Freeform Invisible Tetris (Phantom Mania Cleared)

Thank you for your last explanations QM.
What i'm gonna explain now is my personnal opinion and it don't engage anyone else or the TC community.
I'll be very straightforward. I'm very grateful to you to have introduced the "Art of memory of loci" to me. But i'm very annoyed by what you are doing here. I could try to demonstrate and argue about what's wrong/fishy in all your explanations and videos demonstrations but i'll directly jump to the conclusion :

You have the behavior of Pathological lying (also called pseudologia fantastica and mythomania).

My words are rude, but i have enough inputs to be fairly convinced and prove my point of view despite all the effort you make to keep your story convincing. So it would save us a lot of time and energy if you just recognize the facts. Again this is only my personal opinion but i care enough about this Tetris community to try my best to make you Persona non grata in TC and Harddrop if we come to flame war. You certainly have some tetris skills but definitely not what you are trying to demonstrate here. . Please make amends, delete those videos.. so we can move peacefully to something else

 

I'm glad you acknowledge your opinion as your own. It's completely understandable and frankly I expected it sooner or later. I looked up "pathological lying" and I see your point, but you misunderstand me.

I'm going to be similarly terse here: I don't cheat. You're convinced I'm pathologically lying and there's nothing I can do about it, but that's your call. I can only do so much. That goes for anybody else who shares your opinion. And don't worry about wasting time and energy on this matter - I won't bother and neither should anyone else.

 

Well you should not just look up but consider seriously about this pathology. You should start reading the wikipedia page for instance.

Apparently i was expecting too much but it seems to be normal. If you just prefer avoid arguing, it's fine to me as long you don't post anymore about your superhuman feats. it's disrespectful to this community.

 

Whatever you think about me, you should know this - I didn't come here to start a flame war. As a very busy person I'm happy to leave and not come back (I first joined to talk about invis anyways). For those of you who have now unconvinced yourselves of the veracity of my gameplay, I'm terribly sorry. For everyone else, I'll continue to make videos and hope to join you there.

-?

 

It's not true that there is nothing you can do. There seem to be two or three primary points of conflict that have come up in discussion regarding your gameplay, and it should be fairly simple for you to explain them:

1. You claim that you're storing upwards of 6 pieces of information (or loci) per second, yet the world's fastest memorizers can barely store 500 digits in their head in 5 minutes (roughly 1.5-1.7 per second) while not doing anything else. Can you explain this discrepancy in a way other than that you have some superhuman ability that no one else has? This is not an attack; it just seems impossible to explain and I'm curious.

2. Your Secret Grade video shows that you completed the first half of a secret grade in 20G in 17 seconds...faster than anyone else has ever done it in *visible* play. Upon doing this, instead of trying to finish it off, however, you just suicide out the game while it's in a perfectly viable state. Can you explain why you did this?

3. In this video, the Nullpomino window is half off your screen. Some users (I've been asked about it by more than one person) suspect that you have set up Nullpomino with the "show visible playfield on right side" mod (which definitely exists and is shown in your videos for demonstration purposes) and are using a dual screen setup. Is there any chance you can record a video (any kind of gameplay, doesn't really matter) with the full screen area visible?

I'm sure you can understand the skepticism given the seemingly incredible leap in playing ability that your videos purport exists, especially in a community that has had issues with fraudulent scores in the past.

 

I hope so. This is not stellar timing for me but I'll do what I can.

Before I discuss this, I want to point out that I don't often think about how many operations I'm doing per second, much like people don't often think about how many pieces they place per second. My numbers are mostly an estimate.

The biggest thing is that I "cheat" my own system frequently. When I stack very cleanly, I know in advance that some block of squares is going to be completely filled, so I bypass all the bit memorisation (yes - I don't memorise any of it). I just visualise a mental image of my stack for the next second or so and fill everything in at once. Then it becomes not 12 operations/second, not 6, but 1.

That single update can sometimes hold a lot of information in it, but the squares that get updated are all adjacent to each other. (Remember, I have a journey through each room that takes me through 0-9, and I've retraced it thousands of times.) If I can complete perfect (Tetris-ready) rows while doing it, the updates for that row become completely trivial.

The way I play invisible Tetris is actually not directly comparable to memorisation and manipulation of dozens of abstract numbers or binary digits. The important difference is it's very, very geometrical - that's why visualising the stack is such a viable method to play. As I said before, I can't visualise the whole stack, just a few rows. But I do visualise those few rows, and use the resulting mental image to facilitate exact memorisation. The process happens very quickly, but it's there. And that momentary visualisation is key - I need that split second to memorise.

