I have an idea to reduce tech snowballing, to prevent smaller empires from falling too far behind in the tech race, and to benefit civilizations that choose a 'vertical' strategy:
==> Divide a civilization's research points by the square root of the number of planets in its empire
Here is a simplified example:
Assume there are three major civilizations left in the galaxy: Civ A with 25 planets, Civ B with 16 planets, and Civ C with 9 planets
Assume each planet is identical, and that each planet produces 100 research points per turn
Under these assumptions, research points per turn would be Civ A 2,500, Civ B 1,600, and Civ C 900.
Let’s look at how tech research works today. Say all three races start researching the same tech at the same time, and that the tech costs 10,000 research points
Civ A will discover that tech in four turns ( 10,000 (/) 2,500 = 4.0 )
Civ B will discover that tech on the seventh turn ( 10,000 (/) 1,600 = 6.2 )
Civ C will discover that tech on the twelfth turn ( 10,000 (/) 900 = 11.1 )
The larger civilization will quickly pull ahead in the tech race and 'snowball'.
Now let's see what would happen if we implement my suggestion, namely to divide each civilization’s research points per turn by the square root of the number of planets in its empire....
Civ A, with 25 planets, would have its research points per turn divided by 5
Civ B, with 16 planets, would have its research points per turn divided by 4
Civ C, with 9 planets, would have its research points per turn divided by 3
Say all three civs start researching the same tech at the same time, and that tech costs 2,000 research points **
Civ A will discover that tech in four turns (2,000 tech cost (/) 2,500 research points per turn (/) 5 square root of twentyfive = 4.0 )
Civ B will discover that tech in five turns (2,000 tech cost (/) 1,600 research points per turn (/) 4 square root of sixteen = 5.0 )
Civ C will discover that tech on the seventh turn (2,000 tech cost (/) 900 research points per turn (/) 3 square root of nine = 6.7 )
Given this approach, the smaller civs won't fall as far behind so quickly, limiting tech snowballing. Large empires will no longer so quickly and so decisively dominate smaller empires.
Because the calculation uses the square root function, adding a new planet to an empire will always speed a civilization's total research, provided that the new planet's research points per turn equals or exceeds the average research production of the other planets in the empire. On the other hand, adding a new planet which spins off only a few research points might actually delay your civilization’s research slightly. This is great, because it forces the Player to make strategic decisions. Maybe colonizing a size 4 desolate planet before you have discovered all the terraforming techs isn't such a good idea. Perhaps it would be better to build out existing planets, and cover them with starbases, rather than pursuing an ICS strategy and colonizing every planet in sight as quickly as possible. (By ICS I mean infinite city sprawl, a term people used long ago for Civilization 1, 2, and 3. Maybe there's a different term for space games?)
** I chose a tech research cost of 10,000 points for the current approach example and 2,000 for my squre root example because in both cases Civ A will discover the tech in four turns. Obviously the tech costs in the GalCiv 3 tech tree would have to be rebalanced if my square root suggestion is implemented.