[LURKERS] Sweet 16: Civ Party Fun Time and Philosophical Debate

[Gavagai]

If Gavagai thought that stealing 2 workers will with 100 % likelihood be accepted and not revenged that's obviously a mistake.

This isn't the case. I didn't think about their reaction at all because assumed that would be attacked anyway regardless of my actions.

[Dhalphir]

SO GAVAGAI WINNED ALOT BUT HE DIDNT HAD ANY PROBLEM?

Usualy i kill them, sooner or latter when is easiear , cheaper or more fun to do so.

quotes of the month

[Bacchus]

The above exchange Gavagai contra all made for an interesting read. I can't help but wonder whether the issue here is in some sense 'cultural'. To me, the whole peaceful/aggressive distinction also doesn't make much sense — there are times to be 'aggressive', that is take actions that directly damage the opponent's position, and there are time to be 'peaceful', that is gain through mutual advantage of some sort, say demilitarization. I see Seven's point in that some players are psychologically predisposed one way or another, that's certainly true and it's not of itself 'unprofessional'; after all even chess grandmasters have distinct playing styles, if only because their intuition proceeds faster along some paths than others. I think Gavagai slightly damages his cause by speaking in maxims, which he doesn't elucidate or contextualize too much — but nor is he really saying things that Seven ascribed to him, as in "acting completely at random" or "entirely ignoring what your opponents might do". Which again makes for a fascinating case of people talking past each other.

For my part, and going back to where it all started, the "being attacked is no indication of getting attacked again" is actually correct. As it's on Civstats and it's been a while, it's not really a spoiler, but still, PB13:

At some point WilliamLP attacked me and razed a city. By contrast, until recently Plako has never attacked. Should I have taken an indication that Will is 'aggressive' and will attack again, whilst plako is peaceful? Nothing could be further from the truth!

. Being attacked simply means that you were in a situation where the opponent saw the attack reasonable. The next time you are going to be attacked is far more likely not by that opponent, but by whoever you present with a similar opportunity, or whoever has a strategic case for attacking you. Treating attacks as arising not out of the objective situation in the game, but out of the personal qualities of the attacker is a really pernicious heuristic.

Having said that, a lot of people are emotionally invested in a game enough that an 'aggressive' action within it is seen as a real annoyance, far beyond what people would feel, say, about losing a pawn in chess and with this emotional investment comes the reticence to aggress against players, to avoid needlessly upsetting them. Of course, that differs, coming back to the cultural thing — if I was in GasparNobles situation, I would have known that the worker steal in itself tells me nothing about the viability of future cooperation, should the right situation present itself. I might still have decided that duelling with Gavagai is the only way to go, but not because 'Gavagai has shown his hostile intent'. Largely because I know that his 'hostile intents' arise in cases where he thinks being hostile is profitable, but put him in a situation where mutual help is profitable, and he will as soon take that up. I've had a situation where I took opportunistic advantage of an opponent-ally who we had a bunch of cooperative deals with, only to get response of 'Oh dear, that was silly of me. Shall we get back to it?' and we did, continuing fruitful cooperation for the rest of the game. Someone else might have jumped to the conclusion that I am 'aggressive', rather than 'peaceful' and that no co-existence is possible, but that stereotyping would have been counter-productive. Which again, brings up a point that Seven rather kept talking past — if you take your precautions against aggression, you don't even have to enter into the guessing game of who is feeling 'friendly' today, which is exactly the Machiavellian point of relying on what you can control, rather than on what you can't.

[SevenSpirits]

Bacchus, a simplified way to see the game is as a bunch of iterated prisoner's dilemmas between each pair of neighbors in the game, with the winner of the game being the one who gets the most points. I see the strategy you are advocating as being universal defection. ("if you take your precautions against aggression, you don't even have to enter into the guessing game of who is feeling 'friendly' today.") While this is the optimal strategy for winning any given encounter, and even for coming out ahead in your interactions with any given player, it is not optimal in the game as a whole. That's because even as you are "winning" against each of your neighbors, other players (including your neighbors) are benefitting from gentler relations with each other.

As a concrete civ example, in this game I did not defend my crappy coastal cities very well. While the important ones were internal and safe, most of them had 0-1 defenders at any given moment and could have been razed by a single galley with 2 units by any of a number of players. Why would I allow this? Isn't it safer to defend all those cities and have full control of the situation? I don't think so. I saved enough resources not having to defend them that I believe I could afford to replace a city and still come out ahead. And I think the expected number of cities that get attacked is less than one and closer to zero. Esentially I was repeatedly choosing cooperate with all the other players, with intention to start defecting if they did so, and being willing to suffer the loss when this occurred. I agree that treating every other player as a constant threat is easier, but I sure don't think it's best play.

[Bacchus]

I see the strategy you are advocating as being universal defection.

Not at all. See AlaePB1, which has a really bad case of coastal exposure, actually right between me and Gavagai, something we've lived with for over a hundred turns now. Taking coastal cities that you can't hold, or even the ones you can hold if they are economically irrelevant — what's the point? You are still making a 100 hammer commitment to that Galley and its marines, so why not just build a Settler? Garrisoning these coastals would not be protecting against reasonable attack, it would be protecting against idiocy. PD gets thrown around quite a lot, but in many Civ cases, the Defect/Cooperate pay-off is not actually higher than Cooperate/Cooperate, this coastal example included.

