alre

exactly. if randomness required equal chances, we couldn't say, for instance, that arrows land at random, because you don't have equal chances of different outcomes. looking further into the subject, there is a common concept of things being more or less random based on how predictable they are. techinically, this is uncorrect because randomness is a on/off property, and there are better, more mathematically sound concepts for expressing diversity in expected outcomes. but of course this is where technical jargon and common speech really diverge. I hope @Gurken Khan will forgive me for the OT, I promise I'm over it.

still needs the queries. potentially optimizable by keeping a table of all palisade segments and neighbours, updated by build ruitines, but that would need further work. actual performance fall has to be tested to be known.

those are technical words with precise meanings, and the following are facts: - random does not require even chances - RNG are algorithms that were always used behind random picking civs. I am an expert in the field.

potentially quite heavy because it could need a range query for every hit dealt to the palisade, maybe two (one for each end of the palisade segment).

that's still technically random, and of course it always relied on RNG.

well but turrets and towers are two different things. lazy solutions are better than no solutions anyway, and lazy fixes are still fixes.

I wasn't suggesting to just add arrows to walls, but to make garrisoned walls functionally similar, if not equivalent to garrisoned towers and forts. how? make the guys on walls not shootable at, and raise their attack time. Just a suggestion, but one that works for sure (similar to what we already have) while your solution may as well not.

I am a supporter of palisades having a bonus against cav, as it would differentiate it more from stone walls. I think another another buff for walls, would be to change the effect on garrisoned troops. if it was equal or similar to that of soldiers garrisoning towers, they would be more viable and interesting IMO.

there is not one right time. but when you start to train, you train a good lot then you attack. any time you spend without fighting is time you are falling behind in eco because skirtay gather slowly.

why not both?

at current revision, they do have different stats from other pikemen, but they still have the pikemen class. you do, changing the role of formations in the gameplay would require a lot of work and the effects are not knowable until then. that's no alternative to just changing some stats now, it's not even comparable.

what's that, exactly?

do han halberdiers need to be on pikeman class? ji are not pikes, and they have different stats anyway.

hopefully, balance can be carried on with official rebalance mod that will be released with a26.

lol agree. probably better to scale the number of slots with size, simply in order to avoid people cramming around the resource and being blocked out.

you mean https://trac.wildfiregames.com/ticket/2516#comment:31? yeah it looks like the game has to be modified, because most games wouldn't be rated even if all players resigned correctly, and maybe even if they stayed in the game until the end, it's not super clear from that comment. in any case, the problem is not in lobby-bots.

I think there is a general consensus that something should be done for that. Also something about ratings that are not up to date. However, I don't think anyone can tell now what's the best course of action. If players are presented with a solution that looks more satisfactory (in terms of computed ratings) than the previous one, they will accept it. also I guess that more than one rating system could be proposed at the same time, for trial periods. I'm proficient in python and in statistics, so I can mantain the lobby-bots part, if needed. why should that be different than in 1v1? one can make unrated TGs anyway.

definitely.

this thread is about lobby moderation.

still, it's a lot more efficient, you said it yourself:

I heard that there are a lot of "space weapons" in history. meteoritic iron was fascinating then just as much as it is now.

I wouldn't just say "slightly": a moderator would be able to spend virtually no time at all reading lobby conversations, but just the strict minimum, that is a lot more efficient. I was asking because I don't know the xmpp at all, and I've been a moderator on a software that had all this. I wanted to know what's there and what's not.

agree on -1 by default. mostly, hosts prefer to have that setting because it's less hassle. how is it better to kick slow observers rather than having slow observers lag behind? they can leave if they stop enjoing the lagged stream, instead of getting kicked directly. join lag is a different thing entirely.

what about warfare? there must be archaeology based reconstruction!

since smirfing has explicitly been left out of the thread topic by @user1, and because fighting smurfing requires a different skillset than simple moderation anyway, I'm going to ignore it now. moderating in chats is usually a pretty straightforward activity. in most online community the way this works is: users report violations, the moderators log in once in a while, see the reports, and decide how to act. moderating is particularly comfortable from smartphones, as mods can log in in any moment of spare time and quickly see if there are any reports, and swiftly act on those, greatly reducing the wait time for users. on some communities/apps more mods can see the same reports and act independently, so to provide independent confirmation. I don't know xmpp really, but i've made some research and i didn't find any of this about xmpp. If you provided some practical indications on how this whole moderation thing is going to work, I think more people would be encouraged to step forward. it's all very unclear to me now, and I don't want to have to study an internet protocol by myself to make this work.