A strategic dilemma

[ComradeRob]

Here's a fun little question:

Your alliance is ranked 3rd. The top two alliances are opposing each other, and the #1 alliance is beginning to pull away. It's not an unstoppable lead yet, but they're fighting very well and showing no signs of slowing down. The 2nd-placed alliance is starting to worry; they've wasted part of their round picking fights with others and now doubt that they have the time or the motivation to win alone.

You want to win the round. What should you do?*



* Try to answer this question as if you're in command of this imaginary 3rd-placed alliance. I'm not referring to any alliance in particular and though the scenario is obviously based on this round, it's a scenario that has occurred before. Since the 3rd-place alliance never wins, I want to know if there's some clever strategy they're missing.

[Synthetic_Sid]

You need to provide more information for a proper answer.

How clear are you of #4/5/6 etc? Specifically, are there any other alliances who may be in a similar position to you?

Do the alliances at #1/#2 have any agreements which could potentially impact significantly on the situation?

How important is finishing #1 to your alliance? This is a key issue in PA politics. Some alliances at #3 would turn down a 100% chance of finishing #2/#3 for a 5% chance of winning and a 95% chance of not finishing top 5. Other's wouldn't.

Which of the #1/#2 could you beat 1 v 1 if you ended up in a head to head fight with them?

Do you have the necessary staff, and do your members have the motivation necessary to maintain a high-level of sustained and concentrated activity of the remainder of the round?

In any situation like this you need to work out scenarios, work out the possible outcomes, calculate (or guesstimate) the likelihoods of the involved factors and hence the likelihood of the outcomes. Then measure the rewards/risks against your own alliance's level of risk-aversion.

Without more information it's impossible to give any meaningful answer.

[ComradeRob]

Originally Posted by Synthetic_Sid

How clear are you of #4/5/6 etc? Specifically, are there any other alliances who may be in a similar position to you?

Let's assume that there are a couple of other alliances in a similar position (just to make things a bit more difficult).

Originally Posted by Synthetic_Sid

Do the alliances at #1/#2 have any agreements which could potentially impact significantly on the situation?

No. At the most, they may have some temporary understandings with others.

Originally Posted by Synthetic_Sid

How important is finishing #1 to your alliance? This is a key issue in PA politics. Some alliances at #3 would turn down a 100% chance of finishing #2/#3 for a 5% chance of winning and a 95% chance of not finishing top 5. Other's wouldn't.

I'm starting from the belief that finishing #1 is the ultimate aim of all alliances who regularly finish in the top 5, in much the same way that teams which finish in that area in football leagues generally aspire to win next season. They might not be favourites to do it, but they believe it's possible and that they should make every effort to do so.

Originally Posted by Synthetic_Sid

Which of the #1/#2 could you beat 1 v 1 if you ended up in a head to head fight with them?

1 v 1 fights are rare, in my experience. If necessary, politics could be used to avoid a 'fair' fight; that's part of the job of an alliance leader.

Originally Posted by Synthetic_Sid

Do you have the necessary staff, and do your members have the motivation necessary to maintain a high-level of sustained and concentrated activity of the remainder of the round?

I'd say there's only one way to find out.

[Synthetic_Sid]

Looks like noone else wants to play Rob.

You didn't address my question about risk-averseness - which I remain convinced is a key issue. But let's assume for now that you'll take the course which has the most likelihood of getting you #1.

The first thing to realise is that you can't possibly win by making a long-term deal with either of #1 or #2 - you can get 2nd like that but not 1st. And neither of #1 or #2 is going to want a short-term deal which has a favourable outcome for you.

As there's other alliances in a similar position to yours, it's those that you need to work with. And you need to agree that jointly you'll take down first one then the other of #1 and #2. I'd take down the weaker of the two first for three reasons:

1. If the round finishes before you win you're likely to have a higher rank than if you did it the other way round.
2. The weaker of the two is less likely to be able to regrow after you've bashed them - and you MUST bash them far enough down that they can't re-pass you when you switch to hitting the other.
3. By leaving the stronger to last there's less incentive for your partners to back out of the deal prematurely - as they know they need to team up on them to win.

The fundamental difficulty in any such scenario is that a set of Prisoner's Dilemma type conditions exist in the interactions between #1, #2 and (#3+#4+#5) - whereby the optimal short-term action is very rarely the one which will achieve your best long-term result and you have to rely on one of the other parties also making a sub-optimal short-term action as well.