It's honestly not superhuman. According to this site, somebody memorised 48 binary digits in 1 second. The majority of my memory load is more like binary information than anything else. Of course they didn't have to sustain it for minutes but that's way faster processing than I will ever be able to do. And also they didn't do anything else apart from memorise the digits - but there's at least one person who does.

Rudiger Gamm can do unfathomably difficult calculations like 12th-20th powers of numbers in his head in a couple of seconds - suggesting he not only accurately memorises huge amounts of numbers in intermediate calculations, but also manipulates them in his head (this is the "doing something else" I'm getting at). Vastly more impressive than what I've demonstrated. I'm not setting a precedent here - I just wanted to bring "professional" mnemonist techniques to Tetris and to see how well it could be applied.

The crux of the matter is this: The aforementioned mnemonists/mental athletes have the handicap of dealing with something very abstract. They have to memorise formless numbers, presumably in some way that ties them into their memory palaces. Invisible Tetris is very geometric and already adds a layer of visual memory which I abuse the heck out of to make memorisation faster.

Yeah, it's a bit embarrassing though. I put together the rest of the video and synced the audio but had about 20 seconds left I needed to fill, so I recorded this bit about an hour before finishing the video - you can see how rough I edited that clip. I planned to top out anyways so I could have the "big reveal," I just forgot to cover the hole (as zowayix points out) but when I saw the timer I thought I had a perfect clip.

Speaking of which, the time is good, but not insane - notice I'm warping to Level 300 which has a big leap in potential speed and a still reasonable forced speed. I don't know if a time like that would be possible (outside of maybe TAS play) with higher ARE, even as fast as the first section of Shirase. I'd say I was about 3-5 seconds too slow on most runs, but this was quite lucky with the S and T, and I was a bit too eager to stop the timer.

I will point out that 1) I've been mucking around with secret grades for years, much longer than this invisible business and 2) I stopped before the "hard part" of turning around. TL;DR that whole clip was literally made for the video. :/

Also to expand on 1), it's not like this was a PB secret grade - I have a very funny/painful video (not online yet) where I mis-memorised a hole which turned an m9 into an S7.

*facepalm* Yes, leave it to me to make a video demonstration, aimed at vindicating myself, and leave out something as obvious as half my screen. The thing that bites is that camera angle was really hard to set up because I wanted to get both my control scheme and the full Nullpomino window in the same shot. Oh well - different camera angle next time.

I was hoping there'd be something in the replay file I attached with that post which might prove I wasn't doing that. I included the first few seconds of me starting Nullpomino to show I wasn't re-recording a replay or something like that.

I didn't know a mod exists that shows visibility on one side - in my videos I just toggle visibility by pressing "E," which is a feature of vanilla Nullpomino 7.5.0.

I'm busy with school/work at the moment, but I'll try to record at least an M in Phantom Mania from my POV in the next couple of days - with the whole screen visible this time. (I'm a fair bit worse when recording live since retakes take so long to retake.) Hopefully there will be enough evidence there that I'm not doing something sneaky.

Oh yes, absolutely. I'm fully aware that people have tried to fake not just Tetris runs, but Phantom Mania games at that. It's a little sad for me but totally expected. I'll continue trying to prove the honesty of my gameplay in the future, but to avoid feeding the flames I won't do that by posting here.

 

Well people, I would like this conversation to focus on the initial subject problem : The whole method explanation is completely "Faery Tale" and nobody will be able to reproduce it even slowly. Bringing pseudo - science with figures from other fields, nice fumen, and of course lot of assured "I can assure it work for me but I can't explain why but probably because I'm superhuman " is BULLSHIT. Please don't be confused. Of course QM is a busy person so he will not be able to produce proof spontaneously (on demand) because he need time for frame - stepping/TASing.. whatever you will ask him.

So "Question_mark the unquestionable" (how ironical is your email address), stop whining on what you deserve, and please consult a psychotherapist :

 

That's unreasonable. People have lives outside of Tetris - everybody does. I didn't ask for the veracity of my accomplishments to be questioned this particular week - I should NOT be playing Tetris right now. But I'll make a final sacrifice for the sake of providing evidence in a timely manner.

I'm uploading an unlisted video now. Not an M, but as Kevin said it really doesn't matter - and there's enough clean invisible gameplay to make it clear I know exactly what I'm doing. YT says 89 minutes to upload and I'm not going to stay up that late, but it should be available at this link. If you want a better run, you will wait until I have more spare time on my hands.

