\'0\\\' alt=\\\"IPB Image\\\" /][!--ImageUrlEnd--][/a][!--ImageUrlEEnd--]


That's more than one attack per piece. That is INSANE efficiency. I realized the other day that attack per line is not necessarily a metric of efficiency in the strictest sense because sending a tetris with a 4-tall garbage column is more efficient than setting up a TSD by yourself, even though the TSD has higher APL.

Conclusively, though it's counterintuitive, it's actually safer for you to do riskier things when both players are near the bottom of the field rather than when your opponent is at the top of the field. So as far as as strategy goes, set up a combo or a chained tspin setup when both players are not entirely at risk of topping out. You might sacrifice the chance of an immediate spike, but it'll be better in the long run.

[myndzi]

Interesting observations... I've used the 5/4 thing to intentionally send people "spike or messy garbage" on KoS, but it's interesting to see how it affects the longer game at high speed like that. With players actively timing their canceling for most of the game it really does present an interesting tactic.

As for the combo thing, I was recently rambling to caffeine along a similar line. It seems that after a certain skill level, players can pretty much stay alive indefinitely (pending large apm deficits, severe misdrops, or randomizer hate). There's a larger window of apm difference allowed by timing skill, too. I was thinking that constant pressure is maybe not that beneficial, since it seems that you need an overwhelming spike to really win under "usual" circumstances, and that usually involves starting from a lot of garbage. Of course, to get a lot of garbage to send you can always send a lot too  I guess the see-saw nature of CoA garbage keeps things lively until you can seize an opportunity.

I wonder if a person could perfect a strategy that involves sending nothing until most of the way up the screen and then downstacking it all at once? The opponent wouldn't have much time to analyze the downstack since it wouldn't be there, and they wouldn't have as much APM since they have no garbage to use. Not much to use for timing defense either. You could get like 3 tetrises + 3 t-spins on top of 3 b2b holes, which is a good 30 lines of garbage...

Combo stacking is probably an appropriate counter to that, but it could be an interesting tool in the toolbox if it was plausible

[Agamemnon]

On no1, Larry, you've noticed what I realized half a year ago. Watching hebo vs Blink and qmk on nullpo, I saw that he used their spikes to downstack with smaller sends, creating hellish nasty garbage to stack through for the opponents. I intentionally set out to learn this strategy myself, and it has made me able to beat people I would struggle greatly to defeat if I was to play with my old standard glass-cannon style... The strategy is much better for other guideline games, as the LCD on TF handicaps you when you try things that don't consist of chained alphas....

[Jupiter_Joe]

I'll add just 2 things, (which cannot be helped when the bag will not allow):

1.) Only do an EZ spin if you have a B2B attack ready (typically a tetris)- otherwise wasted soft drop.

2.) Try to do a tspin before you do a tetris due to slow soft drop.

There is general, but not quite total, compliance with this TF centric advice.

[Paul676]

Added to most useful threads

[massi4h]

I wanna discuss the power of middle 4 wide. I feel like people aren't using it as well as they could be. People are saying that it can be countered by a fast 3 wide, but what happens when the 4 wider takes the first 15 or so lines from the 3 wider before starting their combo. One of the main reasons for stacking 4 wide in the center is so you don't die until your centre reaches line 20 meaning you can receive so much of that garbage (and often be able to use it in your combo). I see too many people see a big red line and start their combo to cancel it out then they lose heaps of output.

I personally am not confident enough in my 4 wide to pull it off like this consistently, but I feel the better 4 widers should take more advantage of this.

[myndzi]

I think it's probably difficult for most people to gracefully continue to accept garbage when they can't stack on the sides. It's something that would need practice. Add to that the chance of breaking combo in the middle of the 4-wide, and the chance of instant KO if you're not precise with the amount of garbage you receive, and it's a bit riskier. You'll want flat, even tops on your sides before they go into the hidden area too.

I don't think any of this is beyond pulling off, but it is a bit trickier than otherwise...

[MarioThePhenom]

[fumen]110@HekzgbKwjbQplz1lIwibRpsskz1lYiibQpusaiib/e?pbAoUqAtu6nDs488AQ+TXEl488AwSGoDloo2ApN98AQ+7dD?0oo2A5H5TAT5AAAHcC3ibG3ibG3ibG3ibG3ibG3ibG3ibG3?ibG3ibG3ibG3ibG3ibG3ibB3gbA3lbA3KcK6NAA7eD+OA3N?AdLARcppAzHnrDU91JEnoo2AseirDFbEcEoe0TASo0JEjS0?rDFbMLEmoo2ApAAAA[/fumen]
ive been practicing this three wide method quite a bit, and its like the 4w in the fact that you dont have to be the one to send the first lines. the blocks follow something like that pattern up until the top, where i start to break up the pattern. its kindof due to the I block spawning in the stack on the left side, but i think that can be avoided with irs and some planning ahead

[StS]

Mario, building upon that, wouldn't you end up with a sort of 3-4-wide hybrid? Also, that wouldn't work if you plan on taking garbage to send back to your opponent, because your I would still top you out if your stack is pushed up too high. Initial rotate would work, though, so still possible to pull something like this off.

