Timeline for What Doesn't Kill Me…
Current License: CC BY-SA 3.0
78 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 13, 2017 at 12:39 | history | edited | CommunityBot |
replaced http://codegolf.stackexchange.com/ with https://codegolf.stackexchange.com/
|
|
| May 28, 2015 at 0:58 | comment | added | Ralph Marshall | Now that I have both first and third place I move that we close the contest ;-) But seriously, I'm looking forward to the next round of results, and agree this has been a very well done KotH | |
| May 27, 2015 at 8:50 | comment | added | Katenkyo | @Geobits When do you think you'll have some time to run an other contest? ^^ Some entry actually changed, and some new ones are here ! Btw, Good job for such a nice KotH, there's a lot of people around here ! | |
| May 26, 2015 at 18:01 | comment | added | Shaangor | Come back from vacation and I'm in 6th! I enjoy the simplicity of the rules; they have surprising strategic depth. | |
| May 26, 2015 at 9:54 | answer | added | Matt Y | timeline score: 1 | |
| May 23, 2015 at 5:20 | answer | added | colorado777 | timeline score: 4 | |
| May 23, 2015 at 0:14 | answer | added | Herjan | timeline score: 2 | |
| May 22, 2015 at 20:45 | comment | added | Xynariz | Oooh, I won a round! I would have expected my Guardian to do better than my Tank, but I'll take what I can get. | |
| May 22, 2015 at 18:38 | answer | added | euanjt | timeline score: 2 | |
| May 22, 2015 at 17:22 | comment | added | Others | @Geobits Ok, thats fine. | |
| May 22, 2015 at 16:52 | comment | added | Geobits | @Others I'd prefer it, just to make running everything easier. If your language forces separate files, that's fine as long as it's a single command to run, and input/output work properly. | |
| May 22, 2015 at 16:14 | comment | added | Others | @Geobits Must the bots be one file? | |
| May 22, 2015 at 2:22 | history | edited | Geobits | CC BY-SA 3.0 |
new results!
|
| May 21, 2015 at 17:25 | comment | added | sirpercival | @Geobits no worries, sorry about your week :( | |
| May 21, 2015 at 14:09 | comment | added | Ralph Marshall | Great. Thanks. Looking forward to the new update. | |
| May 21, 2015 at 13:42 | comment | added | Geobits | @RalphMarshall Depends on how long it takes with all the bots, but the "official" results will be as posted, 15 rounds/games. If I do post results from a larger one, it's for info purposes only <insert standard disclaimer here>. | |
| May 21, 2015 at 13:41 | comment | added | Geobits | @sirpercival Yea, it's been a rough week. I'll try my best to get that done tonight. | |
| May 21, 2015 at 13:40 | comment | added | Ralph Marshall | If you do post a new one I'd like to see results of a larger number of runs. I did a partial version with 150 rounds and that lets you see some results for each of the bots that have reasonable strategies. Great competition! | |
| May 21, 2015 at 12:45 | comment | added | sirpercival | @Geobits are you gonna post a new leaderboard? | |
| May 21, 2015 at 12:17 | answer | added | Katenkyo | timeline score: 5 | |
| May 20, 2015 at 6:48 | comment | added | plannapus | Ok then I am the one who misunderstood. I thought the bouts were run in parallel, not the bots (having written that, no wonder i m confused :) ). | |
| May 20, 2015 at 6:36 | answer | added | rudi | timeline score: 9 | |
| May 19, 2015 at 14:57 | comment | added | OJFord | @plannapus Unless I've misunderstood, that's not going to happen. What's being parallelised is the running of entered bots, per bout. i.e. Analyst & Tank run simultaneously in turn 1 - but neither of them has a turn 2 until both have completed. | |
| May 19, 2015 at 6:40 | comment | added | plannapus | @Geobits My bot bases its action on what happened in the very bout that he is currently playing. So if two instances of the bot write in parallel to the same file while playing two different bouts it screws up the logic of the bot. | |
| S May 19, 2015 at 6:31 | history | suggested | Others | CC BY-SA 3.0 |
Added parallelized controller at Geobit's request
|
| May 19, 2015 at 4:04 | review | Suggested edits | |||
| S May 19, 2015 at 6:31 | |||||
| May 18, 2015 at 19:38 | answer | added | agtoever | timeline score: 3 | |
