Evil Randomizer Tetris post game. [Archive]

Zaphod77

03-26-2008, 10:27 PM

The rules are simple.

1) You post a piece, and the next person places it. State columns used, orientation, and, when not obvious, how you got the piece there.(e.g. IRS Clockwise, DAS right, rotate clockwise)

2) you must avoid the last 4 pieces in the history

3) the history starts at SZSZ.

4) you may not start with a S,Z, or O.

5) If the last 9 pieces have not been an I, you must give yourself an I, place it reasonably, and then hand a piece to the next player.

There is no hold, and no preview.

ARS. TAP style.

Initial rotation system.

And, to make it interesting...

This game is 20G. If the piece placement is not possible in 20G, it is not allowed! This is why you have to explain non obvious placements. There are no floorkicks, and no I kicks. Just like TAP. Be glad lock delay is infinite. http://www.tetrisconcept.net/forum/images/smilies/icon_smile.gif

Do you best to make pieces as difficult to place as possible, and when placing pieces do your best to place them well. Don't try to throw the game. http://www.tetrisconcept.net/forum/images/smilies/icon_wink.gif

I'll start.

I http://www.tetrisconcept.net/forum/images/smilies/icon_smile.gif

Ghett0

03-26-2008, 10:38 PM

|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|� � � � � � � � �|
|I� � � � � � � � |
|I� � � � � � � � |
|I� � � � � � � � |
|I� � � � � � � � |
___________________

Z

Zaphod77

03-26-2008, 11:02 PM

The history starts SZSZ

current history is now

ZSZI

S,Z,I are forbidden.

new piece please.

Edo

03-26-2008, 11:09 PM

It might be best to have everyone agree on the rules before rushing in and starting.

For the benefit of people who weren't there at the time, this is an idea that Zaphod and I were playing around with in #arika. Zaphod was calling out a sequence of pieces, one at a time, and I had to place them, assuming 20G. Obviously, I didn't have the benefit of a preview, and I was trying as much as possible to play the first placement I could think of, so as to simulate some sort of time restriction.

(For the curious among you, here is what I ended up with: MASSIVE LINK (http://fumen.zui.jp/?m105_40bdA3nbD3jbG3ibG3ibH3hbH3hbH3gbI3gbI3qbAAA7 ?e5jzAJQBH9AOBBspA13AG_2FAvPBr+AaKBspAd4A+0Av6Aa7A ?rlApe070482E1FD). It starts with garbage in the well because I was initially trying to give Zaphod an example of something else, and it later evolved into our little game.)

I think this could work quite nicely if we have an independent adjudicator giving out the piece sequence, one piece per day. Anybody that wanted to participate could make a post, and update it each day with a link to an updated fumen sequence. We could run the thing for perhaps, say, 2 months (60 pieces), then at the end we would need some way of scoring each participant's sequence.

For scoring, a penalty point system might work, perhaps penalizing complicated moves or messy play: for example, if a move requires the use of a double-tap without a lock reset, 10 penalty points are awarded; if a hole is created in one of the central 2 columns, 5 penalty points are awarded, and so on.

We'd have to agree on a good system, but I think this might work out as a lot of fun for all the forum regulars.

Edit: Right now I'm hoping no one noticed the deliberate mistake in my sequence...

Zaphod77

03-26-2008, 11:12 PM

and yes i think this wold make a great LJ gimmick for stack training.

worst fit, but obey the randomizer restrictions.

tepples

03-26-2008, 11:17 PM

and yes i think this wold make a great LJ gimmick for stack training.

worst fit, but obey the randomizer restrictions.

Are you voting for bug 0048: Bastet randomizer?

Ghett0

03-26-2008, 11:20 PM

Then I guess I'll choose O.

