@m.buettner In an extremely simple case, let's say that we each have one base and there are two unoccupied bases. Thanks to RNG, one of the free bases happens to be 100 light years from me, and the other is 150 light years from you. I obviously get to mine first. For the remainder of the game, I will have more units than you do. Now blow this up for 100 players and 1000 planets. One of them gets an early lead. What stops them?
If you can come up with a good strategy involving 4 bases, I can generate an even better one with 5.
If you make a good strategy with 5 bases, I'll have a better one with 6.
@Rusher it's a matter of time scales. Travelling will take between a few seconds and maybe half a year. Bases are supposed to generate new units every week or month or so. Unless you decide to fly from one of Neptune's satellites to one of Uranus's satellites, while they are on opposite sides of the solar system, the difference in travel time will be negligible compared to the unit spawning rate.
That is, most nearby flights will take less than a day.
a) we're not talking about a two player game... there will be other players keeping you busy on other sides... it's going to be really hard to amass a great number of units anywhere. b) say you're right, is that even a problem? do I want a player with a single base to still be able to overcome an empire controlling 3 large planet systems? I think that might just mean I messed up earlier, and I can't win at this point.
in fact you suggested earlier to not include infinite unit spawning, because with a limited number of units the battle is bound to come to an end. now you're arguing that with infinite unit spawning a player may get the edge, such that the game can't be turned around any more. (not trying to put words in your mouth, this is how I remember/understood it)
Sure, but the end of the game might not be domination of the solar system, but everyone running out of units and gas. I don't think that's much different from putting a hard time limit on the challenge.
That the goal of eliminating all players may not be achieved in either case. And if people decide to just sit around and not do anything, I have to put a time limit on it anyway, because I can't rely on them using up their resources.
Okay but back to actual issue. Is it really an issue? Isn't it just like this in risk? Once you control a sufficiently large portion of the planet, you simply march through your enemies armies, and they can't do much about it any more. There's also some kind of critical mass.
@m.buettner When humans play Risk and they see another player getting close to winning, they all team up on that player. But they do it in a specific way. They say things like "If you knock him out, I won't attack your vulnerable area for three more turns."
That doesn't change anything about the possibility of one player getting ahead anyway... and when it's his turn, there's nothing from stopping him winning the game.
@m.buettner If the winner is decided in 50 weeks, then the latter 950 weeks are not important. You have no end game. If the winner is constantly ganged up on and replaced by a new winner, then the end game is key to victory.
@Rusher no, but you seem to be complaining about the fact that there will be a point after which you can't win any more if one player has a certain amount of bases.
Then the question is, how late or early in the game is it okay for such a situation to occur?
Allow me to help make things clear. From here on out, don't say things like "You seem to be implying that if this were kinda like that then maybe something."
You can do your challenge however you want to. Just don't question the validity of an entire discussion just because you don't agree with what I'm saying.
a) gang up on the leader requires communication between bots of different players, which I think would so ridiculously complicate the game, it won't be worth it.
@Rusher no that's not why I questioned the validity of the discussion... I questioned it, because I literally wasn't sure what we were actually discussion. Because your statement itself is true for any game, and only makes sense if you think it's quantitatively weird for this design, but I couldn't really figure it out if you thought that was the case or not.
now the same is true for the solar system game, the only question is "where is that point?" and if it is too early (which I don't know but which might well be the case), then we can talk about ways to push it back
but similarly, I don't think a small lead will give you a certain win.
"Owning half the universe and holding on to it for some number of turns is not the same as the situation I was describing earlier." Just so we don't talk cross-purposes again, which situation are you referring to exactly?
I think if the lead is early or late is irrelevant (for now). The question is how large that lead has to be. then we can talk about how early such a lead can be acquired. at least I can't anticipate that intuitively.
Do you think a small lead would guarantee you to win?
A day or two ago, the following question was posted:
http://codegolf.stackexchange.com/questions/28620/a-program-to-transliterate-l33t-words-to-regular-english-words
It was about translating leetspeak into regular English. However, since there was no objective winning criterion, the question wa...
@Rusher okay, so you need a certain advantage. In particular, going by your earlier argument, you need to be able to beat n other players (and here n is probably greater than 5 and hopefully greater than 10, and potentially greater than 20). Or at least you'll need to be able to hold your position for long enough to be able to do that.
Do you think it'll be possible to get there in there in the early game?
(I don't even know how to measure "early" or "late" here... in moves? ... in game time?)
Okay, I personally don't think that a small advantage is enough to just sit it out, but as I said, I can't tell. (And I don't know how you can tell, but I'd like to know.)
So let's skip that part and say you're right, you can decide the game for yourself within a few moves if you're lucky and just wait.
You suggested to grant some form of buff or resource (any incentive) for attacking other players.
