Evaluating CIV Players

[Krill]

Or try to poison us with 8 pints of best bitter. That works too.

Seriously though, it is very complex. the variation on how good a leader is for the map played upon can distort the evaluations. So do you value the ability to choose a good leader/civ combo more highly than someones ability to micromanage?

[Sareln]

When I was originally thinking of the question, I was definitely think about keeping leader/civ invariant. If you put player A in a situation, and then put player B in the same situation, who will come out ahead?

I'm starting to realize how silly a quest this is now...

[darrelljs]

Well that is what the adventures and epics do (for SP at least).

Darrell

[Cervantes]

In multiplayer, the ability to balance economy vs. military seems important. Anybody can slave a big SoD but can he keep his techrate up too? Can he expand while building units, and keep his cities always working good tiles? Then, once he has a stack, what does he do with it? These are all good questions to ask when considering multiplayer games. Skill at single player is pretty much meaningless when it comes to multiplayer games.

I was never able to do this, so I am not very good at multiplayer.

[Sullla]

Some kind of ranking system for MP would more or less go against the core principles of Realms Beyond, dividing up the player base and creating incentive to "climb to the top of the ladder", with all of the negative connotations that that brings. There's a reason why all of the Epics/Adventures are treated as individual events, with no overarching rating system.

Sure, it's fun to speculate and all, but this idea is probably best left as speculation.

[Sareln]

Some kind of ranking system for MP would more or less go against the core principles of Realms Beyond, dividing up the player base and creating incentive to "climb to the top of the ladder", with all of the negative connotations that that brings. There's a reason why all of the Epics/Adventures are treated as individual events, with no overarching rating system.

Sure, it's fun to speculate and all, but this idea is probably best left as speculation.

Yeah, this was speculative on my part, not an attempt at the creation of an actual system! Perhaps I should've made that clear up front. It's just me and my love of statistics I suppose...

[Irgy]

I'd say far and away the best approach is to read their spoiler threads and figure out for yourself their depth of understanding of the game. Helps if you're a top player yourself so that you know the difference.

However, if you want a statistical approach, basing it on results rather than doing direct analysis by hand, then you can use a statistical model. One that would potentially suit is the Bradley-Terry(-Luce) model, with a few modifications. Those being in particular;
1. Some sort of improved prior model to deal with the shortage of data in general and the variation in number of games played in particular.
2. Extending the model to deal with games involving 3+ players, as it is designed purely for pairwise comparisons.

I already know a solution for (1), and I have in mind at least one reasonable way of dealing with (2). So, if you wanted to do the dirty work of gathering all the results you can find into some sort of relatively easily computer-parsed format, I'd be happy to volunteer to do the "hard" part of writing some software to give you an answer.

I think putting all this together will be close to the best you can do from pure results, and it would be interesting to see how well it agreed with people's manual analysis.

[Imhotep]

For RB's PBs/PBEMs, there are a number of skills you need, some more valuable than others, and they all vary in value depending on the exact game in question. These are the big ones IMO:

a) Civ/leader selection, being appropriate for what the mapmaker's going to throw at you. You don't want to be Imhotep (PB1: Shaka Zulu, thinking Impi rush where the tech pace was lightning fast and the map gigantic), or Portugal or Carthage in PB3 (no water).

Obviously anybody can pick along Fin/Exp/Phi/Imp as they are the strongest traits in the game. I wanted to try something different. And it would have been a different story if I had not started next to Agg Rome, and that the game would turn into an alliance-fest with techs being swapped around all the time - well, wasn't all that foreseeable.

On the topic: Worker movement and tile micromanagement is to me the main thing that sets the good players apart from the not so good ones.

[Cyneheard]

Obviously anybody can pick along Fin/Exp/Phi/Imp as they are the strongest traits in the game. I wanted to try something different. And it would have been a different story if I had not started next to Agg Rome, and that the game would turn into an alliance-fest with techs being swapped around all the time - well, wasn't all that foreseeable.

On the topic: Worker movement and tile micromanagement is to me the main thing that sets the good players apart from the not so good ones.

Agreed that worker management is the biggest key, but diplomacy and tactics are both very important, and vary wildly from game to game.

But a player can dig themselves into a deep hole by choosing the wrong civ for a game (and it's not like you had to be the Agg Rome target. There was the WonderSpammer team right there as well), and the alliance breakdown was somewhat foreseeable, especially once the first group formed.

[Ilios]

worker management is the biggest key

+1