Perfect Clears Conclusion

[memyselfandmeagain]

Hello everybody, I'll try to keep it short:

Was there ever a conclusive answer found to the question: It is possible to play forever and keep on doing perfect clears?

Shuey has done some research in this field and there are some ancient threads by him but I cannot seem to get an answer to the question.

I'm interested in this because tetris is basically 'solved' by playing forever:
http://harddrop.com/wiki/Playing_forever

If one could play forever while doing perfect clears, that means tetris is actually really skill testing. It also means that 4-wideing is an inferior strategy to doing these PC's.

Hope some vetereans manage to find an answer, because I certainly couldn't.

Cheers!

[XaeL]

It was found that there exist some small percent of 10 piece sequences that are NOT pc'able.

However, you could just stack another semi-pc on top and then use the extra pieces to continue PC'ing, using 20 piece PC's

[memyselfandmeagain]

It was found that there exist some small percent of 10 piece sequences that are NOT pc'able.

However, you could just stack another semi-pc on top and then use the extra pieces to continue PC'ing, using 20 piece PC's

Thanks for the reply! I wonder if some players have succesfully adopted the PC strategy in online gameplay. Did Shuey eventually finish the project or was there some bruteforcing search applied to the problem?

[XaeL]

It was found that there exist some small percent of 10 piece sequences that are NOT pc'able.

However, you could just stack another semi-pc on top and then use the extra pieces to continue PC'ing, using 20 piece PC's

Thanks for the reply! I wonder if some players have succesfully adopted the PC strategy in online gameplay. Did Shuey eventually finish the project or was there some bruteforcing search applied to the problem?

Someone wrote a computer program to solve it.
http://harddrop.com/forums/index.php?showtopic=7792

[myndzi]

The thing that makes PCs not-so-viable in competitive play is that they're ruined when you get any garbage. "Playing forever" does indeed regularly get perfect clears, but they aren't every 10 pieces, so it'd be even harder to use that in competitive play. You can't get 4-line perfect clears back to back forever guaranteed, but you can do a pretty good job in a bag randomizer, or stack up to 6 or 8 lines, etc.

There was a bot (http://harddrop.com/forums/index.php?showtopic=5435) on TF for a while that was pretty effective at a PC based skimming/hard dropping PC style, but it's unclear if it's really feasible for a human to do that that well.

[Captivate]

The thing that makes PCs not-so-viable in competitive play is that they're ruined when you get any garbage. "Playing forever" does indeed regularly get perfect clears, but they aren't every 10 pieces, so it'd be even harder to use that in competitive play. You can't get 4-line perfect clears back to back forever guaranteed, but you can do a pretty good job in a bag randomizer, or stack up to 6 or 8 lines, etc.

There was a bot (http://harddrop.com/forums/index.php?showtopic=5435) on TF for a while that was pretty effective at a PC based skimming/hard dropping PC style, but it's unclear if it's really feasible for a human to do that that well.

ahem

P.S. 2 high pcs also exist

[XaeL]

P.S. 2 high pcs also exist

Note: 2 high pc's arent possible in the first bag of the game.

[myndzi]

6 high and so exist as well. I was only making a statement about 4-high, which is provable by counterexample of a blank field + a state that can't be PC'ed. Whether you can avoid ever getting into that state is less clear, particularly if you incorporate 2-high/6-high/etc. It's also potentially possible to avoid all non-PCable sequences with careful hold management if hold is part of the equation; to prove complete untenability with hold, you'd have to provide a sequence for every piece you could have in hold that chains into a sequence with the undesirable piece in hold as the only option(s)