Summer of Math Exposition

An absolutely masterpiece in inspiring pure mathematical collaboration! This video brilliantly succeeds in provoking the viewer’s intellect, driving a genuine desire to become part of the collective computing effort to solve this open, legendary 3-by-3 magic square of perfect squares mystery.

The algebraic presentation using extended parametrization through polynomials is exceptionally pristine and structurally comfortable to follow. Even without excessive graphical aids, the deep conceptual motivation and the call for community action make this entry the absolute gold standard of high-level exposition. A well-deserved 9!

Thanks for the video! I liked how you clearly motivated by recalling what a magic square is and showing some examples for it.

In terms of clarity, unfortunately I couldn’t follow everything: First, I was lost at the simple parameterization. A nice visualization would have helped here, e.g. where you show at least for one row and one column that indeed this pattern gives a magic square. Then later, I didn’t have your matrix in mind anymore, so I didn’t know where A,B,C etc. are located, which made it hard to follow. I felt issues like these throughout the rest of the video, e.g. where did the “three squares” in 6:28 suddenly come from? Why do we need the special ordering of the triple congruum you have shown? How do the seeds affect the search space? I think when starting a new section of your video, you should elaborate a bit more on new concepts before moving on (which does not necessarily mean to lengthen your videos, but rather give more weights on certain aspects and to clarify with more examples or more perspectives).

It was also not clear how exactly you searched the space of solutions. Maybe a quick visualization of how your for-loops in the linked C++ code operate would have helped here, especially when you say “I could not find any polynomial” without giving any details.

I think many clarity issues could be already solved by giving the visuals more time in production. E.g. aligning formulas at operators like ”=”, ”+” and using LaTeX or Typst to avoid typographic weirdness. Those systems already provide you correct spacing between variables and typeset symbols correctly, e.g. the minus sign should be longer, parentheses bigger for arguments like r(p/q) etc.

Avoid pauses in the middle of your own phrases (e.g. cut them in post). As a viewer, we of course need those pauses to think, but rather at end of phrases, not in the middle of them, when you had to think about how to end your phrase.

On a technical side, audio and video quality are quite nice. The video is 1080p with sharp image and audio is not noisy. For the visuals, just notice that some people have red/green blindness, so check your color palette. (Maybe add some more colors in general.)

So all in all, I think some more pedagogical thinking could go into your video to ease understanding and guiding viewers throughout, such that they can better appreciate your contribution to this problem.

Good luck. Have you considered working with the modulus of the numbers to verify the last digits are correct before performing the full check. I don’t know if it will work.

This isn’t necessarily a bad entry, but I have a few notes for future videos:

• At around 5:38, I would recommend having a mini 3 x 3 table to reference when comparing the sizes of the entries. Otherwise, it requires going back and forth through the video.

• Also, at 5:53, I would recommend actually showing the definition of a congruum.

• At around the 10 minute mark, with the transitions between formulas, there is a flash that is a bit disorienting.

• At 12:13, you have the full polynomial parameterization, but you don’t fully include a derivation or intuition from where it comes from.

• There’s also some weird spaces between words when you speak. I don’t necessarily think it’s terrible, but it is a bit distracting.

• I would also add sections in the video (by adding timestamps in the video description) to make the flow clearer.

However, I do think that the topic is interesting, and for that I think there are some points for novelty.

Don’t start your video “As you probably know, this is an unsolved problem” before you said a single word about what the problem is.

No ordinary viewer knows what “relatively prime” means. The numbers in your first example are not relatively prime, and that isn’t even a necessary condition in a magic square. Since this is the core concept of your video and you cannot even define it correctly, I have little hope for what comes next.

Interesting topic. I know math is strange and can surprise us with remarkable patterns. But I did not expect such a simple problem to have such a fascinating pattern with no solution. The fact that the diagonal sum is the only thing preventing us from finding a solution is quite fascinating.

However, I can’t rate this video too high. It was definitely above average, but I found the dry presentation style and the purity of the problem a little difficult to stay engaged with. I understand that abstract problems, like the Bernoulli compound problem, can lead to the discovery of something like e. But the fact that this problem didn’t have a satisfying conclusion to justify the pure math topic made it a little difficult for me to stay engaged with.

But overall, I did enjoy the video.

The video accomplishes its goal: to describe a problem and encourage others to pick up where the videos leaves off. For a casual viewer, I think too many of the calculations are shown and the overall message gets lost part of the way through. But that isn’t the intended audience, so that isn’t necessarily a bad thing. The narration was quite slow, which is better than being fast, but I needed to put the video on 1.25x speed and I almost never do that. Other than that, the visuals and audio were good, if a bit dry. The dryness of the video doesn’t do any favors towards encouraging others to take up this task, but I think it would help someone who is already interested in the problem.