what would happen if I sorted before each iteration, and had adjacent players play?

I feel like the winning percentage would approach 50%

assuming accurate estimations

That might work better, because it could skew your overall winning percentage.

also, this KoTH has a huge benefit: I can pick an arbitrary number of players, and have multiple bots of the same type play

it's the ideal for KoTHs, but I always seem to get fewer of these than I'd like

The score should probably be the sum of absolute errors of all your guess throughout the game or something. Otherwise you could just guess random stuff to mislead your opponents until the end.

well, that's what I was debating

I'll get some sleep now. ;)

because guessing random stuff could be a viable strategy

because maybe you just want to get a feel for how well you do across all players

but you need to know when to stop

@NathanMerrill This might also be a channel you could add to the feed:

Another problem related to my previous graph-display question: how might one go about dissecting a circle into n pieces that are roughly the same size and as round as possible? Basically, I want to be able to color a dot with some number of colors, but if I just do wedges, it's entirely plausible that they get too thin to accurately note the colors. If, on the other hand, the dot is divided into hexes, I would expect the colors to be more distinguishable.

simulate n repelling particles and then colour the voronoi cells? :)

@El'endiaStarman is there an upper bound on n?

@flawr Not really, but realistically, at some point, you're not gonna be able to show 100 different colors on a dot.

These colors could come from a tagging system. They could come from identification of maximal cliques including that vertex. Or both. Or some other source.

@flawr Hmm, perhaps. Could maybe simulate these particles in 3D and then project to 2D (Riemann sphere?) and then color in voronoi cells.

@El'endiaStarman why go to 3d first?

Just as a way to repel from the edge of the disc. Or maybe I could do that more simply by inverting the repulsion with respect to the origin?

In any case, this would be easy to precompute and serve up when needed.

@flawr I'm liking what I'm seeing so far. THUNK - 122. The Stochastic Model of Scientific Discovery

Another method: First use slices up to e.g. n=8, then distribute them into an inner disc and an outer ring up to n=8+16

then add a third ring

Hmm, perhaps. And the sizes of the rings could be 1/r^2 to account for the increasing area.

@flawr I wasn't motivated to watch it (based off of the image), but it turned out really interesting

@flawr There might be a better way to divvy up the slices among rings. For instance, with n=4, you could do four quarters or two rings with both split in half. For n=8, quarters on the inside and outside rings. Or maybe three rings, with the centermost being split in half and the other two split into thirds.

This could perhaps be a code challenge on PPCG. Objective criterion is sum of (sum of squared distances between each pixel in a region and the center of gravity of that region) over all regions.

Or ratio of perimeter to area.

@flawr Yeah, perhaps. Did you draw those yourself?

yep, and I used an expensive image editing tool, as you can see.

Nothing but the very best.

I mean your work is only ever as good as your tools.

@flawr doesn't that mean that tools are impossible to create?

@NathanMerrill shhhhh

The slide rule must've been hella awesome if they designed the SR-71 Blackbird with it.

"The hammer I'm working on here will only be as good as this stone I'm using to make it"

@El'endiaStarman That is why it leaked so much!

But you have to admit, it would almost pass as an ancient proverb.

@NathanMerrill That guy in Primitive Technology turns this adage on its head in every video.

I've often wondered how we get precision in the modern world

@El'endiaStarman actually one could scale those rings according to the number of divisions.

like, if the tools we use are only X% precise, how could I ever create a tool that was more precise than that

@NathanMerrill Or how long it would take us if all man made things would disappear, and we'd have to start from scratch, but with the current knowledge people have in their heads.

you don't allow for libraries?

then we're pretty much toast

you're right, things wouldn't work out...

even farmers are adjusted to the tools they use, and aren't proficient in making those tools

your backyard welder might be good at making tools, but he doesn't have the equipment he needs

the currently least developed areas would probably fare best

@flawr A Christian fiction trilogy I read a few years ago (Chiveis Trilogy) actually had that as the premise. Nuclear war and a deadly virus knocked the modern world back to the Stone Age, but people still remembered stuff. Hence, medieval times in 400 years instead of 1600.

it depends on why they aren't developed. It's entirely possible that the region is harsh and difficult to use

and therefore, they wouldn't be well off

but if the region was less-developed but in a good location, then you have the best chance

this totally reminds me of the way of kings

they've had repeated desolations which knock society back down to square 1

and so you have all of this ancient lore and artifacts that nobody understands

@NathanMerrill +1 recommend:)

@El'endiaStarman sounds fun, would you recommend reading it?

@flawr I personally enjoyed it. I'd say it depends more on whether you like religion and spirituality and such. It is, after all, Christian fiction and focuses on that more than the more mundane details of life in post-apocalypse medieval times.

@flawr you read book 3 yet?

