/ game theory

My Problems With Game Theory (or why you really need to cite your sources)

Ok, this is going to be an ongoing document as I gather research on the topic.

For those who don't know, Game Theory (not to be confused with economics or animal behaviour models) is a channel on Youtube who's focus is to theorize about video games and the like, often using real-life research and science. If you have seen any of my videos in LittleGadget, I think it should be clear that I've been inspired by this work. Created by Matthew Patrick (otherwise known as MatPat), the show has a good history of providing really compelling ideas and theories, and teaching and inspiring their audience about science. Which I am all for.

However, there is one thing about MatPat's work that has continually bothered me and has been growing into quite a big sore spot. Matpat's work has come under criticism before for various reasons, whether it be the topics he covers or the theories he proposes. But my problem has nothing to do with how many times he's covered Five Nights at Freddies, or his crazy attempts at Undertale, or how his videos are long. I am actually ok with all of these things. No, my problem comes from something much, much deeper in the core of the show.

MatPat, please. Why don't you cite your sources?

For a lot of people, this probably seems like a non-issue. So what if he doesn't properly reference? These are just YouTube videos, who cares? But as a researcher, it becomes increasingly frustrating when he talks about ideas or research concepts while not providing any way to fact check his claims and explore further. And it becomes downright infuriating when he starts making claims that I'm fairly confident are ludicrous and there is no way to see where the hell he got the idea.

The fact of the matter is, any time you reference ideas that are not your own, that fact needs to be clear. It needs to be obvious when ideas are your own, and when ideas are coming from other sources. This is true whether you are discussing what others have already said online, or experiments and results from primary research articles. Furthermore you need to make it clear what those other sources are. If you got the foundation from your theory from somewhere else online, you should display the author. If you are discussion an experiment, you need to display who ran that experiment. Any time you are talking and its ambiguous where an idea is coming from, whether its something you read online, or found research on, or thought of yourself, thats bad.

And don't think you need to be exhaustive with this. I mean, it's entirely possible that someone else on the internet has posted an idea similar to your own and you couldn't find it. That's fine. But that doesn't mean you shouldn't make the effort to see if someone else has already reported or talked about something. Even if you came up with the idea independently, if someone else has posted a long discussion on reddit about it and you can find it, it's irresponsible not to report that.

But the bigger issue for me is absolutely when talking about scientific research. You don't need to cite core fundamentals like Newtons laws of motion, but you SHOULD be citing current research and ideas. And the fact that you don't isn't just frustrating, it's disappointing.

This article is an ongoing work, as I want to go through a number of videos from Game Theory and the sister channel Film Theory to build up examples. Below are notes from the videos as I watch them.

also I realize the irony of not providing references whenever I rebute some of MatPat's claims. A lot of this I'm writing as I'm watching the video. In the future I will write (or make a video on) a formal response to his ideas/theories where I will then provide evidence. Consider yourself disclaimed.

Evidence:

Film theory

Film Theory: Pickle Rick ACTUALLY WORKS! (Rick and Morty, Feat. DAN HARMON!)

Sept 29 2017 | ~ 3.1 million views

  • 6:28 Awesome, you showed a screenshot! I mean, you named them 'some scientists' when their name was right there, but thats ok. Also, no citation, but I can track this back.
  • 9:25 "Another study" ok, here you show a screenshot of an news article talking about the study. Hardly a primary source, why not show a screen of the original article like you did before? Also, no citation again, and here it's a bit more serious.
    • Side point, but lets talk about how you summarized the study. Your first sentence you said "two subjects had their brains linked through a system of electrodes". Just looking at the screen shot, that's not the case. The subject on the left is wearing an EEG cap. Yes, they're ultimately electrodes, but they're placed on the surface of the scalp. They get very rough voltage information. The second subject is hooked up to a TMS machine. This basically produces a strong magnetic field that can induce electrical current in your brain. You can either use it to stimulate or inhibit signal. It is highly localized, so that TMS is already positioned to the part of the motor cortex that controls the hand. I mean, I get the connection you're trying to make, but you summarized and stretched the methods of that study way way way out of bounds. You can use EEG to train simple commands that a computer can read. And then you can use that computer interface to turn on the TMS that an experimenter aimed. This is not 'linking brains with eletrodes'.
    • Maybe that picture wasn't representative, and I should go look at the full study. But then, you never told me what the full study was.
  • 10:40 Ok woah woah man, you're showing all these figures. Even a graph. Where are these coming from?

