Freeform Invisible Tetris (Phantom Mania Cleared)

Well let's make things clear. As i said QM certainly have some Tetris skills but nothing equivalent to the initial subject of this thread. He apparently designed some endless SRS 20G that some people are grateful but i didn't read about it since i have no interest for the subject. About past/present/future videos, surely some are true, and surely some are fake. As mentioned Rosti, the first time i saw one of QM's video, the bullshitometer instantly triggered an Ghettoish alarm, but Steadshot, Qlex and Rosti convinced me to keep the alarm in silent mode, and really consider QM's claim so i took a closer look and read this thread..

Again i'm thankful to QM to have introduced the "Memory of Loci" to me. The first time i read the thread i didn't understood anything of his explanations. so i step aside and read "Moonwalking with Einstein". I was so enthusiast that i decided to really seriously consider this new field. i analysed lot of articles and especially those about advanced memorization techniques for competition (memorizing digits, binaries, dates, shuffled decks of cards...). I'm glad i was able to met a competing "mental athlete" around my place who give me some good hints and i'll go to Paris watching the French Open Memory Championship in November..


So, with far more understanding of the Memory of Loci concept, i came back to QM's Explanation.. More i was reading and less it was making any sense and mostly deceitful about what you can do with Memory of loci. i was very annoyed but more i dug about QM's post and videos, more it was evident that we were conned.

When i posted the "funny" explosion gif it was more ironical, because at that time i was pretty convinced. But i asked more experienced people about Memory of loci field.. and as i was expecting their conclusions just follow mine. I wanted a confirmation about how many loci he can "write/erase" per second.. As i wrote on my last post, there is no need to argue about video's veracity if nobody can't even reproduce or train with his method even very slowly. he's basically demonstrating a pseudo scientific "water engine creation" with a rocket car on track video.


Since of course i was asked to produce counter-arguments, i first need people to understand what is "Memory of Loci". How it works, what prodigious things you can achieve with and presently, what you simply CAN'T.

I'll try to keep things as simple as i can but in one sentence : The Memory of Loci is a visual memorization technique based on the association/attachement of element of information to familiar/well know place. It use 2 unconscious capacity of the brain :

1. Spatialization (loci)

Your brain is fucking efficient for recalling place. For instance you visited once some acquaintance's apartment months/years ago. You don't remember what was his name, what you were doing there but you surely have a good memory of his apartment configuration. Not particularly everything but how was the room disposition. where were the main furniture, the toilet, the kitchen and so on. Why are you able without much effort able to mentally walk in this place even if you didn't make any effort for it ? Thanks to your caveman's ancestor dudes ! only the one you who don't get lost on that fucking big cave survive. And same for those who go out and need to get back ! it's the same for any species in this brutal world. Darwin's rule !

A memory Palace is simply a mental exercise about recalling a journey path (first person) in a place you know. This place can be your apartment, neighborhood, school, work, body, video game.. anything as long as you can travel through it. In this sequential path you must decide point of interest (loci) where you will store information.The very first time everybody use his own apartment. for instance for me, i imagined myself watching my bed (locus 1) then i watch my windows (locus 2) then turn and watch my desk (locus 3), then i get out of my room to the toilet (locus 4) and so on.. You can easily create a path of 20-30 loci the first time. Just enough to demonstrate how it work


2. The imaging Brain

You probably think your memory is shit sometime, you don't remember were are you key, why you opened the fridge, what is this book on your shelter about ? did you even read it ? Actually you might think you are dumb sometime or you brain is shit but without being conscious about it, your brain a prodigious image recorder.. as it is described in "Moonwalking with Einstein" during your whole life at any instant your brain keep recording continuously image to is neuronal network. Unfortunately into this vast network of information the brain don't follow any kind of linear logic and we can't neither sequentially search them nor browse them. the most natural and effective way is to add emotional cue point to those images. As Joshua Foer explain in his book, you can probably remember what kind of breakfast you got this morning or yesterday or the day before yesterday but more you try to recall back more it's blurry... those breakfast go into the vast network of the everyday boring breakfast that you don't need to remember.. But on the particular day of 9/11 at that particular time he perfectly recall what he had for breakfast without any effort. So are the same for all souvenirs : your first kiss, your first GM , when crazy shit happen like someone smashing some other guy's head with a Casio keyboard.. all those trivial explanations to point that you must create emotional and vivid images to get better recall. The first time you test the method, you use something easy to apprehend like a shopping list. write a list with a number of items equivalent to the loci in your first memory palace. Start your mental journey by following your path an place those item one by one. To make them more memorable you must use your imagination to make the image grotesque, fun, sexualised (human are good at this), gruesome etc... take your time to imagine the image, don't hesitate to make it interact with the locus. For instance the first item of my list is a sponge and the locus my bed : i imagine a big giant wet bob's sponge soaking all my bed... once you've finished with all your items, You just have to get back to the beginning of your memory palace and you will recall all the item in the sequential order without trouble !!!