@NathanMerrill currently reading oathbringer

@El'endiaStarman I'll add it to my tbr list then!

@flawr Alright. :)

@Feeds is this the right URL? somehow it doesn't appear as the other ones in the list.

Yeah it is. It always takes a bit for the title to update

if you click on the link, you'll see it goes to an RSS feed

@NathanMerrill hm strange=/

@NathanMerrill did you also read any of the other cosmere books?

Mistborn (all 6), elantris, warbreaker, started rithmatist

I've also read steelheart and firefight, though I don't remember the latter very much

I've got some catching up to do then :)

I haven't read the cosmere-specific book, nor do I plan on finishing rithmatist

and I'm on an eternally-long waiting list for oathbringer

but other than those, I think I'm caught up

(because steelheart isn't cosmere)

...perhaps the URL is wrong?

because it should have posted 5 videos by now

but I have no idea why its wrong because the link works

Spoke a bit too soon, huh @NathanMerrill? :P

@NathanMerrill i guess now it works xD

lol

https://www.youtube.com/feeds/videos.xml?channel_id=UCMOqf8ab-42UUQIdVoKwjlQ (Practical Engineering)
youtube.com/feeds/… (THUNK)

The URLs don't look much different either.

I hope you don't mind this:

Rebel scum.

not quite 10 attoparsec

"Attoparsec" is a funny unit of measurement. How close is it to 10?

@flawr The title is "Happy Feet", not "Happy Foot".

one parsec is about 3.086 × 10^16m

@El'endiaStarman how long is a feet then? XD

@flawr I'll tell you if you tell me how long a meters is. ;)

hehe

TIL about StackRating - stackrating.com/list/byRating?page=1&userId=1473772

@El'endiaStarman there's a suspiciously large range of answers given

like, on that page, there's another person with 1 answer

and has the same ELO. Not saying that the calculating isn't right, I just don't think it's actually accurate

@NathanMerrill lausek? That user is deleted/suspended.

Actually, looks like the reputation number is totally wrong. Probably hasn't been updated in a while.

yeah. I imagine it's expensive to process every answer on the site

> Live Monitoring // The Stack Overflow site is monitored continuously through the StackExchange API (thanks Sanjiv for the Java API!). Currently it scans all questions from the past 90 days at least once every 24 hours.

@El'endiaStarman I actually got an email from a recruiter because of my ranking there at some point

It has Jon Skeet's rep as 824,659 but it's over a million now.

@MartinEnder Oh wow, that's cool!

Speaking of jobs, I just accepted a job at Microsoft

congrats! :)

is that the C# job?

yeah.

sweet!

do you know/can you talk about what you're gonna be doing?

Congratulations! When do you think you'll start?

I'll start in a couple weeks (I don't know the exact date yet)

I don't remember the name of the product, but its basically Microsoft's solution to companies wanting to manage lots of windows machines (whether they be desktops or tablets)

also, I conveniently don't have to move. They have an office around 20 minutes away from where I live

Wow, sounds great.

Say, weren't you asking a little while ago about whether to write a code demo in Java or C#? Which did you go with and what was their reaction?

Java. They completely didn't care

@NathanMerrill congrats:)

I'm glad I went with Java too: My GUI was far easier to create, and most of the feedback they had was related to scalability

it also expanded my mind to what scalability really meant.

I've always considered it in terms of Big-O notation

but they were concerned about "What happens if I put this on a server, how easy would it be to aggregate events from 30 other machines" or "What if I give you a different type of event" and etc

@NathanMerrill Did you ever have any kind of formal training in some programming language?

yeah, I've gone to university in CS

well I know someone who has a degree in cs without having been formally introduced to any particular language:)

As a more concrete example, I was storing all of the data, and would filter it using some function. However, I realized in the interview that this wasn't scalable: having 50 million records in memory isn't scalable: I should have used a database.

@flawr he hasn't learned a single language? I mostly used C++, Java, and Python in school

like, I get that CS can get theoretical, but if you haven't learned a language, that's a bit too theoretical for me

@NathanMerrill he did learn various languages, but never as a course at uni or so

but I don't know how he managed to do that, as far as I know at least today every CS curriculum includes some language already in the first semester

well, I didn't have a "X language" course. The beginning courses were in C++, and I had a GUI course in Java (which was horrid), and a web development course that covered PHP/CSS/Javascript, a Python/Perl course, and a course that covered language types (Haskell, Prolog, a few others)

in Germany the stance of many CS degrees is "we expect you to learn the programming languages you need on your own" but end up offering one or two mandatory courses on programming and then a bunch of optional ones on software engineering

that's not the case here. Especially the web development course, the teacher was very focused on ensuring we understood the language semantics in great detail

We had an entire 2 days on HTML.

@El'endiaStarman see also: how hard it is to launch things into the sun.