Saturday 11 November 2017

The Long Combo opening

        

A "combo" in Puyo Puyo Tetris is a series of line breaks without an interruption, clearing at least one line with each tetromino.  In Tetris vs. Tetris, every hit of the combo from the 2 combo onward does damage, increasing every two drops. So hit N will do N/2 damage, rounded down, e.g. hits 10 and 11 of a combo will do 5 lines of damage each. This quickly racks up--doing an 11 combo will do a total of 30 damage over the course of the combo, which is more than enough to kill someone. Naturally, there are openings which exploit this.

The standard Long Combo builds 3xN columns on the left and right, and reserves three extra squares of material in the middle. These make it easy to continue clearing. Each clear will place a tetromino--four squares--and clear a line. The next empty line has 4 empty squares in it, so voila, we still have three squares to work with. With some clever use of hold piece, one can often keep combo going until one runs out of the two columns. If, in the meantime, we were sent garbage, we can sometimes keep the combo going beyond this by clearing garbage lines. This is easier if we kept the side columns even in height, so we're not blocked from breaking garbage under the higher one.

Let's look at how we build this.


1. Setting it up

    

This is the most popular variant. Its similarity to the Ace setup keeps our opponent guessing as long as possible, as well as allowing us to do the Ace instead if we choose. If we get the S first, we do the left-hand variant, if the Z, the right-hand.

We can stack up the 3x4 stack in either LOJ or JOL order.

Now, we need to build up. Tetris is a deck-based random game, so it's a question of building two 3xN stacks using one each of L, O, J, I, S, T, and Z. Well, LOJ is pretty easy:



Yeah, we can just stack another 3x4. For S, T, and Z, we may hope to be able to play the T piece in the middle. Then given STZ, we can just play on the right of the right column, or ZTS, we can play on the left. Generally, we can just use the T to shift from placing Zs to Ss and back.

    

The I, of course, gets its own column:



If we had but enough space, and the deck kept coming in the right order, we could keep this up forever. The discerning eye may notice a slight problem: the column where we placed 4 pieces is growing faster than the one where we placed three pieces. Ideally, we want them to grow at the same speed. How could we do this? Well, one answer is the good old Ariana Grande song. We need to go from side to side.



We've gotten the L, I, and Z in the next deck. We still have the O, J, T, and S to play. If we were keeping to the pattern set above, LOJ on the left and IZTS on the right, the right would get even higher. But observe, we have precisely the same shape on top of each column. So, we could just... put the rest of the LOJ on the right, and the IZTS on the left, and suddenly the left column is growing faster and catching up.

    

Here I've also put the S piece sideways. This allows the I column, which will now be on the left side of the left column, to catch up to the STZ pair of columns.

At this point, we may be improvising hard in response to the shuffling of each deck. Some other formations may be useful to keep in mind:

            
Other 3x4 rectangles.

                
Ways of handling a naked O.

Eventually, we should stop building our combo and start breaking. The ideal point is with both columns at the same height, with one square missing from one of them. This can be done if the column height is 7, 11, 15, etc. At height 8, 12, or 16 you'll end up with an extra square on top of one of them, which is also not too bad. You'll have to improvise this based on your particular pieces.




2. Executing the combo

    

We have a series of pieces, and we can swap them out with the hold piece mechanic. A computer could quickly calculate a way to keep the combo going through the next 5 pieces, but we need to think fast, and can't calculate everything. What we need is some rules of thumb:

  • Look at least a couple pieces ahead, and figure out how you're making them fit. Take special care to look at placing the S and Z pieces and what comes next, as these can often be combo-ending trouble.
  • Prefer to hold the I, T, or O if you're not planning something specific.
    • The I placed horizontally clears a line without changing the board. This essentially allows you to skip a piece you can't otherwise handle by swapping it into your hold.
    • The T is generally playable on pretty much any board, and can get one out of sticky situations. It's also nice for setting up playing the S and Z, e.g. from the "handle" board (see below.)
    • While the O is not playable on every board, it's great for getting out of sticky situations, e.g. after playing an upright S or Z on a handle hole.
  • Looking ahead, try to end up at good, flexible boards which handle lots of pieces well.
What boards should you shoot for?