I'm going to get accused of propaganda here - but I firmly believe that with 2 alliances above you, your chances of winning are less if you take the stronger out first than if you take the weaker out first. The worst possible option is, of course, to keep swapping too and fro between hititng the two of them - as then you risk the possibility of #1 and #2 realising that BOTH of their chances of winning improve if they ignore one another for a while and smack down #3, #4 and #5 first.

On a side-note, your scenario is actually far more applicable to earlier this round than it is to now (partly because of time-scale, partly for other reasons).

[Wishmaster]

Make deals with other top 5 to target #1

#1 will most likely be targetting their main competator #2 at that time, and ur alliance should be able to close inn / overtake them.

Depends how early in round though

[Synthetic_Sid]

Originally Posted by Wishmaster

Make deals with other top 5 to target #1

#1 will most likely be targetting their main competator #2 at that time, and ur alliance should be able to close inn / overtake them.

Depends how early in round though

Wrong answer. #1 will only target #2 in a very small number of situations: the main one being that they are stupid.

#1 knows that the only reason #3/#4/#5 are hitting them is because they think they increase their chances of winning by doing so. Hitting one of them and leaving #2 alone massively decreases your target's chances of winning compared to them pulling out of the attack. Plus as noone is targetting #2, they grow stupidly fast - so even the two of #3/#4/#5 that you aren't hitting are not making any overall headway to winning.

NOTE: I'm assuming pure game-theory conditions here, where noone involved cares how any of the others do, but wants to maximise their own chance of winning.

[Synthetic_Sid]

To clarify a bit more on my response to Wishmaster:

The weaker of the alliances in the block gain the least from being in it - as their chance of winning is the lowest. As their benefit is the smallest, you can put them in a situation where remaining in the block is most definitely against their own interests (as their chance of winning increases if they leave it - even if only from none to barely measurable). Becasue your chance of winning is 0 if you don't break the block you HAVE to act in a fashion which will cause the block to shrink. And attacking #2 will NOT do that in useful time-frame.

[Wishmaster]

as I said, depends on what stage of the round.

If its close to round end, #1 would be attacking #2 in order to prevent them overtaking them.

[Wishmaster]

Originally Posted by Synthetic_Sid

To clarify a bit more on my response to Wishmaster:

The weaker of the alliances in the block gain the least from being in it - as their chance of winning is the lowest. As their benefit is the smallest, you can put them in a situation where remaining in the block is most definitely against their own interests (as their chance of winning increases if they leave it - even if only from none to barely measurable). Becasue your chance of winning is 0 if you don't break the block you HAVE to act in a fashion which will cause the block to shrink. And attacking #2 will NOT do that in useful time-frame.

Depends what the alliance wants out of the -partnership- with other alliances.

Also - if more than 2 alliances go with #2 on #1, they can hope they wont be targgeted as heavy as the others helping #2, and therefore have a smal chance of gaining ranks / get #1.

[Synthetic_Sid]

Originally Posted by Wishmaster

as I said, depends on what stage of the round.

If its close to round end, #1 would be attacking #2 in order to prevent them overtaking them.

Yes - but that's not relevant to this scenario. As if the round's so late that #1 can win by attacking #2 WHATEVER #3/4/5 do then, by definition, #3/4/5 have 0 chance of winning anyway.

[Wishmaster]

Originally Posted by Synthetic_Sid

Yes - but that's not relevant to this scenario. As if the round's so late that #1 can win by attacking #2 WHATEVER #3/4/5 do then, by definition, #3/4/5 have 0 chance of winning anyway.

I said that #1 MIGHT be able to do so with attacking #2. If its a matter of days, then yes its not much they can do.

Also - it depends what faith u have in your alliance, and how u consider the other people u could block with. If #1 alliance knows ur alliance for sure wont give up because of some incs, then they will not be targetting u in the first place, and will rather go for a weaker / poorer community (mercenaries..people they think will collapse) and thus u can still be able to get #1 or #2.

[Synthetic_Sid]

Originally Posted by Wishmaster

Depends what the alliance wants out of the -partnership- with other alliances.

In the scenario Rob has laid out ALL participants want to maximise their own chance of getting #1. If we want to deal with a situation where Alliance X likes Alliance Y more than Alliance Z then we'd be dealing with an entirely different scenario.

[Wishmaster]

care to point out where he stated that?

[Synthetic_Sid]

Originally Posted by Wishmaster

I said that #1 MIGHT be able to do so with attacking #2. If its a matter of days, then yes its not much they can do.

Also - it depends what faith u have in your alliance, and how u consider the other people u could block with. If #1 alliance knows ur alliance for sure wont give up because of some incs, then they will not be targetting u in the first place, and will rather go for a weaker / poorer community (mercenaries..people they think will collapse) and thus u can still be able to get #1 or #2.

