View Single Post
Unread 13 Jun 2008, 09:04   #50
Banned
Banned
 
Banned's Avatar
 
Join Date: Jul 2003
Location: ******
Posts: 2,326
Banned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so littleBanned contributes so much and asks for so little
Re: No new alliances?

Quote:
Originally Posted by Gate
By increasing the memberlimit, there is a marginal cost in terms of officer time eg 20 more members may require 1 extra DC. But this cost is very small compared to the initial outlay eg you need 10-15 officers/HC to provide infrastructure, regardless of whether you have 20 or 100 members.
Insanity. I used to help run a 250 member alliance with about 7-10 people. The marginal cost in terms of officers is not linear with members. I'd say you need about 1 command level staffer for the first 20 members and then it starts dropping geometrically (the next 40 need 1, the next 80 need 1) or something like that.

Quote:
With a memberlimit of 30, that means there are 15 slots for non-officer members. With a limit of 90, there are 75 slots. Tripling the memberlimit quintuples the number of slots for non-officer players.
Follows from the counter example above that this is wrong.

Quote:
Of course, the alternative is that alliances get by on less officers, decreasing alliance quality so that members of most alliances have a worse time. Either way, given that active officers are a finite resource, I'm still convinced that lower memberlimits are a bad thing for the game. If you have the officers to generate new alliances (only anecdotal evidence supports this), you end up with poorer quality alliances or with less slots for players who can't commit to DCing/BCing.
Incidentally, the alliance that managed to get by with 7-10 active command staff won the round it played. From that I think we can safely assume that alliance quality is not necessarily decreased by decreasing the number of officers.

Quote:
Originally Posted by Gate
Heirarchy has been extensively tested; when these alliances lack DC/BC coverage, they get hammered.
Hierarchy's fail rate is still much higher than any alternative model....

Quote:
The 'Ascendancy' model hasn't been extensively tested; Ascendancy probably have the most skilled memberbase, a number of whom can BC/DC*. We don't know how an alliance would perform if it took the Ascendancy approach and didn't have a core of brilliant players. However, it would probably be the only viable model to generate alliances if we cut the member cap.

*it worked for DLR when they played out of tag too; but they're a group where every member should be finishing t100 and with no politics to worry about.
Sure, a number of whom can BC/DC, but few ever do. Mostly people sort their gal's defense. If you can't sort your galaxy's defense, you're screwed anyway. A lot of the people in Ascendancy who handle incoming would crumble under the usual tasks of a proper alliance DC (ie dealing with 100 incoming waves over the course of 8 hours in the middle of the night), let alone crunch incoming. I certainly couldn't deal with that shit anymore.

Ascendancy's core of players wasn't "brilliant" in round 26. It's easy to say that Denial's model was worse than ours, but let's look at what really happened. Both side crashed lots of fleets, so let's work on the assumption that the member base quality was somewhat equivalent, in the sense that it was not the largest deciding factor.

The key difference is that Ascendancy doesn't have a formally organized command structure, relying instead on an IRC bot and an ad hoc command structure. Denial's formalized command structure made some really stupid decisions that (imho) were a larger deciding factor in the outcome of the round than the general quality of the memberbase. For more info, see JBG's end of round summary.

How does Ascendancy's model protect against this sort of decision? Ascendancy relies on a trust network rather than command hierarchy. Say a new member comes around and has an idea for an attack, people don't really know him, but ok they'll give it a shot. If he bombs, they'll be more reluctant to join his attacks, if the attack is a success, he'll have his own personal army at his beck and call whenever he needs it. In this sense we're protected against retards wasting our fleet slots day in and day out.

In the Ascendancy world, Exit simply means you don't do what that person says. In the Denial world, Exit means leaving the alliance. The cost of Denial's command being wrong is (potentially) much higher than Ascendancy's. In Denial's world, Voice means complaining to the HC, who are the final point of responsibility. If action isn't taken, members become frustrated. In Ascendancy Voice means complaining to trusted members (preferably sponsors of whoever is being a dick). This encourages using Voice to settle conflicts, making sure both sides are heard.

This isn't to say Ascendancy's model is in itself superior, it relies fully on the people who use it. Achi, mz and myself (assuming they agree with me) just happen to believe that our model is more likely to stimulate people to play and do well.

The Ascendancy model has a better success rate then the hierarchy model (2 wins out of 11 alliance-rounds played (18%), as opposed to 9 wins out of about 66 alliance-rounds played (14%, assuming 6 competing hierarchical alliances per round)). But that isn't why it's better. It's better because our player retainment rate is much higher than the hierarchical model. More people stick around and more people come back to play again with Ascendancy. I don't have specific numbers, but I can make some rough numbers for the last 7-8 rounds if any alliance cares to compare. (I could very well be wrong about this!)

I also don't think that Ascendancy plays the Ascendancy model to the best way possible. We can improve on several areas if we truly want to achieve the highest possible rank every round. A lot of our members don't want to. They like playing PA, but they don't like playing PA at a highly competitive level. That's fine. At least it's not an unpaid job.

Last edited by Banned; 13 Jun 2008 at 09:09.
Banned is offline   Reply With Quote