| May 18, 2015 at 14:35 | comment | added | Geobits | @Manu As long as each bot isn't being executed simultaneously, it shouldn't matter. I assume the parallel version still has all bots finish one turn before moving to the next. Like I said, I'm going to verify before using it if posted. | |
| May 18, 2015 at 14:23 | comment | added | Manuel Allenspach | I don't think that the parallelized version will work, since some bots rely on their files (which may be a problem). | |
| May 18, 2015 at 12:48 | comment | added | Geobits | @Potatomato Hmm. I'd say it's okay as long as you're only using the sample bots. | |
| May 18, 2015 at 12:48 | comment | added | Geobits | @Others You can post it as an addition to the main question if you'd like. If I can easily determine the behavior is correct, I'll use that for my next run (hopefully tonight). | |
| May 18, 2015 at 11:21 | answer | added | euanjt | timeline score: 7 | |
| May 18, 2015 at 4:16 | comment | added | Potatomato | May I train my genetic programming based bots using the sample bots as a base for my bots? | |
| May 17, 2015 at 23:58 | comment | added | Others | I have a parallelized version of the controller that runs games faster than the original controller does, if anyone is interested I can post it... | |
| May 17, 2015 at 18:04 | comment | added | OJFord | If anyone's curious how the bots stack up now that there's more submissions, here's the result of my "unofficial" test run: 3-809-Hero; 2-593-Velociraptor; 1-471-SurpriseBot; 1-433-Coward; 1-371-Santayana; 1-364-Wiisniper; 1-262-Analyst; 1-230-Bully; 1-132-Equaliser; 1-71-IKnowYou; 0-0 -- precog, Berserker, BroBot, SemiRandom, MonteBot, Tactician, SnappingTurtle, Psycho, Revenger, Opportunity, PhantomMenace, Tank, Boxer, Guardian, AboveAverage, DefensiveBot. Lots of invalid commands from Opportunity, AboveAvg, BroBot, Boxer, and SemiRandom. | |
| May 17, 2015 at 18:00 | comment | added | OJFord | @Sebb No. Rule 2: "You may not target or otherwise single out a bot by meta-analysis. Use only the information your bot is given. This includes your own bots, so you may not enter two bots that collude." | |
| May 17, 2015 at 16:06 | comment | added | Sebb | Have been waiting for a challenge like this :) But one thing bothers me: May I use the names of the enemies to predict their behavior? | |
| May 17, 2015 at 0:57 | answer | added | Sergey Vlasov | timeline score: 6 | |
| May 16, 2015 at 20:33 | answer | added | Stormcrow | timeline score: 9 | |
| May 16, 2015 at 18:32 | answer | added | randomra | timeline score: 7 | |
| May 16, 2015 at 5:41 | comment | added | Geobits |
@TNT Stronger is in the default package, so it does make it easier if the bots are too. Just put all the class files in a single folder (along with bots from other languages, etc) and run java Stronger -log from that folder in a terminal (removing -log still works, but it just doesn't give output for a while). If using an IDE, you'll need to make sure the class paths and working directory settings are correct, but that varies from IDE to IDE.
|
|
| May 16, 2015 at 3:35 | comment | added | TNT | I'm confused on how to run the Stronger.java file. Do the other class files (Hero.class, for instance) have to be in the same package? Must Stronger and Hero run at the same time? Or am I not supposed to run it? | |
| May 15, 2015 at 23:41 | answer | added | Others | timeline score: 21 | |
| May 15, 2015 at 22:35 | answer | added | Shaangor | timeline score: 7 | |
| May 15, 2015 at 20:59 | answer | added | TheNumberOne | timeline score: 6 | |
| May 15, 2015 at 19:09 | answer | added | Ralph Marshall | timeline score: 16 | |
| May 15, 2015 at 17:23 | answer | added | TNT | timeline score: 7 | |
| May 15, 2015 at 12:58 | answer | added | Manuel Allenspach | timeline score: 8 | |
| May 15, 2015 at 12:49 | comment | added | Geobits | @agtoever Yea, that shuffles them between each game for just that purpose. | |
| May 15, 2015 at 11:48 | answer | added | sirpercival | timeline score: 9 | |
| May 15, 2015 at 10:48 | answer | added | LowIntHighCha | timeline score: 12 | |
| May 15, 2015 at 10:21 | comment | added | agtoever |
Oh. Wait. Now I see the Collections.shuffle(players) in Stronger.java. I suppose that is shuffling the players...