This is a game theory type discussion. Talk of things like internal strength of an alliance is irrelevant - to take those into account here we'd need to discuss a specific situation with the alliances named, the relationships between them defined etc etc. My first post asked about risk aversion: Rob's response clearly indicated that he didn't want to complicate a theoretical situation. As no specifics are provided for the alliances in question we have to assume that all are broadly similar.

[Synthetic_Sid]

Originally Posted by Wishmaster

care to point out where he stated that?

As no information is given for any of the alliances in this theoretical situation, how on earth can we take into account the undefined strengths and weaknesses of them?

I believe you're mistaking this thread for a "what should people do about 1up" thread. If you want one of those, then go and make one. And leave this one for discussing a #1, #2, #3 etc who are all generic alliances that make decisions based on game-theory.

[Synthetic_Sid]

Originally Posted by Synthetic_Sid

I believe you're mistaking this thread for a "what should people do about 1up" thread.

Actually, don't make a thread about it. Have a meeting instead.

[Wishmaster]

Originally Posted by Synthetic_Sid

This is a game theory type discussion. Talk of things like internal strength of an alliance is irrelevant - to take those into account here we'd need to discuss a specific situation with the alliances named, the relationships between them defined etc etc. My first post asked about risk aversion: Rob's response clearly indicated that he didn't want to complicate a theoretical situation. As no specifics are provided for the alliances in question we have to assume that all are broadly similar.

Its hardly realistic to assume all alliances r broadly similar. but anyway, lets assume that for now.

If thats the case, and they decide to ally 4 others to take down #1, and all #4 alliances are equally good, then ur alliance has a 25% percentage chance of being targetted by #1.

Most would take this chance and have a slim chance of ending #1

Also - if all alliances r equally good, I guess the score diff. between top 5 is V smal, and #1 might consider #3 or #4 more / equally as dangerous as #2.

[Wishmaster]

Originally Posted by Synthetic_Sid

As no information is given for any of the alliances in this theoretical situation, how on earth can we take into account the undefined strengths and weaknesses of them?

I believe you're mistaking this thread for a "what should people do about 1up" thread. If you want one of those, then go and make one. And leave this one for discussing a #1, #2, #3 etc who are all generic alliances that make decisions based on game-theory.

to discuss a situation, u have to be able to take diff. "cases" into consideration. #3 clearly has to make diff. choices based on the other alliances around them. There is not 1 given correct answer on this dilemma.

But if u dont want to discuss it further, and think I m talking about this round, then no point going on.

[Synthetic_Sid]

Originally Posted by Wishmaster

Its hardly realistic to assume all alliances r broadly similar. but anyway, lets assume that for now.

If thats the case, and they decide to ally 4 others to take down #1, and all #4 alliances are equally good, then ur alliance has a 25% percentage chance of being targetted by #1.

Most would take this chance and have a slim chance of ending #1

Also - if all alliances r equally good, I guess the score diff. between top 5 is V smal, and #1 might consider #3 or #4 more / equally as dangerous as #2.

Some good points in there. What Rob maybe needs to clear up is whether the difference in score between the alliances is due to difference in ability - or difference in circumstances. Put simply, if all relevant alliances were in no wars at all, would they all grow at the same rate?

We have to assume from the information given by rob that there's a signifcant gap between #2 and #3 - otherwise he would have just said "you're one a group of alliances all a similar distance behind #1". That is a different situation entirely to the one being discussed - and a whole different thread could be made discussing whether your chance of winning is better siding with #1 and hoping the rest hit them not you, or joining with the rest and hoping you don't get singled out by #1.

[lokken]

Stop worrying, join ascendancy and go to bed at 1am.

[Wishmaster]

Originally Posted by Synthetic_Sid

Some good points in there. What Rob maybe needs to clear up is whether the difference in score between the alliances is due to difference in ability - or difference in circumstances. Put simply, if all relevant alliances were in no wars at all, would they all grow at the same rate?

agreed

Originally Posted by Synthetic_Sid

We have to assume from the information given by rob that there's a signifcant gap between #2 and #3 - otherwise he would have just said "you're one a group of alliances all a similar distance behind #1". That is a different situation entirely to the one being discussed - and a whole different thread could be made discussing whether your chance of winning is better siding with #1 and hoping the rest hit them not you, or joining with the rest and hoping you don't get singled out by #1.

