Timeline for Survival Game - Create Your Wolf
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 15, 2017 at 17:02 | history | edited | Luke | CC BY-SA 3.0 | Fixed link |
| Apr 13, 2014 at 10:24 | comment | added | Moogie | @Runer112 Ah, i see :) I have worked out a way to legitimately (i.e. no reflection hacks, and only using the information provided by the simulation) determine who are friendly wolves with a high probability. They now share a map of where friend wolves are. My problem its now how to use this information to the wolves advantage. | |
| Apr 12, 2014 at 2:47 | comment | added | AJMansfield | Looking at your math for computing the best move, I think you could probably do it much better if you used confidence intervals, rather than just deciding between the move with the 'most wins so far' and the 'least uses so far'. | |
| Apr 11, 2014 at 13:36 | comment | added | Rainbolt | @Runer112 It looks like your scoreboard had 15 less submissions from your scoreboard. I run 1000 iterations per trial. I wipe the entire project and replace with a fresh copy after each trial. I'm having problems compiling the one written in Scala, but I haven't given up on it yet. I'll run more trials next time, promise! | |
| Apr 11, 2014 at 11:36 | comment | added | Runer112 | @Rusher Oh, whoops. I think I was still running with an old version of the game code from the sandbox, and either the MAP_SIZE variable didn't actually exist back then or for some reason I ignored it and added in my own static version of it. I also wonder why my wolf scored consistently slightly better than all other wolves in my trials, but did worse in yours... Different wolf set, I guess? Or do you run your games for a number of iterations other than 1000? Or you might just need a larger sample size, 5 isn't statistically that great. | |
| Apr 11, 2014 at 3:33 | comment | added | Rainbolt | Didn't compile because MAP_SIZE is a non-static variable. Since I didn't say it would be an instance variable that, I fixed it for you. Just wanted you to know in case I posted the complete project and you saw your butchered variable names. | |
| Apr 11, 2014 at 1:58 | comment | added | Runer112 | @Moogie No deterministic assessment is made about whether a wolf is friendly. However, we know the number of species of wolves, we know that each species starts at 100, we know how many other friendly wolves are left, and we can guess how many unfriendly wolves are left based on the number we've killed and a guess as to how many have died to other animals. From those, we approximate the probability that any wolf we see is friendly. This information doesn't have a huge effect, but it might be the difference between choosing whether to potentially battle a wolf or to definitively battle a lion. | |
| Apr 10, 2014 at 22:30 | comment | added | Moogie | @Runer112 impressive! I am looking at the code and I am a bit lost :P but one concept that intrigues me is the what you are using to base the decision that wolves are friendly or not | |
| Apr 10, 2014 at 19:24 | comment | added | Justin | I don't know if dropboxusercontent will lose the file, so I posted it on ideone: ideone.com/uRNxvj | |
| Apr 10, 2014 at 16:21 | history | answered | Runer112 | CC BY-SA 3.0 |