This is about as much proof as I can provide, what with the short turnaround time. Whole screen is visible, I show Nullpomino starting up from scratch, and a fresh ruleset so that this mediocre game ranks in the high-scores (which replays/frame steps don't do). The replay is attached. I can't get a better camera angle that does what you asked, so you'll have to live with my hand being partly off-screen. Hopefully you can still see that I really am sending all the inputs.

If you still believe I'm an invisible Tetris fraud, then I'm sorry, but I can't help you. And it's not worth another minute of my time, so I don't want to help you. Let the record show, I have made efforts to vindicate myself. Unless anyone has legitimate questions for me like Kevin did, this is my last communiqué.

 

Hi,

the inputs are not a problem in this case, don't worry about it.

The ruleset is rather what is puzzling. We don't know what the randomizer rules for your game are. (See Edit) In fact, it looks extremely similar to a 7-bag type of randomizer, at which your skills haven't been put into doubt. In any case, I looked up and did some tests, and the probability that this is a TGM kind of randomizer is extremely low. I have an upper bound of roughly 10^(-5) considering the fact that there are no consecutive two pieces, but I could lower this probability even more by looking at the absence of droughts. Are you sure you put the TGM randomizer in your ruleset? The affirmative would be exceptional.

Also, in this one you seem to be doing tetrises only, and when the first hole appears you top out. This is a shame, considering the premise of the topic was to deal effectively with them.

If you are having obligations, then really there is no need to push it. You do seem tired and K's behavior has been rash, as he admitted. How much time do you need before you can have some more spare time?

EDIT : After consideration, I realized that there might be a confusion : Depending on the kind of TGM randomizer that you use, the differences actually become negligible. Is it possible to verify that you have been using the randomizer that has been used in TGM3? If so, then the proof is null and void. I'd like to apologize for the confusion.

 

Nullpomino does not have the 35-bag backed memory randomizer as in TGM3.

 

Nevermind, further testing has been made, and it is closest to the TAP randomizer than any other bag algorithm. You may disregard the argument about the random element, but the other questions still stand.

 

I actually started to have my doubts about this, primarily when I noticed the third point that Kevin raised pretty much the first time I saw the video. I figured I'd keep quiet, because for once I didn't want to be the one to stir shit up (for the same reason I also didn't bring up the discussion we'd already had on HD about perfect clear techniques being a mnemonic device that made it a lot easier than freestyle invis - something I believe you've kind of confirmed by needing to resort to a memory palace instead). I also don't bother with IRC any more so, like jago (assuming I've interpreted what he said correctly), my opinions aren't something I've ever discussed with someone else.

I also find it slightly bizarre that you deem the transition from SRS to ARS to be too much effort, to the extent that in discussing possible acceptable PM Gm rulesets on Hard Drop it was pretty much your one dealbreaker, and yet the construction of an extensive memory palace and a complex memorisation system apparently ain't no thang. I can see valid reasons for that (e.g. an arduous but interesting task is better than a less arduous but uninspiring one) but it flags as slightly fishy for me. Likewise that whenever you intermittently post about these sorts of things you frequently mention how you don't have time for all of it as a kind of caveat for further progress not coming, and yet you're clearly finding time somewhere because your rate of progress is very strong.

Still a bit on the fence - as has been said, it wouldn't be the first time people have claimed outrageous scores and been found out. I lean towards things being legit but I can't deny doubts are there and I'm sort of not surprised that all this got thrown up by someone eventually. FWIW I'd also say that at Eindhoven jago was initially very dismissive and sceptical that the videos and scores were real (I believe his reaction was along the lines of "Whut??? Non... eet iz TAS, right?"), before we pointed him at the thread and he started looking into the memory stuff a little. Don't know if a month of trying memory trickery is enough research to conclusively confirm or deny such scepticism.

It's also a sad fact that anyone with some reasonable coding and/or video editing skill can probably fake anything that's supposedly "proof", and as a result those who don't believe you will hold onto their beliefs in a fundamentalist way because they'll just believe you've cheated the video somehow. The only way to really prove it to someone who is adamant you're lying would be to demonstrate it live and in person to someone whose verification would be considered adequate to vouch for you (if you're based in the South-East somewhere then I could probably spare a trip).

