--- Log opened Wed Oct 17 00:00:07 2012 20121017 00:30:02< Alarantalara> skyfaller: I probably went too far the other way, but bowmen should now never be recruited over spearmen (fish should hopefully still be recruited if there is a fast route by sea) 20121017 01:27:49-!- irker679 [~irker@ai0867.net] has quit [Quit: transmission timeout] 20121017 01:44:13-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20121017 01:54:54-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-umc-dev 20121017 02:19:13< skyfaller> yeah, that's a little too far, but it's a reasonable move in most circumstances. Better for the AI to recruit no bowmen than to recruit lots 20121017 02:20:20< Alarantalara> Well the next step will be to get it to recruit some ranged units if they are lacking 20121017 02:20:24< Alarantalara> which should bring them back 20121017 03:24:28-!- un214 [~un214@108.221.231.191] has joined #wesnoth-umc-dev 20121017 03:29:59-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-umc-dev 20121017 03:47:57-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-umc-dev 20121017 04:00:34-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host] 20121017 04:00:34-!- vultraz [~chatzilla@unaffiliated/vultraz] has joined #wesnoth-umc-dev 20121017 04:15:55-!- un214 [~un214@108.221.231.191] has quit [Remote host closed the connection] 20121017 04:33:38-!- loonybot [~loonybot@wesnoth/bot/loonybot] has quit [Remote host closed the connection] 20121017 04:47:26< skyfaller> Alarantalara: the main problem is that *one* bowman can be good, especially when paired with a mage 20121017 04:47:46< skyfaller> and when facing pierce-vulnerable melee units with low dodge 20121017 04:47:59< skyfaller> such as cavalry, clashers, and grunts 20121017 04:48:33< skyfaller> if you are seeing a lot of grunts and goblins, for instance, a bowman may be necessary to offset the cheap unit spam 20121017 04:48:49< skyfaller> a mage might be able to kill a goblin every turn and still get overwhelmed, for instance 20121017 04:49:26< skyfaller> and if the mage misses, then you have no ranged damage that turn 20121017 04:49:44< skyfaller> but with a bowman paired up with them, a mage+bowman team can clean up cheap units 20121017 04:49:48< Alarantalara> That's why I want to add something that tries to get ranged damage next 20121017 04:50:28< skyfaller> the thing is that ranged damage is very tricky to keep safe, the default AI won't properly shield/protect its glass cannons 20121017 04:51:14< Alarantalara> I also have this fear that there is an era somewhere where rnaged damage it normal and melee is the special one 20121017 04:51:20< skyfaller> the default AI would throw away goblins rather then trying to keep them alive long enough to level 20121017 04:51:56< skyfaller> Alarantalara: even if there were a ranged specialist race, and you could argue that describes Elves, melee units are still your tanks in most cases 20121017 04:52:17< skyfaller> melee units still usually deal more damage and have more HP for the price 20121017 04:52:18< Alarantalara> I was thinking of an entire era, where all the factions were ranged specialized 20121017 04:52:30< Alarantalara> I'd still want the AI to play properly 20121017 04:52:35< skyfaller> oh, I see 20121017 04:52:49< skyfaller> that is tricky 20121017 04:53:05< skyfaller> but you do come across situations like that in real games 20121017 04:53:19< Alarantalara> I'm thinking that I might need to try and make sure no range is unrepresented 20121017 04:53:30< skyfaller> I think I mentioned the ladder game where my Undead was beaten by a Loyalist player who had a Red Mage leader and spammed mages 20121017 04:53:58< Alarantalara> Indeed 20121017 04:54:03< skyfaller> or maybe he was a Rebel, I forgot 20121017 04:54:34< skyfaller> but the point is I didn't handle it correctly, I kept on fielding Ghouls and such, but he had enough firepower to kill them before they could get a hit in 20121017 04:55:32< skyfaller> DA spam was the correct move in that game, roughly equivalent damage for a lower price, and he'd get killed by retal 20121017 04:55:57< skyfaller> that's not a typical matchup, but our AI should be prepared for strange/unusual situations like that 20121017 04:56:00< Alarantalara> I was thinking of a function like 1/x^2, where the first unit of that range is really valuable, and later ones are much less so 20121017 04:56:26< skyfaller> I mean, that's how I like to play, I like variety, and having different tools ready to handle different situations 20121017 04:56:34< skyfaller> but often spamming one unit can win the game 20121017 04:56:48< Alarantalara> That way it should have at least one of each range, but then spamming whatever seems best is still valid 20121017 04:56:50< skyfaller> if it is the right unit and your enemy hasn't figured out an effective counter 20121017 04:57:24< skyfaller> sometimes if you wait to spam until you have one of everything, it is too late for that spam strategy to work 20121017 04:57:43< skyfaller> like, an early DA rush would be very difficult for drakes