Experiment: single lab vs many labs per planet

Introduction

One of the requested feature was the ability to build multiple labs on a single planet. The original game allowed only one per kind. Update 0.95.160 lets the player lift this restriction in a skirmish play. I wasn't sure if it would work from gameplay and AI perspective, therefore, I run a simple test.

Experiment

I started a normal skirmish with a single human player, set the objective to technology victory and turned on AI assist in-game.

Limited labs - AI

This was the classical case with 1 lab per planet. The AI spent about 1.5 years till 10th February 3429 with a total of 26 planets and labs and $66.56M money spent.

Unlimited labs - AI

The AI successfully built multiple labs on the same planet, but usually never got above 2 per planet. This is due the fact that AI first tries to fill in empty planets first, then tries to swap out existing labs based on the target research requirements. This swap is generally advisable as selling an old lab yields some money which helps in other tasks. The second thing is that AI colonizes planets from time to time so there is always some empty planets where it can build new labs.

The game run from 13th August 3427 to 27th August 3428, about 1 in-game year which is faster than the limited lab case. The AI had 14 planets colonized and 26 labs in total. $60.7M spent.

Allowing multiple labs per planet reduced the time (and cost) of colonizing and building up new planets to support a new lab (usually a 250k investment before the lab is built).

Unlimited labs - me

Again, the main difference between my gameplay and the AI gameplay was that I didn't bother building fleets and defenses but only concentrated on economy and research.

Finished on 7th April 3428, which is about 9 months. Using only 7 planets and a daily $70k income at the end, I spent $6.2M for the planets and research together.

I truly felt here the lack of two features already asked among the issues: use spare lab capacity for something (speed up current or just allow another research to progress as well) and queueing up research. The former speeds up research while the latter reduces number of clicks in the UI. I don't yet know how to deal with these things yet (due to UI and development time limits).

Conclusion

Multiple labs per planet seems to be beneficial to both human and AI players; reduces the times and costs to reach a certain technology level. Note that this might fluctuate a bit depending on the starting planet's capabilities. Still, a human player with experience may finish just under 7-9 months when the chance of battle is zero. 

I'm still thinking about whether to lift the restriction in the main campaign as well. I think a human player would gain unwanted benefits in the scripted missions (especially against the Garthogs) and would not facilitate expansion (and thus, higher tax income later). 

However, as the experiment above shows, AIs tend to work better with the settings and perhaps allowing them to build multiple labs (on Hard) while the player can't could be a good way to harden the game. Unfortunately, campaign non-human races don't really research so there isn't any benefit for them, but use it in skirmish against another human race is a possibility.