Pineapple

03-26-2008, 11:23 PM

I have some comments about the rules... long post coming. Small nitpick: in TAP, the Z is removed from the history first, not the S.

|I� � � � �|
|I� � � � �|
|I� � � OO |
|I� � � OO |
+==========+
History: OISZ

Next: T

Food for thought: We need to have a rule about not being able to play until a few others have.

Zaphod77

03-26-2008, 11:35 PM

Sort of, excep that it obeys the rule of other randomisers.

so if we had "evil bag" then it would use the worst fit algo, but otherwise obey the rules of the bag. (hand out the seven before repeating, but try to hand out the worst of the avilable ones).

and evil ARS would NEVER repeat history, and would never go for 10 pieces without an I, but would otherwise try and hand out the worst piece.

The idea is it only gives out piece sequences that are theoretically possible in the original game it was based off of.

and if you can deal with the really nasty but possible sequences it hands out, you will learn to stack better.

The I restriction i put in was chosen out of a hat to keep people from getting starved for Is for the entire game, which is theoretically possible under the arika randomizer (but very unlikely).

This whole thing started because he was giving me stacking advice for TAP, and i commented that i disagreed with his O placement, because the stack was less stable in my opinion with it there, and a z could be placed more cleanly if the O was where I would put it instead. and he said to keep handing out pieces and he'd place them. and so away it went. and i decided i'd be fair and not do stuff the arcade wouldn't. It was pretty easy to mentally do the history avoid, as it was right there in the window. but i starved him for 12 without an I before finally giving in, and almost that many for the second I. I managed to be pretty evil without being able to keep up with where exactly he was placing them.

If I had the field in front of me, i could have been even more evil, waiting until he blocked the tetris hole before handing out the I.

so, yeah. here we go.

the history actually started ZSZS, not SZSZ.

so it's SIOT

with the T waiting to be placed.

Zaphod77

03-26-2008, 11:58 PM

I think this could work quite nicely if we have an independent adjudicator giving out the piece sequence, one piece per day.

Nice thought, but each person needs to get their own hand picked worst piece depending on how they placed the previous ones. if every player got the same piece sequence, then someone can come up with a way to stack it cleanly.

Ghett0

03-27-2008, 12:07 AM

Pineapple

03-27-2008, 12:22 AM

and evil ARS would NEVER repeat history, and would never go for 10 pieces without an I, but would otherwise try and hand out the worst piece.

The idea is it only gives out piece sequences that are theoretically possible in the original game it was based off of.

The unreasonable drought rule is a nice idea, but I would apply it to all pieces, as opposed to just the I. I would also increase maximum allowed distance. A distance of more than 10 (I, 9 other pieces, I) has a 9.8% chance of happening. distance of 12 (I, 11, I) is 4.6%

When you are placing a piece, it's realistic that you would know the next piece as well. However, I can see why you dropped this, as you are trying to encourage difficult piece sequences. It would also take a lot more bookkeeping, and heavy bookkeeping is bad for forum games.

Forcing to avoid the history makes the sequences very slightly easier than TAP, I think. The chance for it to resort to history is 0.6%, so you would expect something like 1 history repeat every 2 sections. Again, I don't think that this can be represented without heavy bookkeeping, so disallowing history repeats is a good idea.

As for the piece placement, I think that TAP-ARS 20G is a suitable choice. However, I think that you should also disallow placement that require exact timing. jago has a page in his webspace explaining the various possible moves (http://jagorochi.free.fr/absolute/index.php?title=ARS_Synchro), and of these, the blocked synchro should be allowed as its a very consistent technique once you've practiced it, but the others (just frame, fling, and wallkick) shouldn't be allowed as they are only consistently possible in a TAS.

tepples

03-27-2008, 12:24 AM

so if we had "evil bag" then it would use the worst fit algo, but otherwise obey the rules of the bag. (hand out the seven before repeating, but try to hand out the worst of the avilable ones).

But how would it know which is worst, six pieces in advance?

