Test your piece placement efficiency in 20G

[jujube]

as a continuation of this debate:


Faster in 20G: ARS or SRS? (http://www.tetrisconcept.net/forum/viewtopic.php?t=708)


i'm proposing that anyone interested try playing lockjaw (http://www.pineight.com/lj) in 20G with entry reset lock delay. the goal is to survive 40 lines with the shortest forced piece placement time, which will be max entry delay + lock delay. the requirements are:


Well width: 10

Well height: 20

Speed curve: Rhythm

Lockdown: Entry reset

Line clear delay: Set by speed curve

Clear gravity: Naive

Garbage: None


to make a true comparison between rotation systems the following options should be kept consistent with rules in games that use the rotation system:


Enter above ceiling

Max entry delay

Randomizer

Hold piece

Initial sideways motion

Initial rotation

Next pieces


though personally i don't see a problem with anyone making these options their own. it doesn't really have to be a comparison between ARS and SRS games, instead it could be an experiment where the only variables are the spawning orientations of the pieces and the kicks.


my shortest (FPPT?) so far is 450ms with SRS and initial rotation, and i was using the rotate left twice key. i forgot to take a screenshot or save the replay though http://www.tetrisconcept.net/forum/images/smilies/icon_redface.gif i'll try and do it again.

[jujube]

http://img81.imageshack.us/img81/7549/fppt450ry7.jpg

[mat]

could you re-explain exactly what it is you would like us to do? i'd love to partake, but i don't fully understand.

[jujube]

pretty much just play on rhythm with entry reset lock delay and see if you can survive 40 lines with as short a delay as you can handle. i'm interested to see how efficiently people can place each piece with forced speed instead of move or step reset. it might give us a clue as to which rotation system lets you play fastest in 20G.


those 5-rotation-over-the-obstacle-under-the-overhang SRS moves won't save anyone now http://www.tetrisconcept.net/forum/images/smilies/icon_lol.gif but maybe there are still ways to survive.

[mat]

so i'm correct in thinking that this is still a speed contest? in that case the results aren't going to be THAT interesting, given the faster players are still going to be on top regardless of their rotation.

[jujube]

true, although it could throw off a player that places a few pieces quickly then does something that isn't optimal. if they're not able to recover from unpreferred placements they might not reach 40 lines. in that respect it could demonstrate the potential of speed in a rotation system rather than what people are capable of when playing in their comfort zone.


so if i'm ever able to play at 400ms per piece forced speed and my 40 line time isn't improving then it's really because i'm not seeing the optimal moves fast enough while manually locking, but they're there.


edit: my only fear is that there may be a large group of players at 400ms. in that case maybe we'll see who can leave the least blocks behind (or in ARS who can keep a clean stack if they upload the replay).

[Taratang]

Maybe I just haven't understood correctly, but I don't see how enforcing a minimum speed for each piece equates to faster play overall if the player is manually locking anyway.

[tepples]

A standard 20G 40 line play, ranked based on the time at completion, optimizes the average case. Entry reset 20G, ranked based on the lock delay setting, optimizes the worst case. Rhythm without modifiers is a compromise between the two: it optimizes the worst case averaged over a five-piece sliding window.

[jujube]

Maybe I just haven't understood correctly, but I don't see how enforcing a minimum speed for each piece equates to faster play overall if the player is manually locking anyway.

maybe we won't ever know for sure which RS allows for fastest sustained play. but one thing this test could determine is whether or not a player will consistently have a good option for placing every piece quickly (or quickly enough). yes, the maximum average speed at which you can place pieces when not affected by the minimum forced speed also matters.


actually i thought this would be a good opportunity for ARS 20G players because usually they can't compete in a 40 line mode when the only thing that matters is your time. caffeine's lock delay challenge was interesting but to win you had to use a system with move reset.

[jujube]

Cleared 40 lines at level 60
on 2007-10-24 at 00:11
(rot=SRS, lock=400 ms Entry reset, rnd=7-piece Bag, spin=3-corner T, speed=Rhythm)
(grav=Naive, drop=None, are=0 ms, das=16 ms 1G (60 Hz), shadow=Faint color, next=8 pieces)

Played for 0:52.65 (active 0:47.08)
Played 113 tetrominoes (128.77 per minute,
144.00 per active minute)
Pressed 191 keys (1.69 per tetromino)

Made 40 lines
(single: 27; double: 4; triple: 1; home run: 1; T single: 2; T double: 0; T triple: 0)
Sent 10 garbage (11.39 per minute)
Left 52 blocks behind

LJ score: 6000

edit:my only fear is that there may be a large group of players at 400ms. in that case maybe we'll see who can leave the least blocks behind (or in ARS who can keep a clean stack if they upload the replay).

hmm i didn't realize Lockjaw 0.38 lets you adjust the lock delay in smaller increments, so scratch that.