Tetris Party Skill Graph

[lampi]

Those familiar with Tetris Party probably know there's a "skill graph" under Records. I contacted their support e-mail asking what specific gameplay data is used for each attribute. Here is their answer:
[!--quoteo--][div class=\\\'quotetop\\\']QUOTE[/div][div class=\\\'quotemain\\\'][!--quotec--] Thank you for contacting Tetris Party Support.  Here is the information you requested.

Attack Power
Will increase according to the cumulative number of Penalty lines sent to opponents in VS mode

Defensive Ability
Will increase according to the cumulative number of Penalty lines cancelled in VS mode

Judgment
Will increase when the number of Tetrimino rotations done in each game is below a set number

Reflexes
Will increase according to the cumulative number of lines cleared at the top portion of the Matrix (upper 6 lines)

Improvement
Will increase according to the cumulative number of lines cleared

Expertise
Will increase according to the cumulative number of games played

We hope this answers all of your questions.  Thank you.

Sincerely,

Tetris Party Support
[/quote]

It's kind of disappointing to discover that most of these are based on cumulative figures, though. Anyway, hopefully this will be interesting to other people too (:

[iphys]

Huh, I had no idea what it was, but it took me a long time to finally max out reflexes.  Everyone I know had trouble with reflexes, and it's probably because everyone I know is a decent Tetris player, and you wouldn't have to clear as many lines in the top 6 rows if you're good and keep your stack low.  Makes the whole thing seem kind of meaningless now if it's basically a cumulative thing.  Looks like judgment is some sort of measurement of efficiency at least.  Thanks.

[caffeine]

If I were to make a radar chart for Tetris...

The center of the chart represents 0%, while the edge represents 100% of the average of players in the 98th percentile's rolling average of their last fifty games they've won rated matches. No 99th percentile to exclude hackers with inflated scores and no lost games to exclude players who do a T-Spin and then top out or something to that effect. You can exceed the edge of the graph.

Offense: total lines you've sent to opponents divided by total lines cleared over the last fifty games. A lot of T-Spins and high Combos result in a higher statistic. This is superior to APM in that it actually rates the players offensiveness per se without becoming "speed inflatable."

Defense: average highest occupied row over the last fifty games. This means you take the total number of game states where your highest occupied row was 1 or 2 or 3 etc and divide by total game states. In the long run, a defensive player will keep a lower highest occupied row at any given time. This is better than using a "garbage lines cleared" variable because this also takes overstacking into account.

Flux (or Efficiency): ((garbage lines cleared) + (garbage lines added to opponents)) / minutes for the last fifty games. Represents your "practical cost-effectiveness" (or alternatively, how much bang you get for your buck divided by time).

Speed: tetrominoes / minutes. An oldy but goody.

Accuracy: using an algorithm like that seen in TOJ, total misdrops over the last fifty games.

[lampi]

Huh, I had no idea what it was, but it took me a long time to finally max out reflexes.  Everyone I know had trouble with reflexes, and it's probably because everyone I know is a decent Tetris player, and you wouldn't have to clear as many lines in the top 6 rows if you're good and keep your stack low.  Makes the whole thing seem kind of meaningless now if it's basically a cumulative thing.  Looks like judgment is some sort of measurement of efficiency at least.  Thanks.

Yeah, I sent them the email asking about the graph precisely because I couldn't seem to improve the "Reflexes" stat. It's just how you described, no good enough player will clear many lines at the top 6 rows.
Glad to see someone liked the info (:

[jujube]

caffeine-i really like the idea of rolling stats. i wish they did that in TOJ.

how would you define a misdrop though? they can often be 'fixed' with T-spins and/or combos.

[caffeine]

how would you define a misdrop though? they can often be 'fixed' with T-spins and/or combos.

I'm going with whatever TOJ's algorithm does, and TOJ knows the difference between a misdrop and a T-Spin set up.  It could just be a hole or un-fixable gap (there are only so many fixable gaps in SRS), but you're right that sometimes players make holes on purpose. I can't think of any advantage in making a hole for a Combo, but I can for doing a platform clear. A platform clear, though, has the tendency to have multiple contiguous empty cells below it when setting up. Therefore, I think a safe definition of a misdrop is any drop that results in a single, totally covered hole and doesn't immediately cause a line clear.

Identifying holes, gaps, and platforming:

[fumen]105@7ePxcIUhB1nBTaBXdBibB2VBU8OHANecoDy31CAeCP?NAGufeDCDVTASI4dDQBAAAJxOAAtHB/jBrdBaNPLAwC0dEm?XyGEJmhCAVZBAwNQAwC0dEmXyGEFbsiDs4DXE[/fumen]

[hurtchow]

I love platforming. It makes me feel like I can see into the future!

Especially with T-Spin set ups. I'm always like "LOOK, LOOK WHAT I'M DOING. SEE THAT GAP.. OHHHH TSPIN"

And then they're all like "how did you do that?" and all I gotta say is "I can see into the future!"

[jujube]

I'm going with whatever TOJ's algorithm does, and TOJ knows the difference between a misdrop and a T-Spin set up.  It could just be a hole or un-fixable gap (there are only so many fixable gaps in SRS), but you're right that sometimes players make holes on purpose. I can't think of any advantage in making a hole for a Combo, but I can for doing a platform clear. A platform clear, though, has the tendency to have multiple contiguous empty cells below it when setting up. Therefore, I think a safe definition of a misdrop is any drop that results in a single, totally covered hole and doesn't immediately cause a line clear.

i made a couple examples of making a weird placement that leads to a combo. the 2nd example includes a T-spin. with a few weird placements you could end up for a longer combo. of course even a short combo is beneficial for countering garbage without receiving any, so one could argue that the S placement next to the L isn't a misdrop in TOJ multiplayer. and the 2nd example has 2 placements that could be considered misdrops. i think your definition of a misdrop is good, but it doesn't apply to all situations in all games, if it's considered to be a bad thing.

[fumen]105@JeI3gbI3gbI3gbI3qbSKI7eInLBDPB+OBMGBJEBeIB?NMBpeBAAA[/fumen]
[fumen]105@JeI3gbI3gbI3gbI3qbSKI7eJnLBDOBMPBOGBJyAyJB?HIBFMBpeBAAA[/fumen]