            

Above are the "best" boards. On the left, the pan board--what we're starting out with. On the right, the handle board. The pan does well with all pieces except L (for left) or J (for right), and the handle with all except S, Z, and O. For convention, I'll call their orientations based on where the hole is: the pan pictured left is the left-hand pan, handle pictured right is the right-hand handle.


            

Above are situationally good boards: the funnel and the corner. Various pieces will take you back to pan, handle, funnel, or corner from here.


            

Above are some boards which you definitely need an exit plan for: the molehill, the W, the pit, the T-hole. Get into these if you can see a way out to get out to a good board with the next few pieces.

It's useful to bear in mind a few sequences that may come up.

        

Above, we start at the left handle, play TZSO, and end up at the right handle. Wow, we handled all the problem pieces in our deck.


        

Above, we start right handle, play LJSZ, and end up right back at the right handle. We've gotten rid of all the pieces that aren't "ooh, this is a good piece to hold" and stayed in a good board.

Some interesting problems, handled:

            

Here, we start at the left pan with a ZS. You might think that the Z only generates the left handle, so we're screwed regarding the S, but with an S spin, we can end up at the left handle.


    

Here, we've gotten ourselves into a sticky situation: the left handle with ZJ to come. This gets us into the molehill, and we're not holding an O--the easiest way out of the molehill. But we get out via a specific sequence. After ZJL we're in the funnel, and the T is just one of the many ways back to the handle.


    

Here, we have the left pan with S. This gets us into an exit-strategy-only position: the T-hole. Here's the possible exit strategies.


        

We get back to safety using only either T, or LJ, or (unlikely) LL.

Of course, with SJ, we could've skipped the T-hole and just done a J-spin:

    


3. I say jump. You say how high?

How high should you build the sides before executing? A deep philosophical question, posed in the great Rage Against the Machine song, Bullet in the Head.

We should note here why the hole is in the center. It turns out that in Tetris, only the center four columns will kill you. The others can rise however high, and you can keep breaking. So if your opponent does less than 18 lines of damage before you start breaking, you'll still be alive to hit them back. If you were building the hole on the left to height 11, an attack of 9 lines would instantly kill you.

It should also be noted that the combo counter in Tetris starts at zero. So, 12 lines of break will do an 11 combo for 30 overall damage. We also see above that the cleanest places to end our build are at 11, 12, 15, 16, and 19. Here's some rules of thumb, and their justification:

  1. Against a T-spinner, plan to go to 11 or 12, and start breaking once you have an even top and see a good route through the next few pieces. Why? By the time you've built to 12, you've used up 19 blocks, or almost 3 decks. The most damage they could have done to you with the same number of blocks without an all-clear is 2*T-spin double + T-spin triple = 4+5+7 = 16. so you'll survive to hit back if you get started right away. And building past 11 once you've gotten hit by a couple T-spin doubles for 9 damage isn't something to count on.
  2. In Long Combo vs. Long Combo, you want to be slightly greedier than your opponent. If he does a 10 combo and you do a 12 combo, you end up doing 11 more lines of damage. Not bad. But don't plan on going much past 15/16. Why? Suppose you could drop 3 more pieces and break exactly as good a formation at 17 as your opponent breaks at 15. You'll be 3 pieces behind, so each of his combo breaks will alternately do 1 and 2 more damage than yours. At equal speeds, over the course of his 14 combo, you'll take about twenty lines of damage, before your extra lines can pay off. So you'll probably die.
  3. Against the Ace, you'll probably only be able to go to 7 or 8. Just do it, and hope your garbage line clears give you some extra.

4. Handling garbage in the hole

            

While you're building a long combo, you may take damage. If the damage isn't oriented under your center hole, you can just clear it once you've cleared your sides--having nice, even sides helps for this and allows you to extend your combo.

If it's in the middle, though...