How Much is YOUR SOUL Worth? (Fullmetal Alchemist Brotherhood)

Sept 21 2017 | ~ 1.2 million views

  • 6:50 Dunbar's number (Robin Dubar), and the other anthropologists estimates

Film Theory: The Star Wars/Valerian CONSPIRACY

Jul 18 2017 | ~ 1.7 million views

  • Valerian vs Star Wars, did you actually read through the comics and find this info yourself? no, you probably looked through countless articles online. Maybe it's hard to cite the first 'source' of these claims, but you should clearly indicate you didn't make these connections yourself.
  • 12:38 Althrough I guess the quote at the end kind of justifies you here?

Film Theory: Can The Lion King SURVIVE on Bugs?

Jun 21 2017 | ~2.8 million views

  • 3:24 Lions live within 100 square miles

Film Theory: He is LYING! | Better Call Saul's Phony Disorder

Jun 8 2017 | 1.5 million views

  • 4:50 UHHH Why did you show screenshots of what I presume are reddit posts while removing the authors?! Did you actually go out of your way to not cite people???
  • 6:53 this is ok, since you show the source right here. nice
  • Then you show a research paper. This is ok.. I mean, a lot of the information IS on screen about the source of the claim, so good. But if you want to be proper about this, you really should include a text citation. Either in the description, or on screen.
  • Then you talk about project nemesis... Uh where is this info coming from? Have they published anything on this? are you just making this up? is this all from the previous paper you showed? This seems like a pretty robust experiment
  • 8:36 "7/63 performed better than chance". This is no fault of your own, but actually that result should be expected. Statistics basically employs a 5% significance critical value. When you hear somebody say 'they were statistically significant', they're saying that there is a less than 5% chance of getting that result from random chance. That also means if you run 100 samples, about 5 of them would come up significant. Just randomly. z
  • Ok you read a long paper. Whats the paper?! There is no way for me to know what this paper is and to look it up myself...
  • 10:41 you're showing a paper here, what paper is this?
  • 11:00 nocebo effect. Reference?! where are you getting this info?
  • 13:30 There IS a nemesis paper!!! What is it! Just, just move show me the top of the page. cmon. just a name, a name!!

Film Theory: How to KILL DEADPOOL!

Feb 25 2016 | ~ 15 million views

This one made me the most upset, I've actually been thinking about tackling this problem since this specific video, because MatPat studied Neuroscience in his undergrad! This should be where he can show his knowledge and put that degree to work!! But his idea is so, so crazy. And wrong. and this has 15 million views!! MatPat, what have you done.