So as i wrote before, The Memory of Loci is a visual memorization technique based on the association/attachment of element of information to familiar/well know place.
Without any advanced optimization method, in it's most simplest form, Memory of loci allow to memorize directly in the long term memory very long list of sequential information stored as image and retrieve them efficiently :
- You Memorize (write) information by Following your path in the previously defined order.
- You Retrieve (Read) information by Following your path again.

When i wrote storing information you must understand that your brain recording device can't handle "abstract" things like numbers, figures in a deck cards or a tetris playfield.. poetry is one of the most difficult exercise for instance. Now storing bananas or toothpaste from your shopping list are concrete object good enough for a simple demonstration but how can we store something more useful like a long list of digit ?
A simple example (for the demonstration) is to create a list of 100 object and assign them a number from 0 to 99. If you want to store a friend's phone number you can just create a memory journey of 5 loci in his apartment and store the 5* 2 digits that way (google major+system for more information). Since creating an image consume time, from a competitive point of view, storing more digit in one locus is mandatory ! So here come all methods from all thoses differents champion. They fundamentally don't change how Memory of Loci work. They "just" create a layer of compression in each image. For instance the Person - Action - Object (PAO) method require 3 lists of 100 items and allow to compress a number from 0 to 999999 in one locus. for instance a number like 748628 would be compressed into Person n°74 doing Action n°86 with object n°28... it create some strange and unpredictable bizarre image that can be stored on the fly.. (eg Einstein Moonwalking with a microwave).

Those are simple examples but as you already noticed you first need to remember a lot of things before being able to remember anything..

The supreme discipline in competition is the shuffled deck of 52 cards. The 2010 world record famous video is 21.90sec. demonstrate how fast someone can remember ~1.7 locus per seconds but it still need time to retrieve and decompress the information (5min).

The worst scenario for a mental athlete is to "forget or confuse" the image and stored information because it was just too common or mixed with a previous stored item for instance. That's why a mental athlete don't use the same Memory palace without cleanly erasing it first through a mental process. In competition it's very risky to use a previous used Memory palace because they are too much risk of mixed parasited image... Serious Memory athlete use hundreds of memory palace...

From what i explained, i hope you have a better understanding about how really work memory of loci. Remembering digits of Pi, is indeed impressive but it's a process about learning it first then recite it later... nothing compared to Tetris but let's get back to business...

20 Row loci ok, but what mean “How many hole” ? “how many” need to be compressed into an image. In the most simple case of your explanation, you can’t just write a number from 0 to 10 because it’s too abstract. But if you store data as bit information it still look like “1111 1111 11” converted into decimal mean 1024 different possibilities because hole and block can be mixed in the row. Of course you can compress it but “how many” is vague in the first place. It become worst for the column loci with 2^20 possibilities… But even without using memory of loci with a simple paper and pen nothing make sense about how you compress/redraw a playfield content. At least with “20 row and 10 column” I can at least draw a frame...

There come the compression !! but wait if you just write A-T and 0-9 “single digits helps”, it’s not compression ! simple letter and digits are abstract things your brain can’t directly record, remember ?

For an human point of view and the decimal system shifting 10 block in an array is “easy to understand” but this paragraph is a complete nonsense (saying it politely) in the memory of loci…

No popcorn but aspirin for me plz

This paragraph make sense in Memory of loci but as QM confirmed, he deliberately use 20 times identical rooms. Unfortunately for him, it’s exactly what you must not do when creating a memory palace if you don’t want to get lost or confuse item from one room with another !

So the Bin is the row loci. But am I supposed to understand with this magic 8 ball encoding mumbo jumbo ? so 2 or 3 magic 8 ball resume 2^10 configuration ?!?

Wait ! what ? a copy of the rubbish bin row loci from each room ?!? how the information is supposed to “duplicate” here ?

If each table is a locus.. Putting several people in the same image with a meal indicating his height…. Just make me think about the painting “the last supper” from Leonardo da Vinci.. definitely just too much information in one locus without compression to be credible.

So you confirm you prefer manage 10 to 20 object in each locus.. this is complete nonsense for “any memory athlete”. 10 to 20 ? why not 30 or 40 ?!?

If you don’t know what any of the displays say ahead of time, how can you explain the information is “magically” popping through light button with displaying letter paired with number?!?
And combination trigger alarm ?! how convenient !
But sorry nothing related with Loci. Remembering information need an active mental recall process.

I don’t understand this justification now.
Talking about 19 different possible orientations among all pieces don’t change the fact that you still have to write to 4 loci arbitrary or not..

Where is the fridge ?
Again memory of loci require vivid image. All those simple toggling information in 20 identical rooms just lead to confusion. This is strongly not recommended for any advanced usage.