IMO u cant talk abouty a scenario like this without taking into consideration the diff macro threaths. U also have to analyze #1, who will they target, who will they think r going to fold first. Then u have to analyze ur own situation and c if u r able to take the heat, and if u r able to come out of it on top.

Its IMO just stupid to discuss this at all, if u dont have a scenario to go by, or many diff. ones.
Its would be like discussing whats the smartest moves to make in a chess game, without taking into consideration what ur opponent does.

If #3 join with #1 I dont c how they could win.

Most likely #1 and #3 would be strong enough to keep the others at bay, and giving #1 the win. If not, the 3 others would take down #1, and then they would with ease take down current #3 because they have already destroyed #1 which leaves #3 with as good as no allies.

[Synthetic_Sid]

Originally Posted by Wishmaster

to discuss a situation, u have to be able to take diff. "cases" into consideration. #3 clearly has to make diff. choices based on the other alliances around them. There is not 1 given correct answer on this dilemma.

But if u dont want to discuss it further, and think I m talking about this round, then no point going on.

I'm happy to discuss it further. I agree you have to look at different scenarios. The problem I had with your earlier posts was that you seemed to be basing your argument on an assumption that #1 would act in a way which isn't their best option.

I suspect that it's actually impossible to come up with a "correct" answer without some values, growth rates and time-scales.

The reason why I say that is because the reaction of #1 actually depends VERY much upon time-scale. The key question being: can they reduce to zero the chance of #3/#4/#5 winning the round BEFORE their own chance drops low. If they can, then if they hit any one of 3/4/5 then that one HAS to stop hitting 1 - or they have no chance to win the round at all. If 1 can't hurt 3,4 or 5 quickly enough then 3,4,5 don't need 2 in the first place - and can take 2 out THEN 1, giving them a 1/3 chance of winning after they beat 1 rather than a 1/4.

[lokken]

To be serious:

I'm tempted by "NAP with 1st and go for 2nd because the fact you are so far behind suggests you and your 'associates' can't realistically fight your way out of a paper bag" as a realistic option. The fact is thus: one alliance is going to be pretty concerned that two of the others are plotting their downfall once the current leaders are done and dusted. They'll pull out and go for rank eventually to protect themselves, which will pretty much screw your own campaign. One alliance will inevitably lose momentum and cause the 'plan' to go wrong, because that's how alliances in the 3-5 region usually pan out. The fact that you're relying heavily on others suggests that you probably can't win the round.

If you can fight, then things are pretty much going to change on a daily basis and one mistake or one irrational/unexpected move from someone else might well end up dooming you. Edit: And it usually does.

[Synthetic_Sid]

Originally Posted by Wishmaster

If #3 join with #1 I dont c how they could win.

Most likely #1 and #3 would be strong enough to keep the others at bay, and giving #1 the win. If not, the 3 others would take down #1, and then they would with ease take down current #3 because they have already destroyed #1 which leaves #3 with as good as no allies.

I'll make this my last post on that alternative scenario - or the thread will get too confusing. The goal of 3 in a situation where they make a deal with 1 is to act in a way to control the speed at which 1 loses. They need 1 to lose - and need to adjust the degree of help they give 1, so that although 1 loses they don't do it fast enough that the others have time to finish 3 off afterwards. And the first stage in any such deal with 1 is to make a condition of your support that 1 has to attack all of the others at once rather than focussing on one of them.

[Wishmaster]

Originally Posted by Synthetic_Sid

I'm happy to discuss it further. I agree you have to look at different scenarios. The problem I had with your earlier posts was that you seemed to be basing your argument on an assumption that #1 would act in a way which isn't their best option.

I suspect that it's actually impossible to come up with a "correct" answer without some values, growth rates and time-scales.

The reason why I say that is because the reaction of #1 actually depends VERY much upon time-scale. The key question being: can they reduce to zero the chance of #3/#4/#5 winning the round BEFORE their own chance drops low. If they can, then if they hit any one of 3/4/5 then that one HAS to stop hitting 1 - or they have no chance to win the round at all. If 1 can't hurt 3,4 or 5 quickly enough then 3,4,5 don't need 2 in the first place - and can take 2 out THEN 1, giving them a 1/3 chance of winning after they beat 1 rather than a 1/4.

Their best option varies, which was what I tried to point out.

I also think its impossible to come with a correct answer on this. It will varie from alliance to alliance whats the correct thing to do. also based on situation ofc.

If we give 3,4,5 enough time to take down #2 and then #1..its hardly needed a block at all at that time.