More generally, I'm not even advocating a strategy, there is no 'best' way to play iterated PD even when pay-offs are constant across rounds (with tit for tat coming out 'best' only in statistical terms and dependant on the field), and Civ doesn't even offer that constancy, it has highly variegated pay-offs in the many instances where interaction allows for a choice between 'defection' and 'cooperation'. Some defections can outright win you the game, there is no question about the need to protect against those, nor of taking them; some, as with the coastals described above, net you nothing in practical terms; the interesting ones lie anywhere between these two extremes, and one of the fun parts of the game is choosing when to make them. Sometimes, an early harass can ensure that 70 turns down the road you have a decisive advantage for profitable conquest; sometimes a very similar harass is nothing but an annoyance to your neighbour that isolates you diplomatically. And some defections vis-a-vis one player are cooperations with another.

What I am advocating is that when somebody does harass you or attack you in another way, 'oh they must be aggressive and there will only be war' is about the last conclusion you should draw. Not least because it puts a stop to profitable cooperation later. Now, you might say that in branding opponents that way you are openly declaring your commitment to the kill-switch strategy of cooperating until a defection and then defecting all the way, which is a Nash Equilibrium strategy, FWIW, but that's a personal preference and the 'aggressive' brand in this case is really just a brand.

[SevenSpirits]

Which again, brings up a point that Seven rather kept talking past — if you take your precautions against aggression, you don't even have to enter into the guessing game of who is feeling 'friendly' today, which is exactly the Machiavellian point of relying on what you can control, rather than on what you can't.

OK, I'll quote this. This is what I responded to.

[Serdoa]

I don't believe that a player attacking you automatically means that player is more likely to attack you later. Because that attack doesn't say anything about the situation later on, which will be more important than if he attacked earlier or not. I also don't think that a player that has not attacked you is more likely not to attack you in the future. The same applies, if the situation calls for it, he SHOULD attack you, no matter his actions in the past.

However, the same is not true for the attacked. He might very well go for you, just on the grounds to get revenge, even if the attack itself is rather likely to fail. See Gaspars attacks on Gavagai this game. Btw Gaspar, I think you refer to my comments on those when you say I stated you throw the game. That's true insofar as I think you throw your own game. Your attacks could have been made somewhere else with more gains imo. Or been better invested in settlers, workers and economy in general. It was actually just a sign of my frustration with your play because I knew you could do better if you'd put the revenge behind you (or at least that was my read of the situation).

Anyhow, I think the points made by Bacchus and Gavagai are interesting and good. Not necessarily something we all have to agree on of course, but discourse over Philosophy seldomly has that effect anyway. But even if one does not agree, it might give a good insight in how others look at the game (or the world really).

[Ituralde]

I really enjoyed reading this discussion about the different approaches you can take to the game. For me I always remember one quote from Seven Spirits where he said something along the lines, that the biggest drawback to attacking someone in any given situation is that you make an enemy. For me this rings true, which would probably put me into the peaceful category here.

But then there's always costs associated with any decision you make in these games and the points Bacchus and Gavagai bring up also sound true to, which is why I enyojed the discussion so much, I think I got some deeper insight into what drives my decision making process, which I want to thank all participants for!

[MindyMcCready]

We were discussing if there is a good argument against attacks of these kind, not whether I have any problems. So far, I don't see how argument "it will make them love you" can be convincing for reasons Machiavelly pointed out.

What I am pointing out is that you have complained a lot about these "unpredictable" actions by the team you attacked. I think that if you are ignoring the effects that your own actions have, you probably shouldn't complain about anything that other people do.

I find your absolute rules to be ridiculous. "You should pay no heed to how your actions influence your opponents." "It's not possible to fully understand your opponent's decisionmaking process, so you should ignore it entirely." "Subjugation is always superior to manipulation." "There is no reason not to seek out opportunities to attack people and take them." This is not a good way to make decisions. Good decisions are made by weighing outcomes.

The premises and logic are very lacking, too. I don't even know where to start on it. I guess I don't care. Enjoy your philosophy.

A classic post of a person who feels he is losing an argument and looking for comfortable way out.

Seven is really not losing that argument.

[Bacchus]

Which again, brings up a point that Seven rather kept talking past — if you take your precautions against aggression, you don't even have to enter into the guessing game of who is feeling 'friendly' today, which is exactly the Machiavellian point of relying on what you can control, rather than on what you can't.

OK, I'll quote this. This is what I responded to.

I can't see how you can read an advocacy for universal defection into the above even at a stretch. At most, you could read my phrase as asserting that one should be prepared for universal defection, because I missed out 'reasonable' before 'aggression'. My bad, it was implied, I did not mean to say that you should be prepared against all possible types of aggression all the time. There is of course a whole other discussion about what would constitute reasonable aggression, albeit I think we agree that committing 100+ hammers for a bunch of peripheral coastal razes in the first 100 turns of the game is not reasonable (edit: outside of special circumstances, like a duel).