Planetarion Forums - Assassin Mode (A solution to 'Bad Politics')

Blazde · 9 Mar 2004, 05:45

(Skip down to suggestion if you're not interesing in my musings that add little to the actual idea).

Background

This suggestion is inspired by the weekend's disastrous speedgame finals, which imo, totally undermined Planetarion's ability to function as a game. In PA, if you ally more than 50% of the universe, in all probability you'll defeat the rest and can claim victory. The structures within PA such as battle groups and alliances mean that such blocks can form through the interaction of surprisingly few players. (Potentially only Thunder, deZz and cbk were needed to create and sustain the speedgame block, though no doubt it was a little more complicated in practice). As I see it the situation is akin to the existence of a simple 'perfect strategy' undermining a board game where, perhaps, the player that goes first can always win or can always force a draw. For example, in tic-tac-toe both players can force a draw by playing a perfect strategy. With a decreasing player base in PA the problem is going to sneak into real rounds more and more.

The motivation for players to hijack the game in this way is exactly the same motivation which drives most PA players, namely to win. We might hope that players would have more respect for the game itself, than to the agreements they strike with other players but it's been proven time and again that many don't. PA instills a fanaticism in people that sets it apart from other games, and ultimately my analogies about having a friendly game of Counter Strike or playing footie with your mates, and switching sides often, are irrelevant. Honouring agreements with fellow players is a very righteous and easy defensible reason for allowing these blocks to persist. Whether it's the true reason or just an excuse to win doesn't matter because once someone's stated their core values like that, no amount of bitching will change their mind.

The game however, can be changed. Spinner I know has talked about this as a fundamental problem a number of times. The problem is that players can cooperate, and that it's exactly this cooperation that makes PA the game it is, so you cannot simply take it away. The solution is to limit it in some way. PAX went some way toward this with the -1 ETA for inter-alliance defence, coupled with a limit on alliance size. A big step in my opinion, but more importantly something is needed to limit offensive cooperation. Unfortunately I don't think there's anything quite as simple, so this suggestion is more of a radical redesign that will never happen, but regardless I'd rather put my anger over the weekend's events to some fanciful constructive use rather than continuing to moan in this thread. Kal's thread incidentally, has some similar ideas. I'd be interested to see any more suggestions people have for this problem.

My inspiration came last night when I fired up Rise of Nations in an attempt to forget about PA. As I configured a game with me and 3 computer players it dawned on me that exactly the same problem afflicts many multiplayer RTS games. In any game that allows diplomacy, the players who brought the most mates to the game are likely to win. In a game with preconfigured teams, you can be put at a severe disadvantage when one of your team leaves the game unexpectedly. However, I noticed RoN has an Assassin Mode precisely to solve this. For those that haven't seen it, each player is assigned a target player, such that everyone has exactly one other player targeting them, and attacks are always directed around the map in a clockwise or anticlockwise direction. When you defeat a player, their target becomes your target. You can, infact, attack other players besides your target but your units suffer a severe (attrition) disadvantage so it's impractical to conquer them. The balance this gives the game is dramatic and you have to play it to fully appreciate the subtleties. Adapting the idea to PA is problematic because there's no concept of 'defeating' someone, and clearly it would be boring to have the same target all round. This leads to probably the most radical aspect of my idea (retargetting) but it allows some other problems in PA to be addressed (bashing/multiing/exponential growth). There's one or two other things thrown in that aren't strictly necessary (semi-private galaxies, geography, clusters), but like I said it's fanciful, and I'm bored. Anyway, here's the idea.

Suggestion

Players form and join alliances before tick start, alliances are limited to a size that's configurable and dependant on the number of players expected to play the round etc...
HC of smaller alliances may elect to band together to form blocks up to the alliance size limit. The alliances are still separately named and individually ranked but for most game play purposes they count as a single alliance. In particular, they get -1 eta defence between them, and may not attack (or covert-op) each other.
Players are randomly shuffled into galaxies consisting only of their alliance mates (perhaps including other alliances in their block), with further ETA advantages within galaxy of course.
Galaxies are placed together in clusters with other galaxies from the same block (this validates the defensive ETA advantage, the planets are closer to you).
1. Clusters are ordered by strength (total roid count, score etc...) with the strongest in cluster 1, weakest in the highest cluster, with 1:1 being moved elsewhere so as not to confuse things! Players may only attack planets in clusters lower than them without severe penalties. Note that the strongest cluster will be in a purely defensive situation, unable to efficiently roid anyone else. The weakest cluster in a purely offensive situation.
2. Alternatively clusters may only attack (without penalty), clusters one or two up from them.
3. Alternatively the clusters are ordered with the two strongest together, surrounded by the next two strongest, the next strongest surrounding them etc.. For instance, if there are 10 clusters numbered 1 (strongest) to 10 (weakest) they're arranged like this:
  
