I'm thinking about doing more videos, not only screencasts of rounds. For some time now I'm using OI Checklist by Rezwan.Arefin01 (cannot recommend enough, especially if you are preparing for hard IOI-style contests) as an archive, so I'm thinking about doing videos with editorials (more like my thinking process) for some old POI problems.
So, here are some questions for you:
1. Is this interesting?
2. Should I read (and think about) problems beforehand? Pros: I will have more structured thoughts about problem, I will know for sure if I will be able to solve this problem in a short time, ??? Cons: It is kinda unfair, it may look like I'm crushing it when in reality it could take hours for me to solve, ???
3. Should I do one-problem length videos? The problem is that I prefer to open some (3-6) problems, read all of them, and then think about each problem for some time, probably starting with the one I liked more.
4. Despite these are OI-style problems with partial scoring I'm trying to go for a full solution, most of the time ignoring groups. Sometimes full solution of these problems are really-really hard. Should I do groups always, or only when I'm stuck for long time, or other options?
5. Do you have any suggestions about what kinds of videos I should make? Or other things I didn't think about concerning this particular kind?
You can look at my channel here
UPD
Thanks for your input! I thought a bit myself and here is my current position:
I'm willing to invest some time to learn basic video editing (yeah, didn't do anything before, just recording screencasts), I hope that I will even add some things to screencasts in post. This way I can read some problems, then solve them one-by-one, but make separate videos for each problem. Mostly I will do "blind" attempts at problems, since most of your want to see me struggle :) I will go through groups if they are interesting, but rarely will implement partial solutions.
About lectures and lower-level educational videos: don't think so at least for now because it is not interesting for me.
Because of learning basic video editing I will be slower at first, but I hope to do first video in next few days, so stay tuned! Also later today I will upload USACO Platinum screencast, so you can watch this if you are interested.
"I will know for sure if I will be able to solve this problem in a short time". Lol if you won't solve just don't upload the video.
About partial scoring, I think often most subtasks are very boring and not interesting, so I think it is more enjoyable to try solving only for 100 pts.
He may want to do it by going live, in such a case, reading/not reading beforehand matters.
Looking forward to it, Keep it up.
Yes.
I think NOT. For me, it's much more valuable and interesting to understand how high qualified CP thinks.
In my opinion, ~3-4 problem length videos will be best...
Only when stuck for a long time.
Try making AOEII videos.
You are talking about a game?
Yeah, Age of Empires II. Not the 20-year-old version, there are patches, expansions, HD edition, a new (read: unbalanced) Definitive Edition and a fun community.
1) Yes, it's interesting. It's still rare to find any videos on that high level.
2)
Pros:
- As you said, more structured thoughts. Better video quality in general.
- You don't wait viewers' time when you just think.
- You won't follow an incorrect path and generally are less likely to make mistakes.
Cons:
- You need to spend more time because you first think and then repeat some stuff in a video.
- People like seeing thought process and mistakes too.
- Reading and solving problems as-you-go shows a more real contest, including some time pressure.
I think that you should sometimes do live-like contest to show people how you approach a problem from scratch, but the default should be knowing problems in advance. Also, it doesn't matter that "it may look like I'm crushing it" because people know anyway that you're good.
3) Whatever you prefer. If you solve multiple problems, make sure to display on screen which problem you're solving. It's then easier for people to find some part of the video.
4) Just do it similarly do a real contest. Solve subtasks only when you're stuck with the full thing. Maybe the exception should be a subtask that looks very interesting or educational.
5) Lectures on hard topics?
6) Additional suggestion. As IOI/OI-training season is starting, I'm planning on making some POI videos too so let's try not to repeat each other. I will be happy if you don't solve all hundreds of POI problems out there :D
Yay, pros here!!!
I have hard feelings about lectures, generally I think that lectures are overused as a teaching device. But maybe in a video format it could be fine...
Hello Mr weak mind!
it depends on the problem
Think it beforehand, when explaining describe your thought process.
Yes, if it is hard enough, do it for open cup, OI, WF problems
Just do the problems as if you are in contest (assumming you did'nt follow point 2).
About solving problems (there are many that don't have editorials), some none trivial tricks, etc.
I can't understand
Very interesting idea! I think if you solve them during the video it's more engaging and even if it takes a long time you can fast forward the parts and just show the parts where you have new ideas.
For someone at level like me, by watching some random div1 level screencast is still very difficult to understand and learn something (sorry we are too weak against div1 level problems). Instead, could you do some categorized and topic related videos, it would be very helpful for beginner or mid level person. For example, everytime pick one topic from this library and do some general intro and explainings and show 1-2 problem solvings. http://cp-algorithms.com/ If there is this kind of video library, I would pay money for it!
The question is for who are you really doing these videos? If it is for those 20% (specialists and higher) then, I guess, it is ok with live streaming and little details (though I don't really know), but for the rest of 80% of mostly grey competitors anything above div. 2 C/D is rocket science now. So if you really want to help the majority you don't even need to solve all the problems. Div. 2 A B C (sometimes D) with thorough details would be more than enough. Now I understand that this kind of great preparation of the videos will take much of your time which usually you combine as a training for yourself, but by solving very quickly A/B/C/D and then struggling with E and higher you just amaze weak competitors as I am, but eventually, it is kind of the same as just us going and looking for someone else's code. Greater help requires greater preparation, I guess. But what do I know, I am just a grey.
Counterpoint: there are far more training resources for those who are at a really low level. So many introductory courses and tutorials for simple things have been written that my head is spinning. But there is far less material for purples and blues already. Not everything needs to be aimed at you.
Also, more like 40% of active users are specialists and higher, not 20%. Many of these pupils and newbies also only have a few contests so they don't really count.
Regarding "blindly solve" vs "describing prethought solution", I feel the later suits better during lectures. Or when you try to fit maaaany problems in short time (say analyzing a particular icpc set of at least 10 problems). However, when you are trying a very hard problem, blind solve is much better choice. It shows how the thought process runs through your head. The problems you are going to attempt are tough. Most of the contestants spent lots of time before fully/partially solving them. Giving them just gist of solution is not helpful. What's helpful is, how a top contestants think to solve them faster/easier than themselves.
Auto comment: topic has been updated by Um_nik (previous revision, new revision, compare).
Maybe you will find here someone who will edit your videos for free or almost free. That being said, editing isn't that time consuming and I do it myself for my videos, while I outsource captions (to help people that don't understand my accent well).
Please do! POI solutions are super hard to find, especially those before 2011.
I think that opening 3-6 problems at a time is fine but please make sure they are connected (same year?) so that it is easier to navigate your videos.