Skepticism regarding recent video of player AVIZL

[yotipo91]

I'm PC illiterate.. but it fascinates me how you all haven't come to a clear conclusion whether it's a bot or not. When I opened the thread and read all the posts I imagined it was a bot, then I read more until the last post and thought it was human. Could there be a program that interprets input presses as individual keytaps? Could that program effectively lower DAS delay? That's the only way I can imagine those TF players can play that fast.

[tfn]

As far as my perfect clear ability, I have never seen anything like that. So I think it's a bot. lol.
I don't have any proofs though. I just observed his match and its the first thing that comes to mind. Given the idea that he's not that well known ever since and became an instant star. It is pretty much alarming.
Though I'm not in any position to accuse. All I have is my experience in perfect clears. I already can say that I'm well accomplished but no way near that performance. That's just superb.

[KoRoBeNiKi]

http://www.youtube.com/watch?v=yjcEgLGWqak this video of his is also quite impressive (and no, not just the time)

[Blitz]

Whether or not AVIZL's play is by an actual human or a bot is not interesting to me.
What's interesting is that I think avizl's pc playstyle can be executed by any human with enough practise. Think of how hard doing T spins was when you first started playing Tetris. After some practise you have familiarised yourself with T spins and you can execute them relatively easily. You probably started learning basic T spin set ups such as TKI and later you learned to recognise T spin oppertunities in mid game allowing you to set up T spins in mid game.
Now try setting up this and attempt to do a perfect clear from there. It's not that hard right?
[fumen]v110@JeRpjbai8essRpibYi1l8etslzhbIw1l8esslzibKw?8epbA4G[/fumen]
When you can successfully do that, try to stack up another 4 or 6 lines leaving a gap somewhere in the stack and then try to fill that gap in a way that allows you to end with a perfect clear. If you have no experience doing pc's you probably wont get it on first try, but if you keep trying you will eventually succeed. If you keep practising this your brain will start to sort out what works and what doesnt work and develop the ability to recognise pc patterns allowing you to do continuous perfect clears. Once you have learned to continuously make perfect clears you will start to notice your speed is increasing and if you keep practising you will be able to do perfect clears faster over time.
I believe learning this is not much harder than learning to do T spins.

[farter]

I say legit.

There are two real tells that say 'not a bot' to me. First, it might be easy-ish to program a bot that did only perfect clears, but T-spins are a hell of a lot harder.

i say no.

a bot who calculates for PC must use and calculate many (should be more than 3) previews.

and with that, so many previews used, simply add a t-slot detection and a positive score in the code lets it run much better for human t-spin player, because usually man can't even really "consider" 6 previews. instead, mostly the 3rd to the 6th one are just for reference if some piece needed is coming or not. however a bot, if only it DOES use the 6th preview, it can easily detect a t-slot formed with that piece.

a human player who really simulates the playfield in his/her brain to the 3rd preview for most of the time, keeping this speed, is already a korean. (

[es2mac]

Has anyone considered an "AI assist" midway of a real human and a bot, Google Glass-style?

Say in this case, a program that watches the field when you play, and when a quick PC setup is detected, overlay the solution on your field similar to ghost pieces?

Just throwing the idea/possibility out there, not saying it's likely in this case.  Though someone should totally try coding one, and bring on a golden age of TF shadiness.

[myndzi]

i say no.

a bot who calculates for PC must use and calculate many (should be more than 3) previews.

and with that, so many previews used, simply add a t-slot detection and a positive score in the code lets it run much better for human t-spin player, because usually man can't even really "consider" 6 previews. instead, mostly the 3rd to the 6th one are just for reference if some piece needed is coming or not. however a bot, if only it DOES use the 6th preview, it can easily detect a t-slot formed with that piece.

a human player who really simulates the playfield in his/her brain to the 3rd preview for most of the time, keeping this speed, is already a korean. (

The thing about it is, it's really obvious when a bot does T-spins because the distribution of common to uncommon setups is really skewed. A human player can have moments of brilliance and foresee a good T-spin opportunity, but a bot will, programmed naively as you suggest, make a bunch of unnatural-seeming T-spins that are actually very efficient for the very reason that it CAN calculate all the possibilities with all the previews.

[officegunner]

I was intrigued, so i went back and checked his profile again, and found out that he has a 21 combo?

from what I can see by just watching his gameplay and profile,think if he's a human, he's:

-A center 4 wider who has developed incredible flat stacking by center 4w stacking

-instead of taking the conventional mindset to get b2bs he chose to combine his flat stack with very fast skimming until the I piece gets in the bag to survive aftercombo for a mid 4wider

With such a incredible flat stack, it isn't that hard(imo) to make a flat stack 6 columns high with a 2 wide gap to clear down with 3 pieces to achieve a perfect clear? Maybe i'm being a bit optimistic here but i think that is very learnable.

the way I see the game as an actual player it looks like a player who has  extensive 4 wide ability combined with a lot a lot of "square"/"playing forever" practice?

http://harddrop.com/fumen/?m110@Je/eaiKwRp...nz/etsnzpbAwNAA

anyway i fumened this.... idk what to make of this but it could be learnt, potentially a whole new breed of tetris players

[tetris_loverrrrr]

i've seen AVIZL before...he was once in the same room as that hacker The_World.....AVIZL got his ass kicked though lol  =D

[pwn_by_numbers]

Just played a few matches with AVIZL today. Before playing him for real, I was pretty convinced he was a bot, mostly because of the argument about how he switches flawlessly from PC's to T-spins the moment there's an odd amount of garbage. Anyway, today I watched him play a bit, and I'm starting to think he's a real person. After about 4 matches of flawless play I saw him make a pretty bad misdrop, which caused him to lose:

[!--ImageUrlBegin--][a href=\\\"http://i.imgur.com/MsOvzQR.png\\\" target=\\\"_new\\\"][!--ImageUrlEBegin--][img width=\\\"400\\\" class=\\\"attach\\\" src=\\\"http://i.imgur.com/MsOvzQR.png\\\" border=\\\'0\\\' alt=\\\"IPB Image\\\" /][!--ImageUrlEnd--][/a][!--ImageUrlEEnd--]
[!--ImageUrlBegin--][a href=\\\"http://i.imgur.com/HP3eR6f.png\\\" target=\\\"_new\\\"][!--ImageUrlEBegin--][img width=\\\"400\\\" class=\\\"attach\\\" src=\\\"http://i.imgur.com/HP3eR6f.png\\\" border=\\\'0\\\' alt=\\\"IPB Image\\\" /][!--ImageUrlEnd--][/a][!--ImageUrlEEnd--]



Now, either there was some error in the bot, or this was a human error. I think what was most telling was after he made the misdrop, he paused for a few seconds, then tried to spin a piece under to fix it, and failed the spin. He lost the match and quit after that. So either he's actually human, or has some sophisticated code that fakes making mistakes. (I also wouldn't be surprised if he just has some sort of mod that predicts PCs and tells him where to place pieces.)