Lost Cities - Rankings & Play

[ericdhepburn]

We all know that one of the great things about Lost Cities is its complexity - much of which is a function of luck and randomness - some of which can be compensated for by good strategy, and some of which can't.

Due to the significant luck factor in the game, I think that the current system of rankings doesn't work very well.

For example, in a survey of the current leader board for the game - everyone who has played a 'critical mass' of games (i.e. ~100 or more) has a win percentage between 51% and 60% - there are a few outliers who have win percentages in the 70's and 80's, but their sample sizes are too small to be statistically significant (usually about 30 games). What this means, is that the leader board is more an expression of a player's ability to meta-game the score keeping system, than it is a reflection of the player's skill at Lost Cities.

Proposal:
1. When the game shifts from 'beta' to 'production' - reset the rankings
2. for 'single games' have the maximum points gain/lost be around 5 points
3. Offer a best of 3, 5, or 7 as a formal option when you setup the game, then have a weighted point shift based on the length of the series and the number of wins (2-0 or 2-1 in a 3 game series, 3-0/3-1/3-2 in a 5 game series, etc.) Best of 3 should cap around 15 point shift, 5 around 25, 7 around 35.

Tracking/Statistics Requests:
1. It would also be really cool to track your own progression in the game - i.e. win % over time periods or # of games.
2. Since you have the data, it would also be really cool to see a metric for 'draw strength' i.e. cumulative value of cards drawn (I know I feel strong when I win a game and my opponent had all 5 10's... it would be cool to get a statistical look at that aspect of the game)

The great thing about lost cities is that good players can run a best of 7 - even if it goes to a 7th game, in well under 30 minutes.
In my RL experience, best of 5 or best of 7 is the way I've seen serious players engage each other.

Thanks for your consideration,
Eric D Hepburn
ericdhepburn

[ericdhepburn]

Really Great Implementation BTW!!

[Oso]

First, BGA never did a rank reset, so, they never will do it -I think-

Second, the real game has a score set -like get 200 points for win-. Someday BGA will add this option...

[ericdhepburn]

Yeah, I hear you on the rankings reset - probably not likely - would cause too much anger from the meta-gaming crowd.

I think the best-of-X (where X is odd) levels the playing field to evaluate player strength better than the 'race to a set score', although that certainly has its merits.

thanks for the feedback.
Eric

[Oso]

BGA tries to build their games exactly than the original game.

I said, the real game has the "score option", where you get victory if you reach X score... -100, 150, 200, etc-

[ericdhepburn]

Right, and the superimposition of a ranking system is an inherently meta-gameable extension of every game on BGA.
So the ranking system, its development, evolution, and direction are outside the original scope of the real game.
It is an ongoing-tournament or never-ending-tournament design problem which necessitates a thoughtful adaptation and approach to each game to determine whether and how the ranking system is or isn't optimized for that game.
It isn't a question of fidelity to the original game, it is a question of how the meta-game of a ranking system can best be implemented for each game so that it resonates with the nature of that game - otherwise it is just a point-gathering overlay, not dis-interesting, but not nearly as interesting as allowing the ranking systems for different games to evolve according to the nature of the game (I think Hanabi is another excellent example of a game where you can't just overlay a point-gathering system and hope to get a value-added meta-game ecology established).

Eric