How convenient to switch 2 room in your palace Memory. But sorry, this is simply Impossible. You bend the rules so hard that I even hardly doubt you really understood anything about the subject :/

Now in complete back pedaling… Cheating your own system ? hmmkay

Sooooo !
QM is using the memory of loci like a Random Access Memory, jumping to whatever place he need to store information and retrieve it, conveniently compress information as he like, redefine his memory palace on the fly, retrieving information in a control panel, playing hardcore Tetris at the same time. Complete fairy tale ! It's not how it work ! and you simply can't read/write/erase the way he claims .
Sorry dude you can't use Memory of loci for invisible Tetris. Even the finest memory techniques like Ben Pridmore for instance can't help you because of the underlying way of storing/retreiving the information.
There is no need to argue about the veracity of his video.. the whole theory is nonsense in the first place.. now he's just backpedaling and trying to prove the veracity of his supposed skills but he crossed the line.. a fallen angel pathetically wading in the mud.

(minor edition.. Loci is the plurial of locus)

 

Don't worry guys, if this technique doesn't work then we still have this.

 

Well, you certainly didn't wait for me at all, did you?

I'm not in the least bit surprised you decided to attack the method itself. What does surprise me is that you have demonstrated to everyone (including me) that I'm not using the method of loci, at least not in the conventional way. This seems central to your argument, so I'll address it using information which I've mostly already posted.

Let's go all the way back to the start. I began looking for a memory technique that looked like it held promise and might be applicable to invisible Tetris. I stumbled upon the method of loci, which at the time I'd only heard of, and "researched" it for about a day. After I felt like I had the general idea, I set to work.

I've mentioned this before - I worked alone in secret. I was not in touch with memory trainers or mental athletes, nor did I do further "research" about the method. I had not actually read much from "Moonwalking with Einstein" beyond the first day, though when colour_thief brought it up I echoed the recommendation nevertheless because I knew it was related to memory palaces and thought it could be relevant. As I explained, from time to time I would look up something via internet search, and maybe use it if I felt I understood it. (Contrast this to researching it for a month from a reference book and professionals in the field.) As a result you now have a rigourous understanding of the method of loci - how it really works - and I a distorted view, without ever realising it.

I did not expect my approach to agree exactly with the method of loci (all of the clever choices which give rise to it, I would have had to make on my own.) In biology, mutations occur when there is a copying error in the genetic code. Likewise my method should be considered a mutation of the method of loci. Some mutations are bad and some are good, depending on if the copying error gives a net advantage. If one accepts your proof that the method of loci (in pure form) cannot be used to play invisible Tetris, then I will say that my error led to a "good mutation."

If you don't want to call it "method of loci" anymore, fine. My goal was never to learn the method of loci exactly as professional mnemonists do - my goal was to learn a method that works for invisible, beyond what direct visualisation could give me. And I did. Now I'll consider some of the discrepancies you've raised.

Identical rooms. Did I say these were a bad idea? Yes. Did I need them to be identical? Probably. It was a lot of extra work and a hard decision to make, but I knew it would be necessary so I did it anyway.

Rapid encoding/decoding of information. Maybe if you have a system that has hundreds of different images, it's hard to quickly write to and read from a memory palace. But I have very few images - the different states of loci in a room, 10 different things in the bins, and 20 different foods. One of the first things I did was to make reading and writing those things second-nature, which you can do by brute-force and lots of repetition. It's not hard. And don't tell me it's impossible - I return to blindcubers such as Marcin Maskow and Marcin Endrey, who have demonstrated before official judges that they can translate a Rubik's cube into 20+ images and memorise them - in 8 seconds total - and decode the information on the fly, all the while solving the cube. Do I need to provide more examples of precedent?

And don't tell me that you know how my brain works better than I do. If you have a hard time efficiently recalling information from a control panel full of lights and buttons, I'm so sorry. I don't share your difficulties.

I shared my accomplishments with the community because I wanted to demonstrate that much more is possible than previously thought. It's hard to overstate my disappointment.



In total this has cost 30+ hours of my time and, as I predicted from the very beginning, has gone nowhere. This is why I did not want to start an argument. The only reason I made any efforts to defend my achievements was that I was misled into thinking it would be easy to prove...

...and having been so addressed, it would've looked more suspicious for me to adamantly refuse to comment. As the weekend is drawing to a close, I will upload a video demonstration and replays. This one should more closely reflect what you were looking for, and in the future there will be still more. (If it's still processing, check back in > 2 hours.)

But I won't engage further arguments claiming that what I do is "impossible."