|
|
| May 15, 2015 at 10:19 | history | tweeted | twitter.com/#!/StackCodeGolf/status/599157158622691328 | ||
| May 15, 2015 at 10:11 | comment | added | agtoever |
Nice challenge. I like it that there is no moving around. While working on a bot, I noticed that the order of the bots make a great difference in the outcome. That's because Hero and Bully are using life >= best and life <= best. Robots that are entered later in the contest (higher ID) have a disadvantage. Do you randomize the order the robots are entered into the competition?
|
|
| May 15, 2015 at 10:00 | answer | added | Logic Knight | timeline score: 11 | |
| May 15, 2015 at 9:35 | answer | added | Potatomato | timeline score: 26 | |
| May 15, 2015 at 9:33 | answer | added | plannapus | timeline score: 13 | |
| May 15, 2015 at 7:21 | answer | added | Multihunter | timeline score: 6 | |
| May 15, 2015 at 5:02 | comment | added | Elias Benevedes | @sirpercival look at what I did, I posted a completely broken bot :P | |
| May 15, 2015 at 4:28 | comment | added | Geobits | @Mark The current bots are about half defensively oriented and half attack. None of the defensive bots have a single win. I agree that it encourages some defense, but the benefit of aggression is that pure defense cannot win any points at all. You only win points if you are the sole survivor. | |
| May 15, 2015 at 4:07 | comment | added | Mark | Your scoring seems to rather strongly encourage defensive play -- there's no benefit to aggression, only to survival. I've seen this sort of thing before, with RoboWar's Alliance of Pacifist Scum. | |
| May 15, 2015 at 2:33 | answer | added | BrainSteel | timeline score: 9 | |
| May 15, 2015 at 2:28 | comment | added | Geobits | @sirpercival You can do either really, but if you don't want to chance the down votes, it's better to post a fully working bot. At least "working" in the sense that it can be compiled and run, even if the logic isn't airtight yet. | |
| May 15, 2015 at 2:26 | history | edited | Geobits | CC BY-SA 3.0 |
added disappointing initial results
|
| May 14, 2015 at 23:41 | answer | added | Xynariz | timeline score: 15 | |
| May 14, 2015 at 23:13 | answer | added | Xynariz | timeline score: 8 | |
| May 14, 2015 at 21:48 | comment | added | sirpercival | i have a bot but it's not working yet. should i post it, or wait til tomorrow when i have more time to test? | |
| May 14, 2015 at 20:10 | comment | added | Geobits | @SuperScript Not on a set schedule. Will do an initial scoreboard tonight, and every one/two days after as long as new entries are coming in regularly. | |
| May 14, 2015 at 20:08 | comment | added | Robin | @Geobits : how often will you run the Tournament and update the leaderboard? | |
| May 14, 2015 at 19:47 | answer | added | OJFord | timeline score: 19 | |
| May 14, 2015 at 19:36 | answer | added | OJFord | timeline score: 23 | |
| May 14, 2015 at 18:32 | comment | added | Geobits | @Tim The bot list is space separated. Each bot's stats are comma separated. | |
| May 14, 2015 at 18:30 | comment | added | Tim | You mean comma separated, not space seperated. | |
| May 14, 2015 at 17:17 | answer | added | Elias Benevedes | timeline score: 8 | |
| May 14, 2015 at 16:48 | answer | added | Ypnypn | timeline score: 13 | |
| May 14, 2015 at 16:15 | history | edited | Geobits | CC BY-SA 3.0 |
ellipsis?
|
| S May 14, 2015 at 16:05 | answer | added | Geobits | timeline score: 18 | |
| S May 14, 2015 at 16:05 | history | asked | Geobits | CC BY-SA 3.0 |