Which is likely to be every 24 to 48 hours now that there are more players. Possibly longer once we go over 16 players (next restart of the tournament)
I'm trialling running 3 tournaments in parallel, so I can get a headstart on the next leaderboard after the one I'm currently running (the currently running one doesn't have Vampire in it)
@ΗγρεŗN̛ευτŗιͷo Did you notice any difference to the startup delay? I did a little optimisation but not sure how much difference it makes on a different computer
There's also a fork of the controller with experimental extra features for tracking ants and debugging, if you're interested in trying it out.
Just going to link the git repo fork that has an additional option to enable tracking ants on the map for those that want to see where the heck the Steamroller is. Download zip, extract, open the html page in your browser, the js does the rest (props to the guys that created the baseline "parse and run answers" code). :) — Draco18sJul 23 at 6:10
@ΗγρεŗN̛ευτŗιͷo Yes, the score at the end of a game is the number of queens who have less food than you (any food held by workers at the end of the game doesn't count)
@ΗγρεŗN̛ευτŗιͷσ Although I'd recommend reading the spec first, if you write something that tries to break a rule there will be a disqualification reason shown along with the input the function received and its response, so you can see what went wrong
It looks the same as this one, but with extras. Works well as long as you don't use "Fit to width", which only works properly on the original controller
Yes I was expecting Vampire to perform as well as Steamroller since it's based on it, but it's not so far. Takes a while for things to average out though, and possibly even longer now there are more than 16 players
@DestructibleLemon The spec only bans teaming up against others - each player has to be trying to win in its own right. Taking advantage of your own previous player is acceptable if the previous player is still trying to win, rather than just setting up something purely for the later player
@DestructibleLemon Yes it's been slowing down gradually as more players arrive that produce a population of workers. The problem with Pierce is that it often converts most of its food to workers, so that half the population of ants are Pierce workers. Other times it stays with just 2 workers, so maybe the average doubling of the game time is partly down to Ziggurat and the improvements to Black Hole.
None of these are a problem for the game - it's all acceptable strategy. I only mean "problem" in the sense that the leaderboard will have to wait a day or so longer.
@DestructibleLemon That can only be subjective. I've tried to avoid strong incentive for that. For example, I changed the design before posting so that workers steal from queens instead of vice versa. This means creating an extra player purely to supply extra workers is not as effective anymore. There is still opportunity for building a purely supporting player though, so I'll just have to judge this as it happens
Also this way around it's harder to build a supporting player without having other players take advantage of it too
That might well happen, especially once Vampire is updated to target more different players
I notice that Vampire still allows Ziggurat to create new workers on schedule, so the worker population grows and brings more and more food back to the Vampire queen
@Draco18s Upvoted so it will show without people having to press "show more comments". Is it worth deleting the earlier comments (a couple mention downloading) to avoid confusion?
I'm going to run a tournament overnight with every existent bot against my new black hole, minus Trail Eraser (I want to see how good the black hole is against everything that isn't specifically built to murderize it)
know what's awesome about pause-on-disqualification? I can copy the input over to the "test input" box and click "test" and figure out what went wrong quickly :D
@DestructibleLemon That's a handy coincidence. The palette in the paper has an extra colour that I didn't need - if I'd dropped another one instead it wouldn't have lined up so nicely
Draco18s - Last Monday at 11:06 PM How do I merge the commits on the original into my fork?
Keybounce - Yesterday at 12:48 AM What I do: 1. My work is in a new branch, something like 'git checkout -b master new' 2. To update: a- git checkout master b- git fetch c- git pull d- git checkout new e- git merge master
I changed my readme file online, and made all the other changes locally. Git just told me to pull from GitHub before pushing my local changes. I didn't have to do anything else
Great. Started a new tournament with the latest version, while leaving the existing tournament running. Hopefully this will allow posting a leaderboard based on the old version tomorrow or Friday, and then updating to a leaderboard based on the new one without having to wait another 48 hours.
Trail-eraser
var i, j
var orthogonals = [1, 3, 7, 5] // These are the non-diagonal cells
if(view[4].ant.type == 5) {
//Queen moves straight to get food
// Color own cell if white
if (view[4].color != 6) {
return {cell:4, color:6}
}
var specified = null;
// Otherwise move to a white cell op...
So I was watching a game with all the players, and paying particular attention to explorer.
And I discovered that Trail Eraser got interested in it during Stage 1.
Followed it all the way up till Stage 3, where it had deployed it's workers.
Then, trail eraser just say there, and kept keeping the paths it was trying to deploy around it, so efficently, that it accidently reset Explorer to Stage 1, which let explorer escape.
There's some bot trying to follow one of Explorer's dead worker's paths..?
And then because it doesn't use one path for the workers, Explorer rockets up the score board.
@Draco18s Last night's tournament gave "A laden worker cannot move onto food." Input: [{"color":1,"food":1,"ant":null},{"color":2,"food":0,"ant":null},{"color":8,"food":0,"ant":null},{"color":8,"food":0,"ant":null},{"color":8,"food":0,"ant":{"food":1,"type":1,"friend":true}},{"color":8,"food":0,"ant":null},{"color":8,"food":0,"ant":null},{"color":2,"food":0,"ant":null},{"color":8,"food":0,"ant":null}] Response: {"cell":0}
I haven't excluded from the latest tournament since you've already made an edit since this earlier tournament started, so this is just to let you know in case it's relevant.
Interested to see how Vampire is affected by it. The random moves to avoid loops makes me wonder if deliberately lacing it with loops would form Vampire traps
Trail-eraser
var i, j
var orthogonals = [1, 3, 7, 5] // These are the non-diagonal cells
if(view[4].ant.type == 5) {
//Queen moves straight to get food
// Color own cell if white
if (view[4].color != 6) {
return {cell:4, color:6}
}
var specified = null;
// Otherwise move to a white cell op...
Formic Functions
Formic Functions