  • 6:00 Memories... here we go.
  • Premise is that you cannot regenerate memories. Reasonable hypothesis, I would even agree, but lets see why he thinks this.
  • 6:15 True, they aren't stored. I even said so in my own video.
  • 6:30 'Memories are signals' Aaaaaaand you lost me.
  • Ok, his explanation of synapses and action potentials is good. And yes, a memory is essentially a specific chain reaction of signals...
  • Yes different patterns are different memories, I'm still with you..
  • 8:05 Ok I think I see what's happening here. His premise is mostly correct, but he's missing some key information. Yes memories are essentially networks of neurons, the very same network of neurons that processed the memory to begin with. But how then do you retain that network? MatPat is making it out to seem like if this signal stops firing, you'll lose it. But this isn't the case. That network actually have a physical connection reinforcing it. As has been said by Dr. Carla Shatz, 'Neurons that Fire Together, Wire Together'. Which is essentially summarizing Donald Hebb's concept of neural learning. When neurons in the same network, or near each other, repeatedly fire in unison, there is some growth or metabolic changes that connect them in such a way that the probability of both neurons firing when one fires increases.
    • The takeaway of that is that there is a PHYSICAL BASIS for memory formation. It's not just a signal.
  • 8:30 MatPat does get this right, that when you experience something a network activates and its that network that forms the memory.
  • 8:40 I'm sorry what? You need to reinforce memories by experiencing it again? Uh, no, I think we have a lot of experience forming memories of things that we only experience once. We have this thing called a Hippocampus that facilitates learning. Basically, it networks or interfaces back and forth with a lot of your brain, including your sensory areas where memory networks are mostly going to be stored. So it's capable of pushing something into long-term memory without you needing to reinforce the networks yourself through repetition. In fact, simply repeating what you want to remember, either by repeating actions or by rehearsing (or commonly by re-reading and memorizing text) is a decidely BAD way to put something in memory. Very ineffective
  • 9:00 WAIT WAIT WAIT WHAT, DID YOU JUST SAY THAT YOUR BRAIN STORES THE MEMORY PATTERN IN YOUR HIPPOCAMPUS
  • DID YOU JUST SAY THAT NEURONS BEING MOVED PHYSICALLY CLOSER TO EACH OTHER MEANS THE NETWORK GETS STRONGER BECAUSE CHEMICALS DONT NEED TO TRAVEL FAR?!?!?!
  • MATPAT WHY WHAT
  • 9:10 Okay uhhhh I have no idea where he is getting this, again I think he is trying to say that memories are only the signal, and the signal needs to keep firing. Let me illustrate why this ISN'T true. If this were true, then you remembering something would mean that you are constantly thinking about it. Remember, memories are stored in the same networks that were involved in processing them. I think he agreed as much. So if you see a red ball, there is a network in your visual cortex that will fire specifically for that ball. That network is the memory. Anytime that network activates, whether you see the ball or think of the ball, you will manifest that visual in your consciousness. Or awareness, however you want to say that.
    • So here MatPat is claiming that to remember things your brain needs to constantly be imagining them. Think, for a second, of as many memories you can remember. Now imagine having them all constantly firing to keep them 'in your memory'. We don't do this.
  • 9:26 See, in essence the way the memory is stored isn't intrinsic to the neuron, fair. But there is a physical basis to it. 1) new channels can be added or others can be blocked at the synapse making it easier or harder for one neuron firing to activate the next. 2) new connections can be made between two neurons that maybe weren't directly connected before, facilitated by your hippocampus.
    • To illustrate this, Say you have a memory of your 10th birthday. this memory is essentially a collection of networks that you can group together. It's not a perfect recreation by any means, and in fact any act of remembering is a game of fill-in-the-blanks. Anytime you remember something, you are recreating it. Now this memory of your 10th birthday is not one you think about regularly. Actually, it's probably been years since you thought of it. It has simply not been activated in a long time. Those neurons that make up the network may have activate for other things, sure, but the specific network has not fired completely and it has not elicited the memory. Now, reading this, maybe you remember now. You remember the cake, your friends, your family at the party. Despite not being fired for years, you are capable of recalling something that happened long ago! Now again this isn't a perfect recall, you're actually probably making a lot of this up. But at the same time, if you go back to corroborate the memory with evidence, you will probably find you are mostly right. You are, in fact, capable of remembering to an extent. Despite, again, this network remaining inactive.
    • To summarize, no there is no 'place' where memory is stored. A memory isn't a physical thing, it's a network of activation. BUT, there is a physical connection that binds this network together. There IS a physical basis. The memory IS NOT THE ELECTRICAL SIGNAL ITSELF.
  • So because there is a physical basis, both in network and the strength of the network, I don't see a reason to assume that you wouldn't be able to regenerate this (assuming we're talking about a fictional healing factor that restores you as you once were, same musculature, same age, same everything). If this were not the case as MatPat is claiming, not only would you not have any memories when you regenerate, you wouldn't even have any brain.
  • Imagine you had some complicated circuit of wires all plugged in. Now unplug them all and toss the wires on top. That's now your brain. It's not just your memories, its any and every function that your brain has developed over time. All your brain is is networks. Without networks, there would be nothing to fire. You'd have no motor control, no visual processing, no cognitive abilities, nothing would be networked.
  • So if you could regenerate the network that the brain was in before, there shouldn't be a problem.
  • 12:30 Why the frontal cortex for memory? Also if you read research on H.M., hippocampal damage leads to anterograde amnesia, which is the ability to form new memories. You might also lose some recent memories that haven't made it to long-term status. But you wont lose all your memories, since those are stored in your sensory systems.
  • And now, after ripping through ALL of that, I want to ask you MatPat. Where did you get all of these ideas? I saw no image of a paper, no reference of a name, no research, no nothing. You made a lot of claims, many many. Where is your evidence? If you provided evidence for your assumptions and ideas, left a citation, anything, I could happily go and explore further to see where you're coming from. Maybe your idea is more substantiated in research than I realize. But as it stands you have spewed a lot of nonsense and I have no idea where it came from.