Zaphod77

03-27-2008, 01:10 AM

that's why all "worst fit" things turn off the history.

you can't figure out a worst fit 6 pieces in advance. you can't even figure out a worst fit 1 piece in advance.

All such "evil randomizer" options would have to have 0 previews.

Implemented by computer, it's easy to apply the drought rule to all the pieces.

However as a forum game, i only applied it to the I, because it's the piece everyone cares about, and is the one most likely to be starved, to set up things for a minimal amount of bookkeeping. http://www.tetrisconcept.net/forum/images/smilies/icon_smile.gif

Zaphod77

03-27-2008, 01:28 AM

i believe "fling" synchro is too diffiuclt to perform reliably, but "just frame" is definitely possible with practice. "wall kick" is simply another type of "just frame" where the rotation involves a wallkick.

But perhaps a limit on the number of allowed synchros other than "blocked" is needed.

jujube

03-27-2008, 02:46 AM

there are no evil randomizers, only evil tetrominoes colored in evil ways.

http://img140.imageshack.us/img140/8501/evilsrswz1.jpg

Ghett0

03-27-2008, 03:30 AM

there are no evil randomizers, only evil tetrominoes colored in evil ways.

My god... It's the Super Rotation System, with Infinity! RUUUNNN! Lol. Anways, is this topic a game, or a place for discussion? http://www.tetrisconcept.net/forum/images/smilies/icon_razz.gif

jujube

03-27-2008, 04:46 AM

[discussion]

Zeta

03-27-2008, 04:52 AM

Ghett0

03-27-2008, 04:57 AM

jujube

03-27-2008, 05:03 AM

Ghett0 - your S is a Z...i can't make my move http://www.tetrisconcept.net/forum/images/smilies/icon_wink.gif

Zeta

03-27-2008, 05:04 AM

@Ghett0>>I had my J and L peices backwards, and changed my (faulty) move to a valid move before your post showed up. please update your board... <.<;;

Ghett0

03-27-2008, 05:04 AM

Ghett0 - your S is a Z...i can't make my move http://www.tetrisconcept.net/forum/images/smilies/icon_wink.gif

XD Fixed.

DIGITAL

03-27-2008, 05:08 AM

You guys should use fumen (http://fumen.zui.jp/default.htm) for this. http://www.tetrisconcept.net/forum/images/smilies/icon_razz.gif

jujube

03-27-2008, 05:11 AM

Zaphod77

03-27-2008, 05:20 AM

fumen link so far

(http://fumen.zui.jp/?v105@7eFZaB7iB1mBWlBPdBsVB)

to place, click on the link, click the button with the down arrow until the last piece darkens, click the color radio buton matching the piece, and click it on the board, then click the button with the star, then press ctrl-c

you can then paste it here.

Zeta

03-27-2008, 05:22 AM

um, someone help me with fumen, please?

*is a bit... confused. nihongo no wakarimasen*

[edit: nvm]

Edo

03-27-2008, 05:26 AM

Ghett0 and jujube, how are your placements possible in 20G?

Zeta

03-27-2008, 05:30 AM

Ghett0's is possible (I think) if he holds A+Right on entry.

as for jujube, dunno. I had a move planned for that "L" piece, but I'll hold on it until someone can validate Jujube's move.

Rosti LFC

03-27-2008, 05:31 AM

Ghett0 and jujube, how are your placements possible in 20G?

They're not

Zaphod77

03-27-2008, 05:32 AM

quite right. both are impossible. even with synchro.

