More thoughts on the scoring system.
I know that some will disagree with me - please know that I'm just throwing out my ideas, I know that what I say is opinion, not fact.
It is clear that many players feel that the scoring algorithm is flawed. The overall problem seems to be that the score does not measure skill well, especially when comparing different types of games.
So, before considering how to try to measure it, do we need to define skill in this game? I suspect that people with different styles of play will define skill differently, and thus will lean toward different scoring systems to measure what they feel is skill.
To my mind higher skill implies an ability to simultaneously use many of the game's facets (units, improvements, resources, tech, culture, diplomacy, etc, etc.) well, to produce a desired outcome.
Any definition like that would seem to rule out fast conquest. If a game can be won so soon that much of the tech. tree hasn't been researched, then that game did not require skillful use of many of the city improvements, unit types, long-term diplomacy, governments, etc. (That game never got to them at all.) So that game is only measuring skill in a subset of the game's rich features. One might enjoy such games and be extremely skillful in using that subset, but it still does not seem as great a test of skill as a test which requires skillful use of a greater range of the game.
So it makes sense to me to eliminate the smaller map sizes. In a general sense, to eliminate games which don't require an application of skill throughout the longer timeline available in the game. (So far, that seems to apply only to small maps, and/or maps with few opponents, and perhaps maps at lower difficulty levels.)
Continuing in the same vein is the question: Does an earlier victory for a given victory condition imply greater skill? Suppose one player wins a game with conquest in 1800AD, with a scattered Civ with many small cities which have been heavily pop rushed, and with little culture. A second player wins the same game in 1850AD with a large happy populace and high culture. Which of these took more skill? A very tough question.
Before rushing to any scoring modification which strongly favors earlier wins I think the previous question should be considered at length. In my opinion the second case in that example probably took more skill. It required a more delicate balancing of more of the game's factors and was a harder goal to acheive. (If indeed one considers it a desirable goal, which is of course part of the problem in this discussion, defining the goals by which skill will be measured.)
If one agrees that the second case took as much skill, the big problem is how can it be measured? I don't have a solution to that. But, I do want to suggest that perhaps the existing scoring system isn't all that bad for the moment.
The two games I described earlier (1800 vs 1850) will be closer in final score after milking than they were in conquest date - the first case has more time available to milk, but is behind in averaged-over-the-years score at that date. The second case has a real chance to win in final score. So, to my mind, the existing scoring system, with its favoring of milking games, does after all have some inherent goodness. Earlier control of the world is favored in one way (more time to milk) but more early building can offset that and might produce a higher score. A subtle tradeoff which at least to some degree can measure skill.
I think that the best tests of skill will be games which can't be won by skilled players until late in in the Industrial Age or later. Given a game which is a challenge up to that point, I don't personally have a problem with milking it. By the time it is "in hand", the game probably has 150 or less turns left to 2050. I imagine that many people invest 40 hours or more in getting a tough game to that point. Spending another 10 or so milking it is not that big a deal to me.
And there is (in my opinion, I know some strongly disagree) some skill and interest in milking efficiently. You need to consider it even before the "in hand" date, anticipating that your average per turn score over the long haul will matter. You need to trade off more towns with faster growth and less unhappiness vs. maximum size towns which use up tiles themselves. You need to optimize build paths vs money available, spending on happiness now vs. happiness later. And so on. It isn't as complex as war and it isn't random, but it isn't trivial either.
One possible cure for the scoring problem is to emphasize date but separately rank the earliest win of each type (conquest, spacerace, etc.) Theoretically that would let us emphasize and work on the different areas of skill in the game. But I'm not convinced that it would actually help. It has a few problems. 1) How do we compare the different types? I think that people inevitably will want to compare them, they'll want THE high score however that is defined. 2) How to balance the types? That would take some careful thinking and might end up imbalanced anyway. 3) Once balanced, would people like the result? I suspect that most high level players LIKE the warfare aspect. And warfare is such a complex part of the game that any game which largely excludes it probably doesn't require as much skill. I'd be rather unhappy if I played a dynamite game, putting 100 hours into a large map with 16 civs and starting on tundra, finally beating down every other Civ by 2000AD, and then someone beat me by playing an easy "builder" type game from the same start to a 1500AD diplomatic victory. I would strongly feel that beating all other Civs into submission had required more skill.
I suspect that any good scoring must most reward a game which includes a significant element of war and conquest. If a scoring system doesn't require that for THE high score, then many existing players (including me) would feel that it must be a flawed scoring system in some way and we'd be back to discussing what's wrong with the scoring system.
In summary: From my perspective the existing scoring system is, although flawed, not all that bad for measuring skill. I would not want to see it replaced with something which might be worse by favoring early conquest even more than it already is - I think that might only create an increased and undesirable focus on skillful blood-letting (and pop rushing) at the cost of other game factors.