[MarioThePhenom]

Mario, building upon that, wouldn't you end up with a sort of 3-4-wide hybrid? Also, that wouldn't work if you plan on taking garbage to send back to your opponent, because your I would still top you out if your stack is pushed up too high. Initial rotate would work, though, so still possible to pull something like this off.

no, it remains a three wide until the 18ty row, where i might make it 4 cells rather then3. you can still take alot of garbage given your next few pieces have no I in them. i still need to practice this more because typically ive been doing it in ffa, and rarely in 1v1's. and id also like to get some videos of me preforming it to compare and see if its really as viable as i hope it to be. i would do it tonight but im rather busy :/

[myndzi]

Koreanyama is known to successfully pull this off, isn't it?

[Paul676]

myndzi: Yes - the issue is that you can't take garbage with every piece with a middle 3 wide like you can with a middle 4 wide.

[larrytetris]

what happens when the 4 wider takes the first 15 or so lines from the 3 wider before starting their combo. One of the main reasons for stacking 4 wide in the center is so you don't die until your centre reaches line 20 meaning you can receive so much of that garbage (and often be able to use it in your combo). I see too many people see a big red line and start their combo to cancel it out then they lose heaps of output.

This is something I've taken into account a little (granted, not as well I should for sure). I think you're overgeneralizing the situation. There are definitely situations where stacking higher at the expense of an even top is not worth the lines you can keep from cancelling because with a flat stack you could potentially reach garbage regardless.

More importantly, it keeps you in a safer position in case the 4-wide needs to be broken.

[vipjun]

I wanna discuss the power of middle 4 wide. I feel like people aren't using it as well as they could be. People are saying that it can be countered by a fast 3 wide, but what happens when the 4 wider takes the first 15 or so lines from the 3 wider before starting their combo. One of the main reasons for stacking 4 wide in the center is so you don't die until your centre reaches line 20 meaning you can receive so much of that garbage (and often be able to use it in your combo). I see too many people see a big red line and start their combo to cancel it out then they lose heaps of output.

I personally am not confident enough in my 4 wide to pull it off like this consistently, but I feel the better 4 widers should take more advantage of this.

This is something I've taken into account a little (granted, not as well I should for sure). I think you're overgeneralizing the situation. There are definitely situations where stacking higher at the expense of an even top is not worth the lines you can keep from cancelling because with a flat stack you could potentially reach garbage regardless.

More importantly, it keeps you in a safer position in case the 4-wide needs to be broken.

I've seen this strategy used in 4w vs 4w battles.  The premise is that if both players do a full combo off their stack the person who started their 4w later would win because:
1) by starting later they usually have a higher stack and able to send more with the full combo
2) by taking some lines initially they need less of a stack advantage to top out the opponent.

This is pretty much an all in strategy, because if you can't continue the combo there's no other place to put your block except right in the middle of your field which makes recovery even tougher.  Also, since your initial combo starts so much later, your garbage meter will likely be full enough to KO if you stop clearing. The better both players are at comboing the more likely this strategy would pay off I my opinion.

vs fast 3w.
I think this strategy is overlly risky vs a double 3wide because:
1) If you can fully combo against 3w you'll probably top them out regardless if you waited to accept the 15 lines or starting the combo whenever.  If you failed then you die difference is that 3w pulls off their combos pretty consistently and faster because they don't need to softdrop any pieces.

2) By waiting to accept 15 lines, it allows them to build a higher 2nd 3w stack before you can respond because of the line clear delay.  The faster both players are the more this plays into effect.

[Paul676]

When I was doing side 4 wides against a middle 4 wider, I would stop the stacking at about 13-14 rows up, and combo it down really quick. This would neutralise their 4 wide because my base speed is usually quicker than theirs, and the fact that my 4 wide is side whilst theirs is middle meant it was even quicker than that. So they'd be forced to either top out quickly or receive enough garbage to put them at a disadvantage if they wanted to stack higher than me, or to cancel down quickly, and they receive some lines. The advantage of middle 4 wide over side 4 wide, as long as the side 4 wide keeps his wits about him in not accepting loads of garbage only exists if middle 4 wide can get to the 19th/20th column.