The SCIENCE! of...

How The Metroids DOOMED Us All! | The SCIENCE!...of Metroid: Samus Returns

Sept 21 2017 | ~ 1 million views

  • ATP eating bacteria

Can FNAF Kill You IRL? | The SCIENCE!...of Five Nights at Freddy's

Sept 10 2017 | ~ 1.8 million views

  • You immediately go into fight or flight response, then the limbic system. No references, although alone this is excusable since these ideas are more in public knowledge nowadays.
  • Next you get into perception and say essentially that your perception of control for a situation dictates the amount of fear you experience. Do you have any proof or references for this? Or is this your own idea?
  • You then go into the whole limbic circuit for how epinephrine and norepinephrine is released (why didnt you say adrenaline instead?). Now I'm not going to suggest that you find a reference for every detail, that's crazy. In fact this is all kind of mechanical stuff. Not theoretical. But things like Idiopathic Ventricular Fibrillation, those numbers should have a reference. You even show an image of what looks like a study in the background, what is that? I assume its related, and maybe where you got those numbers. Just reference it dude.

How to DOWNLOAD A BRAIN! | The SCIENCE!...of Nier: Automata

May 4 2017 | ~ 1 million views

  • Need reference for 2.5 petabytes
  • Good when talking about salk institute research, and I suppose that you just looked at this first link here.
  • All throughout this video you mix in rudimentary science, which you don't need citations for, with some further concepts or ideas that you do. And maybe you just found some webpage somewhere that gave you your fact. That's hard to cite, I know. But... if thats the case, dont you think you need better research standards? :P
  • 11:35 You talk about maximizing the efficiency of the analog signal, even show some graphs. Need citation, where did this come from?! You leave no indication of where on earth you found those graphs. They are only attributed to 'really smart people'. You DEFINITELY can't just show data or figures without having a citation
  • 13:50 University College of London sending that large transfer speed, that needs a citation. You show a lot of figures and even text from the paper. What's the paper?
  • "Orthogonal Frequency Division Multiplexing" Also missing a reference

Game Theory

Game Theory: Beyond Fidget Spinners – How to Create a YouTube Trend

June 15 2017 | ~ 2.6 million views

  • This is actually fine, you make it clear the entire time where the information is coming from. You even show the article. Perfect. It would be nice to go the extra mile and have that citation in the description so others can easily grab the article, but thats ok.