This is the best scenario for damage within the hole. We have exactly 4 lines of damage oriented underneath. If we clear all the relevant damage lines, we still have three squares to work with in the hole. Better yet, we can easily start out or in-the-middle our combo with either a T-spin or a Tetris and kill them a bit earlier, and having a vertical hole to clear in that location may make it easier to extend the combo at certain points. If we wish to ignore the garbage entirely, well, we don't often drop pieces that would fall into the extra hole in the middle, working with the pan and the handle. Only the inverted J and L really see a difference here with having an extra-deep hole and they're typically dropped on the side into the handle.

Wow! We covered pretty much all the logic looking at why that example was so good for us. Let's go to some rules of thumb and apply them.

  • If our in-hole garbage is exactly 4 lines, or a multiple thereof, clear it whenever we choose. If in the center, hope to get a T-spin out of it; if a column, hope to get a Tetris. Plan to finish clearing a set of 4 within a few pieces of starting.
  • If our in-hole garbage is less than 4 lines, ignore it as long as possible. Every line of it we clear leaves us one fewer squares of material to work with and makes the combo harder to keep going. Try to avoid clearing it until we're a few rows from the top of the hole.
  • If our garbage hole is in the center (columns 5 or 6) and we're determined to ignore it, mostly don't worry, be happy, and play our usual pan-handle game.
  • If our garbage hole is at the side (columns 4 or 7) and we're determined to ignore it, try to avoid getting into the handle at that side with the relevant JL piece up. For example, don't end up at left handle with the garbage hole at 4 and a J to come. We can do this by specifically going for pieces where we use a facedown J/L instead of inverted.

Let's apply these rules of thumbs to the above boards.

    

We're golden. T-spin or Tetris whenever we get the opportunity as we combo. Sequence: T.O.SLITSL, opponent died. Note the Tetris followed by the T-spin double for an extra 9 lines in the middle.


    

Ignore this until we're almost done with the combo. We should be able to get away with it, as we see in the example fill. We mostly ignored it, then got a T-spin at the end for some extra lines. Not bad. Sequence: JSZSIOJTZLT for 10 combo.


    

We need to work around this. Here we did our best not to ever need to drop an inverted L in 7.  Sequence: TLTSZOJIOLJTI for 12 combo. How did that work? weeelll, in this case we used L to move from left handle to right handle, and held T/O as appropriate, planning to use the TSZO sequence to cleanly get us back to the left handle. Thus, we never needed the inverted L, and we just willingly cleared the garbage at the end.


5. If your combo breaks...

You may, unwillingly, end up ending your combo early. I suggest going as far as humanly possible. If you're at 6 combo, stopping your combo earlier to set up a T-spin or Tetris is not worth it, since one more hit of the combo will do another 3 lines of damage, and who knows, you might figure out another hit after that.

Your choices here are:

  • Build up the sides some more and restart the combo. Only do this if you're not near the top with garbage, your sides are clean enough to build upon, and you're not at a really awkward place in your deck. Otherwise, you might go nowhere but death.
  • Just keep comboing, starting again from 0. Do this if you have enough side material for the combo to do worthwhile damage. 7+ rows is a good rule of thumb here.
  • Build in the remaining space in your hole to set up T-spins/Tetrises, and go into a normal game. I generally suggest this if the first two do not apply.


Conclusion

Given sufficient time to build, we can do devastating damage, and the Ace/Long Combo mixup means we can keep them guessing throughout our first deck whether we will do this. But getting it done requires practice, thinking ahead, and some agility with various types of spins. In my next post, I'll look at the L, J, S, Z, and T-spins we might need.

1 comment:

  1. The Casino Hotel Las Vegas, NV Map & Directions
    Las Vegas 세종특별자치 출장마사지 (NV) location, map 포천 출장안마 and elevation. 3350 Las Vegas Blvd S Las Vegas, NV 상주 출장샵 89109. Realtime driving 김제 출장샵 directions to The Casino Hotel 문경 출장샵 Las Vegas,

    ReplyDelete