[Synthetic_Sid]

Originally Posted by lokken

To be serious:

I'm tempted by "NAP with 1st and go for 2nd because the fact you are so far behind suggests you and your 'associates' can't realistically fight your way out of a paper bag" as a realistic option. The fact is thus: one alliance is going to be pretty concerned that two of the others are plotting their downfall once the current leaders are done and dusted. They'll pull out and go for rank eventually to protect themselves, which will pretty much screw your own campaign. One alliance will inevitably lose momentum and cause the 'plan' to go wrong, because that's how alliances in the 3-5 region usually pan out. The fact that you're relying heavily on others suggests that you probably can't win the round.

If you can fight, then things are pretty much going to change on a daily basis and one mistake or one irrational/unexpected move from someone else might well end up dooming you.

Where the problem lies, is that the optimal situation for ANY of 2/3/4/5 is for the other 3 to attack 1 while they sit it out. And in a game-theory situation all of them will be trying to find ways to reach that position.

Luckily, in a real round, that could never happen.

[Wishmaster]

Originally Posted by Synthetic_Sid

I'll make this my last post on that alternative scenario - or the thread will get too confusing. The goal of 3 in a situation where they make a deal with 1 is to act in a way to control the speed at which 1 loses. They need 1 to lose - and need to adjust the degree of help they give 1, so that although 1 loses they don't do it fast enough that the others have time to finish 3 off afterwards. And the first stage in any such deal with 1 is to make a condition of your support that 1 has to attack all of the others at once rather than focussing on one of them.

Dont c #1 making such a deal, as it would be just stupid to focus on all 3 others attacking them at the same time.
If #1 was stupid enough to make such a deal, then yes, #3 might win.
But how realistic is that?

[Wishmaster]

Originally Posted by Synthetic_Sid

Where the problem lies, is that the optimal situation for ANY of 2/3/4/5 is for the other 3 to attack 1 while they sit it out. And in a game-theory situation all of them will be trying to find ways to reach that position.

Luckily, in a real round, that could never happen.

therefore, since it never would occur in a real round, whats the point discussing it?

[Synthetic_Sid]

Originally Posted by Wishmaster

Dont c #1 making such a deal, as it would be just stupid to focus on all 3 others attacking them at the same time.
If #1 was stupid enough to make such a deal, then yes, #3 might win.
But how realistic is that?

Surely it depends on whether any of the others offer #1 a better deal? Accepting/rejecting such a deal isn't as cut and dried as you might think. Hitting 3 alliances at once with someone else attacking alongside you isn't really that much worse that hitting 1 alliance but having 4 attack you back. But once again it depends on thngs like rate of growth, how fast roid loss grows with X+1 atatcking alliances instead of X alliances etc etc.

[Synthetic_Sid]

Originally Posted by Wishmaster

therefore, since it never would occur in a real round, whats the point discussing it?

I think I forgot to put [sarcasm] and [/sarcasm] around that comment by mistake.

[robban1]

the best chance you ever get as third is that you somehow manage to toss 1 and 2 on eatchothers and stay out of the trubble as there is very little gain in a 1vs1 war

[lokken]

Originally Posted by Synthetic_Sid

Where the problem lies, is that the optimal situation for ANY of 2/3/4/5 is for the other 3 to attack 1 while they sit it out. And in a game-theory situation all of them will be trying to find ways to reach that position.

Luckily, in a real round, that could never happen.

My point of view is that these alliances that go for #1 have a track record of crashing and burning, generally because they won't be good enough. So you might as well as ruin someone's day and get the best rank you can.

ND did it pretty well in round 12 and secured what at the time was its best ever rank. LCH ruined our day in r15 and massively improved their rank (but didn't overtake ND because they were coming from 5th or 6th), but it wasn't half a good move on their part. In their case LCH allied with the #2 alliance at the time, but one might argue that ND was in an artificially high position.

[ComradeRob]

Originally Posted by Synthetic_Sid

Where the problem lies, is that the optimal situation for ANY of 2/3/4/5 is for the other 3 to attack 1 while they sit it out. And in a game-theory situation all of them will be trying to find ways to reach that position.

Luckily, in a real round, that could never happen.

I'm trying to think of an example of it happening, but I think you are probably right. It might be possible to deceive other alliances into serving your agenda in the short term, but I don't think you could win the round with that approach.

The problem, as Sid outlined, is that the leading alliance can simply threaten to destroy the weakest alliance that cooperates against them. This makes it very difficult to assemble a block against the leading (or even the 2nd-placed) alliance.

