I prefer quality over quantity. And as an extra benefit, taking your time means you may be able to do a few more passes and get a more "fair" ranking. You have our blessing! PS.: And if it helps you sleep better at night, do a a quick rough video where you show the top 5. Whatever floats your boat!
4 years ago
| 196
I agree with what a lot of people have said here already. Why not both? Make a quick and dirty one to give out the prize money and then later give us a fancy pulsating(?) graph and show off the top five strategies. I'd definitely watch both!
4 years ago | 25
Hey Cary, thanks for putting in all this work! This tournament 100% deserves a quality video with cool stats and contestant interviews! On the other hand we'd all love to learn about the results. Why don't you make an announcement stream somewhat soon (after maybe 10 runs of the tournament) and an extended "documentary" styled video later? Both on their own would be interesting, but both together is even better :)
4 years ago | 41
Although I didn’t participate, I would say that quality over quantity is usually better but I would probably watch either because this seems interesting.
4 years ago
| 124
I feel like a lot more people would like to see a polished version, but the contestants would prefer a simpler one that they could watch earlier(I can only assume). Personally, I prefer a well thought out video with exciting visuals. Please don’t overwork yourself and take your time, I will watch anything at this point so its all up to you. :]
4 years ago (edited)
| 37
Thanks for asking us! As a contestant, I'd rather have the rough results now. However, as a viewer, I'd much prefer the longer video. My suggestion is the same as loads here: why not both? A quick video to say who actually won, and a detailed one later down the line would be amazing!
4 years ago | 7
Thanks for consulting us! I would opt for a short, rough, narration video going over the top 10 strategies (maybe for one of your alternate channels), followed by a polished video on the main CaryKH channel!
4 years ago (edited) | 10
Cary, you are the best I've seen at data visualization. That is basically the heart of what you do! Abacaba, Twow, Scale of the Universe, and even the old machine learning videos have this element in a well thought out way. Most of them highlights data and asks questions in surprising ways I never even considered. All of this is to say: I'd watch a rambly stream of consciousness OBS style video, but I'd admire a well planned plunge into the weirdest of nit-picking of the data produced. So no pressure or anything. You do you.
4 years ago | 1
Why not a simple video and later then complex video? I don't think people would mind watching both as they both have value
4 years ago | 70
I mean it seems to me the logical thing to do would be to inform participants of their placings via whatever their preferred communication method was asap (along with distribution of prizes), and then make the nicer video for the proper announcement. That way nobody’s stuck waiting for months on their results, but you also have the best end product possible. Besides, knowing placing would also lead to more interesting conversations as far as the calling people who did interesting things aspect is concerned.
4 years ago | 2
Some people suggested it already but I wanna show my support, I think making a short video announcing the winners in lazykh then posting about it on the community tab on Carykh is a good way for you to gain time while doing the more complicated video. While I did participate and I do want to see who won, I also wanna see all the statistics and stories from the most interesting ones and all the stuff you mentioned. Might be because I'm a nerd though, dunno
4 years ago | 0
You could maby do both, a quick video where you announce the winners and notable mentions and a video that goes deeper into everything that happened
4 years ago | 3
People will always answer “quality over quantity” regardless the proposition because it looks negative to choose otherwise. But I really don’t think the alternative of a 2 month-video would necessarily have lower quality just because it didn’t take the more polished road. The problem itself is interesting and the process checking the best strategy is interesting as well. Yeah, of course adding animation and cool graphics is neat and having a more fine tuned data is better. But I don’t think the viewers know how to choose it in an unbiased way. We don’t know the energy and time required to accomplish it. “Quality over quantity” is always a free choice for viewers, even if it is not free for creators. I am pro the creator’s choice. Not only do they know the limits and constrains of doing one way vs another, they are also able to put effort when doing both ways. It is more like quality at different concentrations. So my suggestion, Cary, is to close your eyes and see inside you what is calling you the most. If it is the long journey, we will wait and have fun then! If it is the short journey we will have fun sooner and appreciate the project together. Don’t do it quickly if you are excited about having an elaborate video on your portfolio, but also don’t do more than you want to or have energy for. Also; if you end up deciding the long video, consider a more efficient way to do it because it seems it is worth it. Maybe allowing multithreaded processes, or running it on a more powerful machine / cluster
4 years ago (edited) | 7
As someone who participated, I would love to see a more in depth video, the data analysis is probably the most interesting aspect of this project regardless of who won.
4 years ago | 0
I did participate, and I think that I'd be fine waiting on the results so long as the video is good, and I wouldn't mind a deeper dive into the interesting algorithms.
4 years ago | 2
If you’re certain that the winner isn’t going to change, then you don’t need to run many more tests, but don’t rush the final video.
4 years ago | 3
You should make extra parts of the celebrity videos. That would be nice. Part 1: Customizing celebrities (already done) Part 2: Creating babies and toying around with celebrities (already done) Part 3: Revealing all 300 components quickly (not done yet) Part 4: Toying around with horizontal shading (not done yet) Part 5: Making three parent babies (not done yet)
3 years ago | 2
Pulsating plot is a must, if we can’t make that happen might as well trash the whole project
4 years ago | 1
carykh
Prisoner's Dilemma Tournament update: So, I have results that I could share!
In the end, 1,468 of the total 1,615 strategies passed all four requirements: 1) was a .py file, 2) ran against the default 9 strats in under 10 seconds, 3) didn't do anything nefarious, and 4) never crashed when paired up against any other strategy. Including the original 9 strats means the entire working roster has 1,477 contenders.
However, due to a 1477*1477 matrix having over 2 million entries (1 million bc of symmetry), and some of those entries taking >1 second to run (most are faster), each single "full-pass" takes 8-10 hours to run. I've done 2 "full-passes" so far, and the top strategies on the leaderboard move around n spots, where n is maybe 20% of their rank (so if they're 30th, they might end up 24th the next round.) #2 moves, but #1 does not. This is also because hundreds of strategies scored very close together. Yes, I'm aware there are many ways to optimize, including multi-threading, caching, and more. However, since I'm already confident the top spot won't ever move, I'm wondering how necessary it is to run hundreds of full-passes.
My main dilemma (haha) is this:
If I tried to get the results out as fast as possible, I could get the video and the prize money out in a day or two! It would have to be an unedited, OBS-livestream style discussion. On the other hand, if I were to make a more polished video with more polished data, it might take months. (I envisioned displaying the data in a 3D environment with fancy transitions and animations to make the dataset feel like a new continent being explored. Also, I planned on calling every participant who did anything interesting, to include that in the video, but that would be a big logistical task!) I know many people would side with "quality over quantity" (meaning wait for a polished video), but I do think quick-n-easy argument is valid. It would suck to enter a tournament and not know the results for months, just so Cary could make his pretty scatter plot to pulsate... (like, does anybody even care if a graph pulsates, versus being a still image?) So, what do you guys think?
Also, if you're wondering why I posted this as a YouTube community post instead of a Github issue, my last GitHub issue only got 2 comments. So, in the interest of this post reaching as many participant's eyes as possible, I'm posting it here! (If I ever do future tournaments, I'll agree on a singular announcement location beforehand lol)
4 years ago | [YT] | 1,035