As a final point I would say that, although it's a bit of an unpleasant business, I think it's far better that this is all raised in the open on here for civil discussion rather than kept in private channels where people can't defend themselves. And jago, I appreciate you being blunt in some ways, but I also think if you're going to repeatedly dismiss someone and essentially tell them to fuck off, you at least owe them some elaboration on your arguments. If you have "enough inputs to be fairly convinced and prove [your] point of view" then either explain yourself or be happy that you've said all you can be arsed to say on the matter and leave it for other people. Neither of your last two posts add any opinion or insight you didn't already make pretty clear in the first one.

 

More suspicious than anything else is that when I asked for a video with the full game screen available, the video posted was of a 230 in not that fast of a time with no seemingly inhuman hole recoveries which are supposedly trivial as a result of the memory palace method of playing. I have no doubt that that video is legit, but I find it hard to believe that it's the same person playing in that video and playing in the video of either 16 grades in the M-roll or the Phantom Mania MM. I understand that a game played on short notice under maybe not the best circumstances is not going to be amazing, but if someone asked me to whip out a Ti Master M I bet I could do it within 20 minutes; I wouldn't post up a video of me slowly topping out at 500 and saying "well, the circumstances are not the best for me to play right now so this'll have to do".

Anyway, I'll let others draw their own conclusions. I don't like to accuse people of being cheaters; I would rather let them prove it themselves.

 

Rosti: I will expose tomorrow (1AM now)

 

I'm no memory expert and I don't use nullpomino so I don't have a strong opinion, but I want to remind people that QM is a dude who came up with a significantly improved "playing forever" strategy that deconstructs 7-bag into the tightest possible loop. As the guy who made the first formal proof of playing forever, I can say there's no way to cheat it. It's basically math. Do you really think a person who has established that they are smart, dedicated, patient and can legitimately achieve great things would waste his time trying to troll a bunch of Tetris nerds?

Even if warning flags are going off in your head, please treat him with respect and ask questions wherever things don't make sense to you. If your suspicions are right the truth will be revealed soon enough, and if you are wrong you will get some amazing answers and avoid making yourself look like a douche.

 

One by one, most important first.

I'm a working physics grad student and this is exam week. The absolute earliest I'll be able to do anything Tetris-related is this weekend (and I make no promises that I will be available then). That includes playing Phantom Mania, recording, and unless there is divine intervention, posting further responses.

Obviously K thinks/thought I frame-stepped the games in the videos, which completely trivialises invisible Tetris. This is why I'm concerned about it.

[Randomiser has been addressed.]

Under the circumstances it was critical I make a video demonstration ASAP. With no explicit directions (get an MV, downstack 18 rows of holes) to follow, I just assumed quite fairly that anything goes. Hence the ragequit-230. Priority #1 is proving "conclusively" that I can play invisible Tetris. If everyone's good with my slapdash setup then I'll use that for future videos.

I wouldn't describe it as "ain't no thang." It's a complicated task, but it builds on my existing visualisation system, it doesn't replace it. Something that takes 3 months is decidedly non-trivial, especially considering I put much more time into it than I should.

Now, contrast this to ARS, which Kitaru (an expert in ARS) has explicitly stated "takes years, not months, to master." Perhaps I shouldn't have believed him and it takes less than 3 months to master (doubt it), but that's beside the point. Mastering ARS after having mastered the vastly more prevalent SRS is very much like mastering Dvorak after learning the vastly more prevalent QWERTY keyboard layout - it's very difficult to be good at both, you'll accidentally do something corresponding to one when you're using the other, and so on. So I assume that learning ARS involves replacing my muscle memory for SRS - which has developed not over the course of many years of playing not-TGM, long before I got serious about Tetris.

I have maybe a few hours to play Tetris in a given week, mostly on weekends. If I could grind away hundreds of games every week I might be improving faster still. The only reason for my rate of progress being fast is that I can already play Tetris reasonably well and I just need to apply it to invis using this system.

I dunno. He seems more focused on my psychological pathology of being a liar. It is apparently okay to commence your analysis of a suspicious player with "you are literally mentally unhealthy" and only much later point out apparent conflicts.

I'll accept that - it's clear to me that K is convinced beyond all doubt that all my videos past and future are cheated, as well as anything I write to explain myself. I will continue to address points he makes that are valid to others, but I've stopped trying to persuade him.

I think a faked real-time game would be very difficult to make not-obvious, let alone convincing, by video editing. (I'd love to see counterexamples though.) As for coding, I obviously cannot prove I'm not good at it because of that whole "playing-dumb" problem. But FWIW I wouldn't have posted this very simple request on the HD forums if I was halfway-decent at coding. Believe it or not, up to you.

Thank you for the offer, but there's this damn pond in the way.