There is no initial L-R move. But if init L-R did eist, then the S would be placeable there.

but assuming you have jp text support, my guide above works.

not as easy as it looks, huh. http://www.tetrisconcept.net/forum/images/smilies/icon_smile.gif

Zeta

03-27-2008, 05:37 AM

meh, I got it figured out...

hold on, updating to last known valid move and flagging the two questionable/invalid moves.

here you go: updated (http://fumen.zui.jp/?m105_407eFZaB7iB1mBWVPbAMSceEFbs9DunVKEFb0iEBDGoD ?FbcEEPSdCAPNPWARgctCUu6cChz7BDFb8KCuhrVCp2AAAsF?P HAplSxBM+RBA)

(change "?m" to "?v" to edit, BTW)

Zaphod77

03-27-2008, 05:39 AM

Zeta

03-27-2008, 05:44 AM

well, going off of zaphod77's suggested moves...

positioned L-block (http://fumen.zui.jp/?m105_407eGZaB7iB1mBWVPbAMSceEFbs9DunVKEFb0iEBDGoD ?FbcEEPSdCAHMPUATABFCFNu+BEoo2ANny+BFb85AMNPUATA?B FCFNu+039236F160)

(rot. rule abuse FTW!)

[full explanation: IRS (B) -> right ->(B) -> (B) -> (B) -> Right to position.]

[looking at it again, I probably could've IRS (A)'d it and saved a couple (B) rotations]

next up: T

Edo

03-27-2008, 05:54 AM

Sorry Zeta, I don't think that L placement is possible in ARS, it's one of the wall-kick exceptions where a square in the central column of the 3x3 bounding box is occupied.

Zaphod77

03-27-2008, 05:55 AM

That fails.

if you try it with clockwise rotation only, it will kick back to the left, and if you try to spin it counter clockwise, then back, it will hang up on the nub and will not rotate.

This is valid, though, with DAS synchro. (http://fumen.zui.jp/?v105_407eGZaB7iB1mBWlBXcBcdB6VB)

jujube

03-27-2008, 05:55 AM

Ghett0 and jujube, how are your placements possible in 20G?oops, didn't read the instructions carefully enough http://www.tetrisconcept.net/forum/images/smilies/icon_lol.gif

i was too busy discussing and posting pictures and stuff.

Zaphod77

03-27-2008, 06:34 AM

that move would have worked... in ACE-ARS, which allows all pieces to floorkick. We are playing by TAP rules, though, which forbid all floorkicks and all I kicks.

Caithness

03-27-2008, 08:09 AM

Double rotate the T. (http://fumen.zui.jp/?v105_407eHZaB7iB1mBWlBXcBcdB6VB1XB)

History: SZLT

Next: J

I think there should be a participant history rule of four as well as a tetromino one.

Also, so far we're following 7-bag.

Zaphod77

03-27-2008, 11:38 AM

probably. i'm only coming in to correct placements. hoepfully i won't have to do it again. http://www.tetrisconcept.net/forum/images/smilies/icon_smile.gif

i'm not sure my placement was the best one in that case, but it is a valid one. have fun, people. http://www.tetrisconcept.net/forum/images/smilies/icon_smile.gif

Poochy

03-27-2008, 11:47 AM

IRS the J clockwise, then lock it down there. (http://fumen.zui.jp/?v105_407eIZaB7iB1mBWlBXcBcdB6VB1XBONB)

|� � JJ� � |
|� � J� � �|
|LL� JT� � |
|IL� TTTZ� |
|IL� SSZZ� |
|IJ SSTZOO |
|IJJJTTTOO |
+==========+History: ZLTJ

Next: S

(No, I'm not trying to throw the game with bad placements. In an actual T.A. Death round, I'd actually do this and try to get a Z or T under the "overhang". Though in this case, it conveniently makes an S almost impossible to place well. http://www.tetrisconcept.net/forum/images/smilies/icon_razz.gif)

Pineapple

03-27-2008, 01:37 PM

Amnesia

03-27-2008, 04:04 PM

|� � � � � |
|� � � � � |
|� � � � � |
|� � � � � |
|� � � � � |
|� � � � � |
|� � JJ� � |
|� � J� � �|
|LL� JT� � |
|ILOOTTTZS |
|ILOOSSZZSS|� �<--- ! line to remove !
|IJ SSTZOOS|
|IJJJTTTOO |
+==========+

Next : I

It is interesting gila to simulate a bit a disrupting of a stack..