The correct option might be to begin with an easier challenge - taking down the #2. This will be made easier if the #1 alliance is also attacking them and even easier again if the #1 alliance will offer you planetary NAPs. This would provide time for the practical issues of cooperation to be solved and for attack systems to be perfected. At a later point (but not too later) when the former #2 has dropped to, say, 4th or 5th, it would be wise to NAP that alliance and drop the planetary NAPs with the #1, in order to begin a full-scale assault. There's still a decent chance of failure, but that might be the best scenario for a 3rd-placed alliance. Assuming by this point that the #1 is some way out in the lead, but there is a "chasing pack" of alliances on similar scores, it might be possible for all of them to believe that they can win. Buoyed by an easy victory over the former 2nd-placed alliance, momentum might just carry them to victory.

To be honest, I suspect that this kind of thing might rely on personal qualities rather than game theory mechanics. An inspiring leader may be necessary to motivate the membership, and diplomatic skills will be necessary to ensure cooperation.

Alliances may have to be willing to make deals that will be unpopular, such as permanent NAPs. If I were running the leading alliance in a block, I would make sure that the smaller alliances knew that, once victory was achieved, they would be protected as much as possible. This is the only counter I can think of to the strategy of attacking the weakest link.

[Synthetic_Sid]

In "real" PA, game-theory just don't apply. In every block I've ever seen which had more than 2 members in, at least one of the block members wasn't playing with any intent or expectation of becoming #1 alliance.

Real blocks form late in the round (and sometimes early) for one of two reasons:

1. Alliances would prefer A to win than B.
2. Alliances believe that assisting A will likely give them a higher finishing rank than assisting B.

If you're HC of an alliance which is (or is likely to be) a contender for #1 then one of priorities during the round is to ensure that at the latter stages you're A not B in the 2 points I just made should the appropriate circumstances arise.

[robban1]

there is always just one allie who gains lot on a blocking. the weeker one doesnt get a bigger chance of passing the target of the block or even the partners they get in the block so my first suggestion is the best one for a third rank allie

[Synthetic_Sid]

Originally Posted by robban1

there is always just one allie who gains lot on a blocking. the weeker one doesnt get a bigger chance of passing the target of the block or even the partners they get in the block so my first suggestion is the best one for a third rank allie

I agree largely with you. In the scenario being discussed, where all the allainces are trying to win, the correct answer for #3 is to block with the alliances of similar strength to themselves and not to cooperate directly with either #1 or #2. By working with alliances of similar size and ambitions to yourself you can be far more confident that your interests coincide - at least until you're both the only remaining valid contender to win, or no longer in the race.

But this is in an ideal scenario where all participants have as a sole goal winning the round - with all other outcomes considered equally (un)desirable.

[robban1]

Originally Posted by Synthetic_Sid

I agree largely with you. In the scenario being discussed, where all the allainces are trying to win, the correct answer for #3 is to block with the alliances of similar strength to themselves and not to cooperate directly with either #1 or #2. By working with alliances of similar size and ambitions to yourself you can be far more confident that your interests coincide - at least until you're both the only remaining valid contender to win, or no longer in the race.

But this is in an ideal scenario where all participants have as a sole goal winning the round - with all other outcomes considered equally (un)desirable.

not as it is in this round then

[Alki]

Interesting question, myself personally, i'd approach it by making no deals with either #1 or #2. In today's game mechanics its not needed, id look at average score gains per ally per day, set the goal of how much is needed and work at it. I would pick out fat targets but making sure i never singled an alliance out per night, id then pick out members from my alliance who would gain max xp from these targets, id then have the rest of the alliance escourt them to there target, and let them go out roiding newbies with there 2nd slot. 3rd slot for defence of course.

In my opinion as the 3rd alliance ur pretty much boned in any scenario if you choose to cooperate with either. Cooperate with the 2nd alliance, and the 1st alliance will target you, if they have any sense, that is how I personally like to do it, work from the bottom to the top.

Cooperate with the #1 alliance, you pretty much secure there victory for pummeling the #2 alliance, the #2 alliance will most probably seek allies to take care of you aswell.

If you leave the parties in a 1v1 fight niether gain anything significant. I do believe if you apply my 'plan' you have a fairly good chance at hitting the top stop with huge gains of score per day, in today's game mechanics that is all you can hope for aswell, which is pretty saddening.

My approach may not be elegant, nor would people like it but if its an attempt to win you the round id take it any day. Obviously no alliance is perfect, it would be pretty ****ing hard to pull it off in today's game, you'd need no gloryhunters/mercenaries, no disloyal members, 100% trust, and people willing to sacrifice quite alot of time.