That's why I originally said that taking over a base immediately grants you n fighters. But Geobits made a valid point, that this could lead to absolute chaos of two enemies repeatedly take away their bases.
Okay, but why does that not happen in Risk? Only because players cooperate? Would you say risk suffers from the same problem if you prohibited players talking to each other?
Also, I think there is fair chance that no single bot will manage to use their units optimally, so I could just be better at using my units more efficiently than you to make up for the difference. I think you're maths assumes that everyone is playing by a known dominant strategy.
Another thing, let's assume players have full information. They know you've got a base more than they do. I think that would increase the chances of them attacking your bases, too.
It doesn't happen in Risk because players cooperate. If you prevent players from talking to each other in Risk, it might suffer the same problem. If played by bots, it would definitely suffer the same problem.
If you are given that information, you have two options as a bot. 1) Take a stab at the leader or 2) Hope someone else does it for you
Conflict with the leader seems detrimental to my health
okay, you might do that. but that's not the point. if you're saying your argument only works, because people won't be following the strategy that is worst for you, that doesn't seem to be a valid argument to me.
oha, well, I might be asleep by then, but feel free to respond, and I'll get back to it tomorrow
So your argument works if people don't attack the leader. If your argument doesn't work any more if people jump on the leader, that is incentive enough to attack the leader. Because it means no one can win if the leader is left alone. So the only winning move is to attack the leader. In which case you can assume that your opponents will do so. Can you still maintain your lead? (I don't know, I'm genuinely asking.)
Also attacking the leader doesn't seem to have any disadvantage, if the number of fighters per base is still equal, so the choice of base to attack is otherwise independent of the player who owns it.
I plan on adding another answer to this question, but i'm worried that it will be too similar to this answer. So i wanna post it here and see what you guys think
my (proposed) new answer is as follows:
#include <stdio.h>
const float inf = 1./0.;
int main() {
int x=1;
int sum=0xBDAAAAAB; // Arbitrary magic number for debugging
while(x --> inf) { // while x tends to infinity
sum+=x;
}
float sumf=*(float*)∑ // convert to float since -1/12 is not int
if(sumf == 0xBDAAAAAB) { // no sum performed, something's wrong with the loop...
fprintf(stderr, "sum is unchanged\n");
return -1;
}
printf("%f\n", sumf);
Assuming you took bases in a cluster and not scattered around, I can much faster reinforce the planets being attacked, and with more units than you can
I certainly wouldn't write a bot that favors attacking the leader. I would write a bot that attacks the bases with the least units defending them
I suppose I could throw a random if statement that says "If all bases in the 89 planet galaxy are being equally defended, and all of the enemy planets are equally distant from mine, then I'll just attack the leader."
The question isn't what bot would you write, the question is what bots could your opponents write. If three people decide to write a bot that always attacks the leader, it doesn't matter whether you think it's a sensible strategy or not. The question is, will the leader then still be able to keep his advantage?
If three people write a bot that always attacks the leader and I write a bot that always attacks them when they launch an attack on the leader, I'll thrive until I'm the leader, at which point I'll be under their attack.
As soon as I lose the lead, they'll focus on someone else and I'll regain the lead
But if you repeat this over and over, one person keeps coming out on top, and thats me
I don't see how attacking another player is any different
your argument also doesn't work because for the time you're not leader, you could reverse the argument for the player who is the leader during that time
I don't see how this conversation is meaningful once you factor in that you don't know who the leader is. At that point, all the talk of "attack the leader" vs "don't attack the leader" becomes... pointless.
@Geobits no I'm actually considering making that a thing. the "you can see only what's going on on your own planets" was born out of the "bots need to communicate" idea.
@m.buettner Then the 3 who are jumping on the leader will be at a strict disadvantage for engaging in combat against a stronger opponent, and the other 7 will gain. As soon as those three die (and they will die sooner than the rest because they search for trouble) then the situation returns to a greedy cascading win for one player.
The one you posted above? It is pretty similar. It's up to you, but I probably wouldn't bother if it were me. I might add it as an edit, showing both version in the original.
@m.buettner Wait, did I say I could capitalize on a small lead if three players were specifically attacking me? I don't recall saying that, and I'm really tired of having words put in my mouth.
@Rusher that's how I understood "If three people write a bot that always attacks the leader and I write a bot that always attacks them when they launch an attack on the leader, I'll thrive until I'm the leader, at which point I'll be under their attack. As soon as I lose the lead, they'll focus on someone else and I'll regain the lead But if you repeat this over and over, one person keeps coming out on top, and thats me"
A day or two ago, the following question was posted:
http://codegolf.stackexchange.com/questions/28620/a-program-to-transliterate-l33t-words-to-regular-english-words
It was about translating leetspeak into regular English. However, since there was no objective winning criterion, the question wa...
