Faction ranking system

[ete]

While I am working on a fully automatic awesome everything included faction ranking formula, it's huge and complex because there are so many variables and things for them to interact with, so it's not going to be ready for the primetime fast enough. I need a way to roughly sort factions into tiers of power for the first faction pack release, and that's going to be by hand for now.

Here are my basic thoughts:
Have a range of power which is considered balanced which covers all official factions, even aliens. Let's call this range 50-60.
That range is based on the average of two qualities, build and rush. Build is how well a faction will do given lots of time to expand, tech, build. Rush is how well a faction will do if placed in early conflict. There are many complexities and edge cases, but these two should cover most everything.
Morgan would have very high Build (maybe 80-90) and very low Rush (maybe 20-30), Mirriam would have the opposite, and someone like Lal or Roze would have fairly balanced scores for both. 40 would be defined as no negative or positive effect on that area.
Taking the averages, we sort the factions into rough tiers of power. Each 10 or 20 points above the max official faction power is represented by a + (and mirrored with -es for below official power).

For this to work, we'd need to go through a large number of existing factions (at least all considered for the quality pack). I would likely do this initially and propose ratings, then ask for feedback. The community would suggest changes and give reasoning, and I would adjust ratings. Does this sound like a good workflow and would you be up for participating?

Also, using the data gathered from this I would be able to refine my model of automatic faction rating.

[JarlWolf]

I think a rating out of ten would be far simpler and easier to determine percentage.

I also think that you should give it three tiers: tech, build and fight. Why? Factions like the Drones are GREAT builders: but they aren't fairly good at teching and their early game rushing strategies are very unstable: you could either overpower the enemy or be easily killed if the opponent gets just a little bit above in weapons technology then you do.

It will also segregate the factions more into more clearly defined lines.

Give each category a score out of ten, 1 being the worst possible score and 10 being the highest.

[Sigma]

This is absolutely necessary for this community if we want to accurately stratify the custom factions that we've created and collected.

JarlWolf, the issue about adding Tech to the scale is that Tech in and of itself only has an effect if a faction puts it to use. It benefits both a Rush and a Build strategy-- if a Rusher can tech up enough early game he can ammend his attack power with superior guns; and of course a Builder isn't able to do much without Tech-- even for Domai, a major part of any good Drone strategy is mitigating his Research penalty instead of trying to circumvent it.

[ete]

I think a rating out of ten would be far simpler and easier to determine percentage.

The score is not meant to be either out of 100 or a %, though I see how that could be assumed. Not sure how best to avoid this confusion.

Give each category a score out of ten, 1 being the worst possible score and 10 being the highest.

I don't want a hard max rating since there is no way we'll ever be using the absolute most powerful faction (best bonuses maxed out), and don't want a hard min rating for similar reasons. Using larger numbers than single digits is mostly to be able to give some more detailed differentiation, I feel that having the building prowess of Morgan or the Drones only a couple of numbers away from Cha or Mirriam would not give.. space.

I also think that you should give it three tiers: tech, build and fight. Why? Factions like the Drones are GREAT builders: but they aren't fairly good at teching and their early game rushing strategies are very unstable: you could either overpower the enemy or be easily killed if the opponent gets just a little bit above in weapons technology then you do.

It will also segregate the factions more into more clearly defined lines.

In the full formula, tech, info war, combat, etc will most definitely be included. I'm not certain what tech brings to the table of "how this faction performs at different parts of the game" that can't be covered more directly by "how good is this faction with early conflict" + "how good is this faction if mostly left alone", perhaps not enough to justify adding another category to rate by? Maybe Sigma put it better.

This is absolutely necessary for this community if we want to accurately stratify the custom factions that we've created and collected.

So you're up for helping?

[JarlWolf]

For tech, the AI that can research better and is more advanced can often reach projects, better weapons, etc. I think it should have its own rating, or should be detailed well specifically in the respective categories.

[ete]

AI also has less drone problems, better industry, better growth (i think?), and better loads of other things. Mostly, I see tech as something which effects and is included in both Rush and Build rating, and adding it separately would confuse things by forcing separation (e.g. Uni is a great builder and fairly strong as a Rusher (fast impact rovers), but if tech was separate both its Build and Rush ratings would have to be low in order to avoid the tech advantage being counted twice). Teching power should definitely be taken into account, but imo it should be taken into account using the two proposed ratings.

[JarlWolf]

Then I'd suggest referring to the latter part of my suggestion. It's just so we can differentiate the more "brute force and numbers" rushers from the "better technology" ones, etc.
I can help test certain factions if need be, its just I need to mop up some projects.

[Sigma]

Sort of like Sophisticated Rushers (Santiago) vs Horde Rushers (Miriam). Then again I don't think it's that necessary to begin with-- if you introduce too much granularity then it becomes too difficult to create a single meaningful metric, which is the goal for this.

[ete]

Yea, as Sigma says, the aim of this is to be as basic and simple a metric as possible. The full faction ranking formula will have a whole lot of detail, and using it you'll be able to read out the type of rusher or builder, but in this case.. if it was possible to just have one figure straight off, I would do that. It's split into two because those are the most basic understandable building blocks: can this faction fight early, can this faction fight later. Teching is more advanced and detailed than is needed for a straight "how good is this faction", which is what I'm looking for here, not "what kind of faction is this".

[Sigma]

Ultimately the goal should be to have, what, three numbers? Total Faction Strength, Total Build Strength, Total Rush Strength. You look at the first number to determine whether you want this faction in your game based on its balance with other factions, and the second two numbers to determine whether it fits the playstyle you want.

[Buster's Uncle]

You know, a faction's strengths and weaknesses interact with a multitude of factors in such complex ways that rating tend to be reduced to a matter of opinion, somewhat.  Which is not to say that we shouldn't try to rate them, but it IS a tricky thing...

[ete]

Right now, I want just one number. Total faction power, used to tier factions approximately. Approaching that directly is basically impossible (last time I tried asking people to order factions just by power, MPers just ranked them by building power not rush at all, and other rankings were fairly inconsistent), so it's split into the two most clear categories, early and late strength.

@BU: Yea, you should see the size of my (far from complete) faction ranking formula, which tries to just flat out simulate all those interactions on a basic level by being huge and including everything. Even when complete it will not be perfect, but it should be a reasonable approximation for all vaguely normal cases. Doing that properly is a big project. This project will be somewhat opinion based for now.

[Buster's Uncle]

Yes.

I just wanted to point that out.  It is still very much a Thing Worth Doing.

[Kilkakon]

Yeah tiers are the way to go, whether we go Smogon style with Uber, etc., or something like that. Vanilla, Underpowered, Overpowered, or something.

[Nexii]

I think one number is more accurate than rating 'rush' and 'build' equally.  The game favors building over heavy warring generally. 

I would rank factions as follows:
70 - Poor
85 - Below Average
100 - Average
115 - Above Average
130 - Excellent

The total faction power would always be 1400 (i.e. 100 * 14) - a zero sum system.   

Voting up a faction would give them +13 "points" and all other factions -1 "points".  Likewise voting down a faction would give them -13 "points" and all other factions +1 "points".

Then you can take weighted points as the basis for a consensus.