If I describe in detail how I memorise the playfield, my memory technique is attacked by a proof that 1) I'm not using the actual method of loci and 2) the actual method of loci cannot be used to play invisible Tetris - which incidentally is not a disproof of, and in fact is perfectly consistent with, my use of a slightly different method which I have inaccurately been calling the method of loci. (Maybe "RAM" is a better name - I mentioned ages ago how parts of it are very similar to RAM in their workings. Maybe you prefer "faery/fairy tale." The name means nothing to me, I just use it.)

If I argue why my method works by bringing in past information, you will counter by arguing why it differs from the method of loci. If I introduce any of the numerous complications I have yet to bring up, you will accuse me of backpedaling. If I make a video demonstrating something which is plausible (but difficult) using conventional invis strategies, I will be accused of exaggerating my abilities in videos past, despite stating in the OP at the beginning that they were significant outliers. And if I make a video demonstrating something thoroughly implausible by conventional methods (in trying to vindicate my very specific method of play), somebody will accuse me of faking the video. You might say that the only option I have is to perform before a referee, but 1) I have no capacity to arrange an in-person demonstration with a third party whose word would be accepted here, 2) even if I could, thanks to your scathing arguments nobody will be willing to referee for me, and 3) even if they did, I'm sure somebody will come up with a similarly convincing explanation for how I "faked" an in-person demonstration and we will be right back here again. And beyond that lies the territory of unfalsifiable theories - arguments from either side which cannot be unambiguously disproved.

If anybody who watches this thread has made it this far, I hope you at least see that what I claim to be able to do is humanly plausible. Beyond that, whether or not you think I did pull it off is irrelevant to me. For those who are on the fence, I will continue to make live video demonstrations (on my own time) on YouTube alongside my quest for GM. I cannot prove they aren't fake, but I am willing to go to some lengths to make them as unfakeable as possible. And if one day, against all odds, I am able to arrange an in-person demonstration in front of somebody you all trust who is also able and willing to be a witness, I will be sure to have a camera ready. But If you still don't believe I'm capable of this level of invisible gameplay, that is no longer my concern. The most credibility I can hope to salvage at this point is not worth my time to continue this endeavour - or rather ordeal. I might just care if Tetris were my whole life but it is not.

Oh, and for the record.

At no point did I ever claim that any of this was easy.

Don't bother - I'm not coming back anyway. Thanks for nothing.

 

Hey QM, I can be your referee if you want. Also your video is private, could you set it to unlisted please?

 

I should've made a longer post before, because it seemed clearly I didn't want to get involved in drama. Well I still don't, but it won't stop me from giving my two cents =p

There is really one major issue that puts a question mark in question_mark's videos (haha I'm so funny) : are those videos cheated or not ?

Well my response is no. I said that at pretty much every video, and any raised skepticism didn't stop me from validating the contents of each video, especially since QM has uploaded it's latest video on that, because clearly, the hands match the movements in the game and they're at human speed, there is no mods, and if we're gonna assume he used a second screen to pull that off, well the only second screen I can see is behind and clearly turned off.
We could also assume he made a tool-assisted replay, then tried to match the hand movements with what's happening in-game, but then again you get to see him literally turn on Nullpomino and immediately play a new game, which nullifies that argument.
So far, Question_Mark was patient enough to upload four proof videos. The only proof he refuses to give is to move just to prove his invisible skills, which I fully understand, because quite frankly you will never make me move either =p

That's my opinion. I see no room for computer magics and trickery here, we can debate ad infinitum about the methods used and the practicability of said methods, if the proof posted above isn't directly challenged, that means it is decisive, conclusive and exonerating from the main accusation QM is facing.

All in all though, this whole thread made me want to fire up TGM3 again, so thank you QM for that o/

 

I'll just comment that part for people, but all the rest is trivia backpedaling

Refering and comparing Blindfold speedcubing with QM claiming performance is still irrelevant. Blindfolded SpeedCubing requires an amount of time for memorizing the VISIBLE shuffled cube configuration, THEN receiting it. There is no "decision branch" to take when you are physically solving a cube blindfolded. Of course i'm not lowering the performance of blindfolding speedcubers.. and as mentioned they are performing with "official judges"
Again Memory of loci if used for storing/retrieving information sequentially... Using the brain like a so powerful and fast Random Access Memory like QM claim to describe (putting invisible tetris aside), is "unlikely" to happen because it simply contradict the 7±2 elements the short-term-memory capacity according to Miller's law. if any art of memory method would have allowed mankind to bypass this limitation : I guess we should have known by now..

Don't worry QM, it's just an online forum. I'm sure you don't have any acquaintance from here IRL. You can just walk away peacefully without tarnished reputation risk anyway..

 

It drives me crazy how you let your distal interphalangeal joints collapse when pressing the keys.

 

(relevant paper)verbal reports on mental processes: http://deepblue.lib.umich.edu/bitstream/handle/2027.42/92167/TellingMoreThanWeCanKnow.pdf