317 karma · joined November 11, 2016
Technical Director @ Team Liquid. Live in Berlin, sometimes around Utrecht! I'm interested in statistics in competitive video gaming, mostly Dota 2; as well as mathematics, film, and trivia. I run datdota.com and windrun.io
My feeling is that it'll struggle with word-plays in OnlyConnect/Connections (like missing letters, added letters, words-within-words homophones, etc) as well as two-step references (such as {Venice, Dream, Night, Nothing} => "last words of Shakespeare plays"}).
* a "tail" -- some initial part of the path before you get into a true cycle (like the 002, 004 in the example above)
* an inner repeating cycle before you reach a node you've seen before.
In the case given |t| = 0 and |c| = 1, but it's easy to construct a more complex example with nodes (A, B, C, D), edges (A->B, B->C, C->D, D->B), and 'ending nodes' being B and D. In this case left and right paths go to the same node. This case would have a tail of length 1, and then the inner cycles would be of length {2, 1, 2, 1 ...}.
As a result, valid 'ending states' (Z-nodes) for this graph would be after {1, 3, 4, 6, 7, 9, 10, ...} steps.
The intended solution was seemingly to calculate the length of the cycle (as in, until you were on a "Z" node) for each starting node separately. Once you'd done this, you could find the LCM of these cycle lengths to find the overall period of the cycle (~10^13).
This solution might not work if your cycles weren't constant length -- for example imagine when walking your graph you found your i_th Z-node after {6, 7, 6, 7, 6, 7, 6, 7 ...} steps; or perhaps {6, 7, 7, 7, ...}. In the test data, this didn't occur - it was always {n, n, n, n, ...}.
I can give another example perhaps - consider you're asked to find the last 3 digits of 2^n for n >= 1. You start off calculating: 002, 004, 008, ... eventually you get to 2^103 which ends in 008. This means that the cycle length would be {103, 101, 101, 101, 101, ...} since it'll never get back to 002 or 004. Solving this is a bit more difficult than the constant cycle lengths, since it's not a simple LCM.
That's a very nifty strawman, but that's not my claim (at least not in such a general way). In this case it's become SOP by experts in a field, specifically for safety reasons. Other experts in the field have said "well, we don't think it's necessary - there are downsides such as aircraft landing rate". These groups of experts don't disagree on the fact that it's more safe - they just disagree as to what the optimal combination of (safety, financial value) is appropriate.
Apparently some Canadian carriers do also have this as a SOP.
Around the time this happened I spoke to some friends who are ATCs (in the US) who all immediately agreed it was a very reasonable request, especially given that the request was made far enough out (so it wasn't like they'd have to quickly scramble to respace the incoming planes correctly in the sequence).
I'm actually surprised SFO still allows visual approaches at night after that Air Canada 759 flight nearly landed on the taxiway ~5 years ago (https://www.youtube.com/watch?v=aGQlQFn0euI).
One thing I was experimenting with (read: struggling with!) was a way to keep per-service (per-repo) architecture workspaces which are also synchronized on-commit to a central workspace and used !include to bind them together. The moving parts are not difficult - but writing your DSLs in a way that can handle this can be. This idea would let individual projects be self-sufficient and generate their own README doc diagrams as part of their own build process; but also have a central site which shows all the services as well as inter-service connectivity.
Did you ever consider https://github.com/avisi-cloud/structurizr-site-generatr to bring together your ADRs in with your architecture, or do you keep them separate?
Almost a year before CSGO official launch there was a showmatch for it and it looked awful. A bunch of small things were improved by the launch, but it was not good at all (look back at 2012 CS:GO footage and you'll be shocked). Sound was inaccurate and inconsistent, molotovs and smokes were glitchy, wallbangs didn't work in a sane manner ... it was pretty atrocious. That said, 1.6 was basically ancient and Valve (who didn't develop the game) pushed for CS:GO to be used in competitions, so it was used. Over time it was increasingly improved and there was an influx of existing (1.6 & CS:S) pro players towards the end of 2012 and early 2013; and an explosion towards the end of 2013. CS:GO came out in August 2012, but only crossed over the 100k concurrent player mark (which is the all-time CS:Source record) in ~ December 2013.
Skin betting became rampant on CS:GO after the August 2013 Arms Deal update, with sites like CSGOLounge having absolutely huge overall parimutuel pots on basically ever game. This pushed up the popularity of the game - even a small local tournament could have 10k viewers if the game was listed on CSGOLounge [and the bigger matches had millions on the line].
Eventually, years later, CS:GO was improved to the point where it could stand alone without propping up by the skin gambling - and since then it's grown much further; but in reality both CSGO and CS:S had very similar and buggy first few years.
* The pairwise comparison: does it freeze updates and calculate all shifts at the same time? For example if you had A > B > C do you calculate the impact of {A>B, A>C, B>C}, sum these impacts together (grouped by the VC), and then apply them? Or do you do it iteratively: if a firm had {A}, then {A>B}, then {A>B>C} do you add 0 then 1 then 2 comparisons as you get new data?
* How do you handle the fact that respondents to the survey are over a large time-frame, so some VCs might get better or worse over that time frame? Is there some Elo-decay applied?