Tetris & Dr. Mario (SNES)

[iphys]

Um, yeah, if it's memoryless it's possible (although not hugely likely) to go 25 lines without getting an I.

[Pineapple]

===== Piece Counts =====
Piece I: 25/200
Piece J: 33/200
Piece L: 22/200
Piece O: 27/200
Piece S: 34/200
Piece T: 26/200
Piece Z: 33/200

===== Interval Counts =====
Interval 27: 1/193, 0.005181347
Interval 26: 2/193, 0.010362694
Interval 24: 2/193, 0.010362694
Interval 23: 1/193, 0.005181347
Interval 22: 1/193, 0.005181347
Interval 20: 1/193, 0.005181347
Interval 18: 1/193, 0.005181347
Interval 17: 3/193, 0.015544041
Interval 16: 3/193, 0.015544041
Interval 15: 7/193, 0.036269430
Interval 14: 5/193, 0.025906736
Interval 13: 1/193, 0.005181347
Interval 12: 5/193, 0.025906736
Interval 11: 4/193, 0.020725389
Interval 10: 6/193, 0.031088083
Interval  9: 11/193, 0.056994819
Interval  8: 18/193, 0.093264249
Interval  7: 10/193, 0.051813472
Interval  6: 11/193, 0.056994819
Interval  5: 16/193, 0.082901554
Interval  4: 14/193, 0.072538860
Interval  3: 26/193, 0.134715026
Interval  2: 20/193, 0.103626943
Interval  1: 24/193, 0.124352332

Annoyingly, a sequence that short is inconclusive. Anything longer? Like... if you had the sequence from that 290 line game that would be awesome...

[iphys]

That actually looks pretty good!  I have the video, but I hesitate to use the sequence from a best game, because it could be kinder than normal.  Even with a longer sequence of pieces, there'd still be a lot of uncertainties in the statistics.

[iphys]

I found this Spectre video of 347 lines, which looks a lot more true to reality than that other video.  If Spectre can't speed tap the pieces to the walls like in that other video, then I doubt anyone actually could.

[Sisu]

wondering about strategy for versus. seems like stacking high is ill-advised.  

MGC finals 2012:

DTL 2009:

zack = sirtetris IYDK

[Integration]

I would formulate it as following: stacking high is ill-advised, if garbage column changes (or if you know, where the next garbage column will be). And there's no point in stacking higher than 8 rows. It's unprobable that you get 3 line pieces within your next few pieces.

[Joshua_Tolles]

Tactics for two player vs. mode:

Two very important things to keep in mind: The canals in the garbage lines pop up in the same place for both players, and both players get the same pieces.

So Job #1 is always set up for a double as fast as you can, and clear the double.  You will see where the canal in your own garbage will be by looking at your opponents single line of garbage..  You obviously do not want to put pieces there, so instead, build your wall around that spot, and your opponent will help you set up for Tetrises by sending you garbage.

SNES can be brutal in its randomizer, so always keep a conservative approach to going for Tetris: make sure you can burn lines easily!

If your opponent gets ahead of you in number of pieces played, you can watch his next box and know what you are going to get later on!    The farther ahead he gets, the more of a preview you get (assuming you have the RAM in your brain to keep track of course   )  This is very handy when the longbar drought comes along. If you are behind in pieces, you will sort of have a "heads-up" about when the drought will finally break.


And an "in case you missed it":
392 lines in SNES, pure DAS charging, no hyper-tapping.

[caffeine]

Two very important things to keep in mind: The canals in the garbage lines pop up in the same place for both players, and both players get the same pieces.

Something else worth noting: the empty garbage column appears change every ten rows. I'm not 100% certain of that. It appears to be random. I'm also not sure if it'll randomly assign the same garbage column twice in a row.

So Job #1 is always set up for a double as fast as you can, and clear the double.  You will see where the canal in your own garbage will be by looking at your opponents single line of garbage..  You obviously do not want to put pieces there, so instead, build your wall around that spot, and your opponent will help you set up for Tetrises by sending you garbage.

In other words, you're paying the price of less throughput (1 garbage sent for 5 pieces) to gain knowledge of the empty column. What this knowledge earns you is a chance to not leave upstack over the empty column after your first attack. However, in many cases, it only costs around 5 pieces to uncover that upstack after the first Tetris. For example:

[fumen]110@7eP2MWoAFLDmClcJSAVDEHBEooRBKoAVB6S9wCTH8L?CK9aFDUuaFDMNmPC0gBynBJUBlhB/aBTWBycBpYB+SB/QBz?MBMOBdMBpeBTnb[/fumen]
HD's embedded Fumens seems to be down. Here's a link

Let's say after that first Tetris, I see my opponent receiving the empty column in the 5th column. I prepare accordingly.
[fumen]110@VessjbQphbkzts1lIwRp8egblzss1lIwQp9eqbT9VZ?AFLDmClcJSAVDEHBEooRBPoAVB6yTxCqAAAA7eBMYB6jByd?Qphb1ljbRphb1lssibRpZiD3gbI3gbB3QpB3tsB3gbA3RpB?3ts0lA3gbA3RpJwpbXnb[/fumen]
Link to fumen.