to deal with, for instance 20121017 04:57:56< Alarantalara> I'm hoping having 1 ranged unit is acceptable towards that 20121017 04:58:04< Alarantalara> (or 1 melee in that case) 20121017 04:59:59< Alarantalara> I am somewhat worried about countering spamming, since often the best counter is vulnerable to another unit 20121017 05:00:41< Alarantalara> For instance, if you had switched entirely to dark adepts, the opponent can easily switch to cavalrymen 20121017 05:01:19< Alarantalara> At which point you could be worse off than before 20121017 05:03:04< Alarantalara> But if you kept recruiting some ghouls (or other units) to go with the DAs, then your counter isn't as effective 20121017 05:03:51< Alarantalara> Right now I'm biased in favour of balanced forces, where the first few units of a type are the easiest to get 20121017 05:04:19< Alarantalara> but if opponent spamming is detected, then it's not to hard to have lots of a unit 20121017 05:04:50< Alarantalara> That second part is leading me into spamming WCs as undead right now - the point to start spamming them is a bit too easy 20121017 05:05:45< Alarantalara> Play about 3 turns vs Ron and it becomes very obvious 20121017 05:08:10< Alarantalara> The other problem is that if the opponent gets enough of jump with the spamming, Ron can end up doing the same thing as you did in that game vs all the mages: keep recruiting units with lots of health trying to keep some of them alive long enough to do some damage 20121017 05:09:42-!- un214 [~un214@108.221.231.191] has joined #wesnoth-umc-dev 20121017 05:12:06< Alarantalara> Also, is this new? http://imagebin.org/232245 20121017 05:12:29< Alarantalara> I don't remember being able to see where my opponent wanted to move when interrupted 20121017 05:15:26< Alarantalara> Anyway, it is time for me to sleep, so goodnight 20121017 05:15:27-!- Alarantalara [~Adium@173.33.158.188] has quit [Quit: Leaving.] 20121017 05:53:24-!- un214 [~un214@108.221.231.191] has quit [Remote host closed the connection] 20121017 06:35:48-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Quit: shadowm_laptop] 20121017 08:46:20-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-umc-dev 20121017 12:03:44-!- loonybot [~loonybot@wesnoth/bot/loonybot] has joined #wesnoth-umc-dev 20121017 12:03:47-!- mode/#wesnoth-umc-dev [+v loonybot] by ChanServ 20121017 12:04:23-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has quit [Remote host closed the connection] 20121017 12:05:02-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has joined #wesnoth-umc-dev 20121017 12:09:16-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has quit [Ping timeout: 246 seconds] 20121017 12:36:35-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy] 20121017 16:38:33-!- vultraz_laptop [~chatzilla@124.109.10.167] has joined #wesnoth-umc-dev 20121017 16:38:52-!- vultraz_laptop [~chatzilla@124.109.10.167] has quit [Changing host] 20121017 16:38:52-!- vultraz_laptop [~chatzilla@unaffiliated/vultraz] has joined #wesnoth-umc-dev 20121017 18:24:49-!- skyfaller [~skyfaller@ool-43551e75.dyn.optonline.net] has joined #wesnoth-umc-dev 20121017 18:24:50-!- skyfaller [~skyfaller@ool-43551e75.dyn.optonline.net] has quit [Changing host] 20121017 18:24:50-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has joined #wesnoth-umc-dev 20121017 18:27:20-!- mattsc [~mattsc@DHCP-211-241.caltech.edu] has joined #wesnoth-umc-dev 20121017 18:28:14-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-umc-dev 20121017 19:53:52-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy] 20121017 20:14:09-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-umc-dev 20121017 20:16:10-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-umc-dev --- Log opened Wed Oct 17 20:56:50 2012 20121017 20:56:59-!- lobby [~wesnoth@wesnoth/bot/lobby] has joined #wesnoth-umc-dev 20121017 20:56:59-!- Topic for #wesnoth-umc-dev: Wesnoth-UMC-Dev Project discussion and support channel | http://wesnoth-umc-dev.ai0867.net | Staff: AI0867, Espreon | Blog: http://wesnoth-umc-dev.ai0867.net/blog/ | IRC logs: http://irclogs.wesnoth.org, http://irclogs.ai0867.net 20121017 20:56:59-!- Topic set by shadowm [] [Mon Aug 27 11:27:23 2012] 20121017 20:56:59[Users #wesnoth-umc-dev] 20121017 20:56:59[@ChanServ ] [ esr ] [ loonycyborg] [ shadowm ] [ wesbot ] 20121017 20:56:59[+loonybot ] [ fendrin] [ Lukc ] [ shadowm_laptop] [ zookeeper] 20121017 20:56:59[+shikadibot] [ iwaim_ ] [ mattsc ] [ skyfaller ] 20121017 20:56:59[ AI0867 ] [ janebot] [ melinath ] [ vultraz ] 20121017 20:56:59[ Espreon ] [ lobby ] [ noy ] [ vultraz_laptop] 20121017 20:56:59-!- Irssi: #wesnoth-umc-dev: Total of 22 nicks [1 ops, 0 halfops, 2 voices, 19 normal] 20121017 20:56:59-!- Home page for #wesnoth-umc-dev: http://wesnoth-umc-dev.ai0867.net 20121017 20:57:18-!- Channel #wesnoth-umc-dev created Mon Mar 31 16:51:24 2008 20121017 20:58:18-!- Irssi: Join to #wesnoth-umc-dev was synced in 87 secs 20121017 22:58:29-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 252 seconds] --- Log closed Thu Oct 18 00:01:00 2012