  9 7 5 3 1 2 4 6 8 10
  
  Clusters may attack the two surrounding clusters (or perhaps the 4 surrounding clusters), and from the first cluster to the last and vice-versa. Ie. in the example 9 and 10 may attack each other as if they were next door.
At appropriate time intervals (weekly or bi-weekly, and perhaps also immediately before protection ends for most people, at PT48/72), the clusters are reshuffled by strength as in 5.
Attacking into a cluster that you shouldn't is either not allowed, or suffers an ETA disadvantage, a combat disadvantage (pilots are tired from the long haul flight so armour/agility and damage/accuracy is lowered, or some percentage of ships simply don't arrive at the target or return home, they get lost).
Players may leave/join alliances, alliances may break (and remake) blocks with other alliances. All changes take place at the next shuffle (eg. Players will be assigned new galaxies, alliances are broken up and their strength reevaluated, new blocks form etc..).
Scanning is limited to planets you can attack (or is very inaccurate into clusters you shouldn't be attacking), and to planets that are already attacking your cluster.
Allianceless players are merged into 1 or more clusters and encouraged to work together (initially through ingame cluster forums).

Alternatives

I list 3 alternatives in 5 that are the obvious arrangements, but I don't think a and b are really viable. With a, no one wants the top-cluster since you simply won't grow. Also, the smaller clusters may gang bang the bigger ones eg. The strongest cluster getting hit for a week by the number 2, 3, 4 and 5 clusters, without being able to counter attack, is certainly not fair. An alternative is to shuffle very often (daily), so the top cluster won't remain there very long if it's being targeted, but that would almost eliminate the interesting politics and make wars too fluid. The problem with b is that unbridgeable score gaps may open up between the clusters, dividing the universe into two or more chunks which is no fun for anyone.

Option c poses the best balance of interesting politics, the possibility for cooperative 2 (clusters) on 1 action with agreements forming between clusters 1 apart etc... but still limits block size severely and means that no one cluster can ever be bashed for too long. Longer lasting agreements can still hold because clusters next to each other don't have to fight (they each have another target cluster), but clusters will have to choose their allies carefully and be prepared to break them, or they'll totally run out of targets (but won't spoil everyone else's game). Also it allows players to counter attack / retal their attackers since everyone you can attack can also attack you.

Cons

The problems I can see so far common to all options are:

Clusters are now the same as alliances and so you lose the original concept of clusters, including intra-cluster wars, cluster alliances, a chance to meet players outside your alliance, and a chance for new players to meet experienced players. (One possibility to solve this last problem is to merge allianceless players into galaxies with have an alliance, instead of out on their own with other allianceless people). I personally missed clusters a lot in PAX and believe removing them was one of the biggest mistakes made.
Alliances might break up into smaller elitist clusters in order to prey on weaker clusters for cheaper roids. This might not be such a bad thing, since they'll become more vulnerable, and probably reduce the size of their roid catches, but it also means they can't be targeted by the alliances in competition with them. An alternative is to rank the clusters by average score instead of total score.

Pros

The major advantages:

Alliances that form massive blocks will have zero targets, instead they must directly attack the alliances in competition with them in order to grow.
Active new players are encouraged to form together into small alliances where they can feast on roids from the weaker clusters, before slowly moving up the rankings without fear of being roided by the huge, well-connected players, or they can band with other fledgling alliances and trying their luck in the stronger clusters. (Here's the counter argument for ranking small elite clusters by average, you want exactly that situation with new active players).
Semi-active players/alliances aren't continually bashed to death, instead they slip into the weaker clusters where the pace is more relaxed.
Maintaining a multi/scan planet is useless unless you keep it in the same alliance as your main planet, which is difficult in the major alliances where multiing is most damaging.

Anyway, I feel like I've written enough (*looks up*... wow), and expended the anger I had toward R6B/VsN. I hope whatever changes round 11 brings, it's a fair, fun, fight for all, and round 10.5 too.

Guran · 9 Mar 2004, 07:40

I don't know how many will actually read your post but atleast I did.

I think you have a somewhat complicated approach to our problem or should I say, a complicated solution but atleast you have given much thought to this. I don't necessarily like your idea as a hole but it has some valid points which could make this game more enjoyable in the future.

Main issues imo are, how to prevent blocking and domination of one alliance over the others.
Some sort of solutions could be found in travel times and smaller alliances. I'll dwell on this and reply again on my lunch break when I've had more thought on this issue.

mist · 9 Mar 2004, 12:17