Infact in todays game that would be impossible

[JonnyBGood]

Originally Posted by ComradeRob

Here's a fun little question:

Your alliance is ranked 3rd. The top two alliances are opposing each other, and the #1 alliance is beginning to pull away. It's not an unstoppable lead yet, but they're fighting very well and showing no signs of slowing down. The 2nd-placed alliance is starting to worry; they've wasted part of their round picking fights with others and now doubt that they have the time or the motivation to win alone.

You want to win the round. What should you do?*



* Try to answer this question as if you're in command of this imaginary 3rd-placed alliance. I'm not referring to any alliance in particular and though the scenario is obviously based on this round, it's a scenario that has occurred before. Since the 3rd-place alliance never wins, I want to know if there's some clever strategy they're missing.

I'd go speak to #4 and #5 and attempt to ally with them to take down whoever requires taking down, either #1 or #2 as the situation arises. However be very carefuly initially to only hit #1. An alliance can focus on one of the alliances in your block but because there are three alliances it's different to really split them up. Even if they do focus on one of the alliances it's a) unlikely to be yours (see comments made by sid) and b) if one alliance does drop out of the war you've still got two alliances going strong. This gives improved expectation to alliances #3, #4 and #5 in terms of winning the round and, obviously, the best shot to your alliance.

[Kjeldoran]

What would help is putting pressure on the #1 alliance and let the #2 alliance gain up on them, considering you gain on both aswell.

#1 will get nervous and might wanna concentrate on hitting #2, not full front but maybe just enough to ensure they don't outgrow them. If this is the case then it means less fleets from #1 heading your way and less defence fleets vs your targets from #2 heading against your own attacks.

I think the key is, in the situation described by Rob, to find a way for #1 and #2 to somewhat target eachother, even for the smallest bit. #1 will not uphold a nap with #2 if #2 is rapidly gaining on them. #2 will on their turn respond to the actions of #1 against them.

Ofcourse, this all sounds nice in theory and I don't really know how you'd pull it off, since the solution here is rather simplistic.

Another important point is to move up as a block (#3 4 & 5) - As a block I mean equal growth, not as in forming a huge powerblock - because therefor none of the alliances can concentrate too long on 1 of them since that won't secure their position either. If #3 greatly increases the gap between them and #4, then it'll only be easier for #1 and #2 to concentrate on #3, without risking #4 to catch up too fast.

rgds Kj

[Mighteh]

well here is the thing.
if those #1 and #2 alliances were naped and cooperating from the get go, then its a reason for serious concern to those on #3, #4, #5 etc spots, and that will require serious blocking of, hopefully, ranks 2 through 7 alliances.
In case #1 and #2 alliances had no friendly relationship throughout the round, there always will be hostility between these two, just due to the simple facts that #1 and #2 alliances usually posess the highest roid count, and, all in all, this game is about gaining and keeping your roids.

In this case scenarion, you have to evaluate the chances of both alliances to end up in #1 spot when ticks stop, and block yourself with the alliance who has the least chance at that. From the point on when #1 alliance will be toppled by #2, #3, #3 and maybe even #4 ranked alliances, then you can reevaluate situation again, and apply same rules.

Winning is not about upholding your principals (eventho, if possible, its only better if those principals are upheld) but about best political and tactical decisions.

There is only one possible case when #3-5t alliances are pretty much screwed. If #2 assumes neutral position and will play catchup with #1 (A much harder way for #2 to actually win the round, but not impossible.) That might, or might not presume a non agression pact between #1 and #2 spots. In any way, as long as #1 does not specifically target #2 and vice versa, #3-#5 spots are, pretty much, fubored...

[Synthetic_Sid]

I think we've probably discussed this about as far as we can without getting into specific alliances.

What I find interesting is that most posters seem to agree with my initial response on this - that #3 (and #4/#5) shouldn't work with either of #1 or #2. Yet if you look at recent PA history, the majority of times this situation actually arises that's precisely what happens. Let's not pollute this thread by discussing specific alliances or mud-slinging, but here are my thoughts on why it is that alliances often make what most of us in here believe to be the wrong move in the above scenario:

1) History/Vendettas/etc: A lot of the time alliances make incorrect (from a game theory perspective) decisions because they assign value to what happens to specific other alliances rather than just focussing on what they achieve themselves.
2) Risk Aversion: While in the above scenario we assumed all alliances' sole goal was to finish #1, in reality most alliances also assign a value to finishing #2, #3, top 5, top 10 or whatever. This is not necessarily a bad thing: it is entirely possible that for some alliances the long-term bad effects of not finishing in the top X are substantial enough that their best decision to maximise value for them is not to adopt a "#1 or bust" approach.
3) Lack of ambition: If an alliance never actually intended or expected to win a round then it's nor surprising if their decisions don't maximise their chances of winning. Again, this is not necessarily a bad thing: we could have a whole different thread on whether or not an alliance should strive for goals which are easily achievable, reasonably welll achievable or a long-shot. I'm sure noone would claim that all alliances should be aiming for finsihing number 1.
4) #1 and #2. In the scenario described, it is often the case than one (or both) of #1 and #2 will try to convince the rest that the only choice they have is to pick a side. While this argument is largely specious, it can be a very convincing one. If sufficient of #3/#4/#5 etc fall for it - and side with either #1 or #2 then the argument then extended becomes "there aren't enough of you left solo to beat either of us, so you must join either with us or with them". That argument is still wrong - but few alliances have the motivation or ability to do anything proactive when there are already dominant blocks around.
5) Lack of will. By the time this sort of scenario arrives, the round has usually been running for a while. Officers/HCs are tired, member activity has dropped. The will just isn't there for any course of action which will most likely increase the work-load.
6) The big member vs alliance conflict. Any war is likely to have the most adverse effect on your bigger members. Many of these would rather settle for the rank they have, rather than risk their own planet for a slim chance of your alliance winning. You do wonder why they play a war game in the first place if this is the case - but the pressure is there. If those big members are also HC then this can be a real issue.

In an idealised scenario it's (relatively) easy to work out what to do. In a real round, there's a lot more factors in play - and the answers aren't always as obvious.

[Mighteh]

Hat's off to Sid for underlining the issue pretty clearly.

Here's 2 points i disagree with:

Point 4).: Eventho, clearly, #1 and #2 alliances will pull #3, #4 and #5 to their side, thus strengthening their clain for win of the round, its, sometimes, only desicion feasable for those 3-5 spot alliances, as, otherwise, they would be reduced to a role of farms for #1 alliance (wheather is its an active conflict and those alliances will be hit for some extra roid, or, either #2 or #1 will decide that the XP gained from capping of roids from #3-5 will significantly help them, scorewise, to get closer to the goal, and hence, become repeatedly roided for fwe nights) Its pretty much, "you either with us, or our ships are going to be coming your way" situation. And there is nothing wrong with it, its just how this game is played.
point 6).: Any truly big planet of any top alliance, ofcourse, does not wanna risk huge confrontation and, potentially, loose their top X spot. BUT there are few but's here actually:
1.: The obvious lack of fat targets that big planets face. Planets in my galaxy alternate between 2500 and 800 roids on a daily basis. So in order to survive, roids capped have to be HIGH up there. And, as we know, #1 and #2 alliances usually have the fattest planets, and, hence, best targets.
2.: Big planets owe aLOT to their alliance. And member's loyalty has to be prooven not only when its easy roiding, but when shit hits the fan aswell. True loyal members, no matter how big they are, will never step aside from a conflict threatening their alliance.
3.: If its indeed, alliance's HC, those people have to pursue alliance goal way beyond their own, personal ones. Eventho it is not the case sometimes, i would rather be HC of #1 alliance then #1 ranked planet, tbh. I know few people do not, but that constitutes a lacking HC. (In alliance ambition, devotion to alliance, and, simply, tactics.)
All in all tho, great post by Sid. I mostly agree with his assesment of the potential situation. Again, not naming any specific alliances, and, hypotheticly, situation not applied to current, or any other round. As when it comes down to putting ideas into being, there are other issues at stake, wich are too numerous to actually list. (such as staff- and member base, personal differences, tactical differences, what would 1 alliance request from another, etc...)

[Veil05]

What I would do is...

Id realise the situation I was in and make a hypothetical post on forums explaining the same situation my alliance is in. Wait for Synthetic_Sid or some other PA-Mogul to reply with what he would do in this situation and act on that.

[gzambo]

imo refering back to robs original question

#3 naps 1 and then hits #2 however as part of the nap only #3 hits #2 giving #3 the chance to overtake #2 so when that happens the nap ends and they pick their partners for the end of rd battle for 1st

[Ali]

I would also factor in any playing BG's and also any connections you would have beyond what is explained above. The more I think about the twisting scenarios the more i can not reply to the post without a specific example. But the ideal situation for #3 would be that 1,2,4,5 are warring and you are just floating. Ofcourse then you will be targeted so you need to have a plan already in mind to counter balance, it all depends on length left of round and your personal connections and so on...