Besides those used for the first Tetris, I used 8 pieces. One was a double. The remaining 3 pieces can be converted into another Tetris. All in all, this is a very similar outcome compared to what I might get had I started with a double.

So then, we might conclude that it's roughly an equal strategy either way. This is not the case. If I use the opening double and my opponent opens with a Tetris, then my opponent has an advantage. I'm giving him the benefit of that empty column knowledge before he's finished stacking for his first Tetris. This can allow him to stack in such a way that nothing covers the empty column past the fourth row. After finishing his Tetris, he'll have prepared in the same manner as myself. The difference is that he'd earn 4 rows sent for 10 pieces = 0.4 sent per piece, while I'd have earned (1 sent for 5 pieces) + (4 sent for 10 pieces on the next-coming Tetris) = 0.33 sent per piece.

If I had to guess, a good strategy would be to keep a stable surface while Tetrising down garbage at the earliest opportunity. When there is no more garbage, use upstack Tetrises.

[Joshua_Tolles]

So then, we might conclude that it's roughly an equal strategy either way. This is not the case. If I use the opening double and my opponent opens with a Tetris, then my opponent has an advantage. I'm giving him the benefit of that empty column knowledge before he's finished stacking for his first Tetris. This can allow him to stack in such a way that nothing covers the empty column past the fourth row. After finishing his Tetris, he'll have prepared in the same manner as myself. The difference is that he'd earn 4 rows sent for 10 pieces = 0.4 sent per piece, while I'd have earned (1 sent for 5 pieces) + (4 sent for 10 pieces on the next-coming Tetris) = 0.33 sent per piece.

If I had to guess, a good strategy would be to keep a stable surface while Tetrising down garbage at the earliest opportunity. When there is no more garbage, use upstack Tetrises.

Thank you for the in depth analysis   Of course, I didn't mean that you actually had to start with a double, I was trying to express the idea in an easy way to digest the concept. I should have stated that Job One was to be conscious of where the canal is going to pop up

[Blink]

Hey I just got Tetris & Dr. Mario as a gift from team STARS!  Thanks guys !

[benmullen]

Hey I just got Tetris & Dr. Mario as a gift from team STARS!  Thanks guys !

Josh, we are ruined.  haha

[caffeine]

I take this to mean SNES Tetris will be an event at this years Classic Tetris Championship?

[Sisu]

I take this to mean SNES Tetris will be an event at this years Classic Tetris Championship?

Yes actually, but I thought perhaps Ben was referring to his and JLT's A-type records.  

If you beat 9-5 in the B-type game...then it lets you play 10-5

QFE

Instead of simply a win screen (space shuttle, dancers, etc.), you are presented with a new challenge:

[attachmentid=441]

...then to 11-5, 12-5, onward and upward presumably. If you top out in this unique "B-type marathon", you are sent back to the level & height select screen. What would be the gravity that would make height 5 impossible to clear?  

[Joshua_Tolles]

"If you beat 9-5 in the B-type game...then it lets you play 10-5"

Instead of simply a win screen (space shuttle, dancers, etc.), you are presented with a new challenge:

[attachmentid=441]

...then to 11-5, 12-5, onward and upward presumably. If you top out in this unique "B-type marathon", you are sent back to the level & height select screen. What would be the gravity that would make height 5 impossible to clear?  

I tried this out last night for the first time, and it is really fun!  I got to 15-5 in my third attempt.  

Without any evidence, and just guess-timating, I suspect that 19-5 through 28-5 would be "technically" playable, but to clear it ten times in a row may be virtually impossible. Maybe, maybe not. I haven't done any math on it or tried it out yet.

18-5 should be challenging, but beatable. I assume this based on NES B-Type experience. In NES, 18-4 is normally beatable for "very good" players, and 18-5 is possible, but is mostly dependent on the trash sequence given and good piece fortune. Level 18 is considerably easier on SNES than NES, so I imagine that SNES 18-5 should be at least as easy as NES 18-4, if not easier.

As far as my estimation regarding 19-5 through 28-5 is concerned, I base that upon my belief that Levels 19-28 on SNES are ever so slightly easier than Level 18 on NES. So, I believe that would make SNES 19-5 slightly easier than NES 18-5. Again, I haven't done the math on it, or even played yet, but I think that would be the case.

I am interested in trying a 9-0 to 31-0 B-Type marathon!

[Sisu]

I am interested in trying a 9-0 to 31-0 B-Type marathon!

9-0 doesn't do the same thing

QFT

BTM is only accessible via 9-5. Is reaching 29-5 the holy grail of SNEStris?