Unpleasant it certainly is, but hopefully the discussion stays civil.

You also explicitly parenthetically said

which could not have been more vague if you tried to make it so. That was my cue to take the path of least effort that would still prove I wasn't using a visible replay or other unfair assistance.

Had you asked for a downstacking sequence, I would've provided one (though definitely not within a day). Had you asked for a good game, I would eventually have pulled it off (provided enough time). But,

considering that you're satisfied with my setup, I'll use it to film future videos that meet your requirements.

You've also spent 9 years mastering a game which is the embodiment of consistency in performance. I've been playing Phantom Mania for that number of months, using a setup where the "Retry" key is within reach of my right ring finger and it takes < 2 seconds to reset a new game. My PBs are and have always been miles ahead of my averages.

The thing is that, as a video maker, I have the power to sort out my best gameplay footage when I upload videos. The mediocre performances never cross my mind as potential uploads. That definitely has not helped my consistency.

Yes I'd much rather have filmed an MK than a piddly 230, but you didn't specify you wanted a good game - you literally said any game and are now accusing me of not meeting requirements that were never stated in the first place. A longer run, a faster run, or an N-row dig are very specific things to ask for. Why didn't you ask for any of them?

Now for my rationale. My priority was to make a video, fast, that demonstrated something important. K makes a good point that a fraud cannot produce evidence in a short time, as fakes need to be crafted carefully. That's the only reason I put this video out yesterday. (That being said, non-frauds who are not incredibly consistent can't do it very easily, and it is a very inconsiderate thing to demand from a person who, again, has significant time commitments in their personal life. I won't honour such a request a second time, full stop.)

The fact that I'm not frame-stepping was supposed to be demonstrated in the M-roll video, but I obviously blundered there. So this was kind of a test for my setup ("Did I overlook anything?") before I make another video with a similarly stupid mistake that makes me look suspicious. Based on your wording I assumed you just wanted to see that I could play invisible Tetris (and not mash pieces, like most players). So I did the easiest thing I could that demonstrates it - stacking for Tetrises. Of course, I also executed a difficult SRS no-overhang triple S-spin at 109 in the hopes that it would demonstrate my understanding of the stack surface, but nobody seems to have pointed it out or, for that matter, even noticed.

Finally,

Thanks. I'm glad to see at least one vote of confidence (this is getting exhausting). To segue into a point I'd have a hard time bringing up otherwise - when I first wrote up the 35-piece playing forever loop, my developmental posts were similarly detailed and in-depth. My characteristic essay-posts are not a byproduct of some psychological affliction, I'm just verbose - that's just the way I write. (Sorry.)

I can't promise amazing answers, though - I topped out midway through a fast secret grade just because "it fit the video." I didn't downstack holes because stacking cleanly was easier.

 

I see a lot of skepticism and doubts in there, but has anyone actually tried to practice this technique ? You won't count on me for sure (I didn't understand a word of how this works haha), but I think invisible clearing doesn't seem that far away... there are people that can remember an absurd amount of digits of pi, that is a more "inhuman" feat than remembering where some pieces are locked =p

 

ARS takes several months to master, but I'd also say that with several previews, Ti wallkicks and a hold function then mastery is far from required. ARS under TAP rules is a harsh mistress, but under Ti rules it's not a huge deal. And as someone who used to regularly switch between ARS and SRS and played both to a very high level, I'd say switching between them isn't really all that big a deal once you're pretty familiar with the two. I have some issues with it these days, but I basically go six months between periods of playing.


Onto the other stuff (and to an extent what c_t says), I think people are pretty accepting that there is a reasonable amount of skill there - I think the main question is whether it's been embellished or exaggerated. And expanding on m.kevin's comment I think people should hold some caution, especially when saying things like these have to be "superhuman", if they've not spent a while seriously attempting these sorts of skills (though comparisons to other people who have are valid evidence imo). I don't think many people on here would claim that our collective Tetris abilities are particularly superhuman, and would concede that most people who bothered to invest the time actually trying to be good at Tetris would get to a skill level like Death M or sub-35 40L eventually. And yet as I'm sure we've all experienced, to your average joe on the street with little understanding of high-level Tetris or 20G play, people playing Tetris at merely sub-minute 40L pace seems beyond belief and superhuman, let alone stuff like sub-20 40L or Ti Gm. Nullpomino is a game not a lot of people play, and within that group of people not a lot of them play Phantom Mania seriously, so I think the pool of people who have a level of understanding that can directly critique this is pretty small.