Game Theory: Super Mario Maker, BIGGER than the UNIVERSE!

August 6 2017 | ~ 3.4 million views

  • "Feel free to cite me in all future uses of this very specific number." Ah yes, like how you cite everyone else? ;)

Game Theory: FNAF, We were WRONG about the Bite (Five Nights at Freddy's)

April 29 2017 | ~ 8.2 million views

  • Here a lot of the video is based on information from magazine article, which he displays on screen clearly, and names the author. This is good.
  • He also mentions 'research' when talking about skull changes as you age, I wonder where he found that.
  • When talking about research by Tobias Matee (sp?), he names and credits appropriately. Good! Again though, if he got this from a research paper he should cite the paper here, not just the guy directly.
  • 10:00 'I looked it up', ok where?
  • 10:25 He found a study! Ah excellent! But here they are just referred to as 'scientific madmen'. There's indication or way to trace back where this research is from. No names, no article names, no years, etc. This is bad. BadPat.

Game Theory: FNAF, The Answer was RIGHT IN FRONT OF US (Five Nights at Freddys Sister Location)

July 25 2017 | ~ 5.8 million views

  • For this, the FNAF book he refers to constantly is credited correctly, absolutely. He shows the cover on screen, it's always obvious when information he is pulling from is from the book. He even leaves page numbers! So if you have the book, it's easy to go and confirm for yourself, as well as gather more context. Beautiful!
  • He references the 'jumbled word' thing that has been shared around a lot. Considering that this effect, and this exact text in particular, kind of took an online life of its own, its ok he didnt go digging for the actual research behind this.
  • He mentions almost offhandedly that this is how optical illusions work, and there are plenty of research on this. It couldn't hurt to have a reference for this rather than stating it as fact.
  • Unrelated note, but the jump from 'optical illusions, your brain fills in gaps!' to 'a sound can cause specific illusions of monster animatronics' is a HUGE stretch.

Game Theory: Can Chicken Nuggets SAVE YOUR LIFE?! | Kindergarten

July 1 2017 | ~ 4.3 million views

  • Obviously most of this video is either core physics (which doesnt need references) to their own experiment. This is a seperate point, but.. I would have really liked to see this data for myself. I mean release of this data is absolutely up to them, but just saying. They did the work, it'd be cool to make that available for others (if, say, I wanted to check their math ;) )

Game Theory: Gaster's Identity REVEALED! (Undertale)

Feb 26 2017 | ~ 3.5 million views

  • Ah yes, neuroscience. Let's see what you got.
  • Two hemisphere theory, ok you name a researcher. You describe the corpus callosum experiment. So close, just needed to name the article... and you know, reference properly.
  • Also, you then describe the subsequent analysis of these split brain patients, could use references too. Honestly just find a more recent review article that covers all of this and you're golden.
  • A side note, MatPat is very good at first presenting a clear hypothesis and then going step by step through some examples he's found to illustrate why that makes sense, which people certainly do. You want to prove your point. But it'd be a lot more interesting and robust (and less misleading) of him if he at least presented some counter arguments. Just saying. Honestly I think it'd be more convincing

Game Theory: Nintendo’s SECRET PLAN for the SWITCH!

March 2 2017 | ~ 5.7 million views

  • I got to this first! I just wanted to point this out, I made a similar video almost a week before :) https://youtu.be/dL20QKBXa5w
    • And I think i did it better ;) Well, you had better presentation, and actually went a different angle with AR, but I gave better evidence

Game Theory: PROVING Mass Effect's Indoctrination Theory! | Mass Effect 3

April 14 2017 | ~ 2 million views

  • Mentions all these studies about infrasound (below 20Hz) frequencies negatively affecting humans. Does not display or specifically refer or reference any of them. No examples... citation needed.
Kyle

Kyle

I do tactile research—It's a touchy subject. Psych/Neuro grad student and creator of LittleGadget, a channel about Science and Videogames.

Read More