But please don't do it every time ! http://www.tetrisconcept.net/forum/images/smilies/icon_lol.gif

Zeta

03-27-2008, 04:07 PM

I-block placement (http://fumen.zui.jp/?v105_407eLZaB7iB1mBWlBXcBcdB6VB1XBONB_2FeBbWBxYB)

next up: Z

also, fumen takes care of line deletes automatically.

I'll upload a translated version of it onto my site when I get home from school today. it's actually quite a useful planning tool ^-^

Ghett0

03-27-2008, 05:09 PM

that move would have worked... in ACE-ARS, which allows all pieces to floorkick. We are playing by TAP rules, though, which forbid all floorkicks and all I kicks.

I could've just IRS'd the I.

Zaphod77

03-27-2008, 06:40 PM

i'm not complaining about the I. i was complaining about the L.

But keep in mind that in 20G is can now be difficult to place.

Zeta

03-27-2008, 10:12 PM

well, while we're waiting on someone to move with the Z block, I would like to let the japanese-illiterate participants (and anyone else who cares) know that I UL'd an english translation of FUMEN to my site.

you can see (and use) it here (http://zeta.hypersector.com/Tools/TetrisField/default.htm).

the codes used are identical to those used by the original JP version.

cdsboy

03-27-2008, 11:45 PM

Thank you for the english Fumen. Now all the buttons make sense!

Ghett0

03-27-2008, 11:48 PM

Zeta

03-28-2008, 04:12 AM

Thank you for the english Fumen. Now all the buttons make sense!welcome. also, none of that is machine translation - all of the translations are based solely off of what the button/switch does. granted, I did kinda use systran to get the just of what it originally said.

also, the two-line statement at the bottom should cover my rear end legally. heck, if the guy looks at the source, all the comments save for one line are intact and unmodified. (I added that one line XD)

With that S piece.. that could've been DAS'd over a space and rotated.I already suggested that, the Thread Creator shot it down, but the Z move after that was also questionable.

once again, the next piece (if anyone wants to go ahead and make a move) is a Z-block.

[edit: I think this little forum game had a sound idea, but ended up getting battered and derailed due to disputes over 20G-possible moves (I'll accept the Rot. Rule dispute on that one L-block move >.<). may I propose we start a different one without the 20G requirement?]

Zaphod77

03-28-2008, 04:47 AM

there is no init left/right move.

If IRSed, it would be in 4-5, even if you charged DAS right.

the nub was in six. it could not cross.

Feel free to start another game with different rules if you feel this one is too difficult. http://www.tetrisconcept.net/forum/images/smilies/icon_smile.gif

If one of the forum gods corrects me on my asessment of 20g play i will apologise profusely. But i'm pretty darn sure I know the rulez. http://www.tetrisconcept.net/forum/images/smilies/icon_smile.gif

Ghett0

03-28-2008, 04:48 AM

I know there's no init LR, but I believe doing a Right + Rotate synchro on the first frame the Z spawned would work, no?

Zaphod77

03-28-2008, 04:59 AM

nope. rotation is processed that frame, but movement is not.

WIthout IRS

|I
|I� �SS
|IJ SST OO
|IJJJTTTOO

First process rotation
|I� S
|I� SS
|IJ� ST OO
|IJJJTTTOO

then try to process movement.
|I� S
|I� SS
|IJ� ST OO
|IJJJTTTOO
(movement fails)

then process gravity.

the processing order is.

1) Initial Hold System. (doesn't exist in this game)

2) Initial Rotation System. (does exist)

3) gravity (20g, drops immediately to bottom in initial colums)

and then loops to

4) rotation (with wallkick)

5) movement.

6) gravity.

for each frame.

In a "synchro" rotation is still processed before movement.

They are not truly simultaneous. but they are processed in the same frame, and before gravity for the frame.