I'll start another thread on the current espionage system here, in the hope that there will be a couple of changes in TotA 2.0. For the developers: Please don't take this the wrong way. Overall, TotA is a terrific game, only the espionage system is (IMHO) still one of its weakest points. I'll try to be constructive in all criticism. That said, now my issues with its current status:
1. You know immediately where a spy is the moment he appears. Sorry, not a good idea in my opinion. Intuitively, once you know where a spy is, he should already be dead or captured. And there is just no way a spy team could pull off what they can in GC2, to bring an economic activity on its target planet to a complete standstill for an extended period of time. A spy should be a hit-and-run enterprise, disrupting operations once and making a quick getaway before getting caught.
2. Spies are neutralized on a 1:1 basis. So, a strong civilization can swamp a weak one with spies, and the victim has no chance to defend itself effectively. I think espionage would be a good area for an equalizing influence advantaging the weak and threatening the strong civilizations. In its current form, the exact opposite is true.
3. Currently, a spy has to be a saboteur, which strikes me as distinctly odd. A spy should be able to lay low, unrecognized, until he can make off with some useful information.
4. Espionage status on an empire never decreases even after withdrawing all spies. How can you maintain accurate information without continuing to pay for it is beyond me. I think the infiltration level should drop again over time when it is not maintanined by spies. Of course, you shouldn't lose information you already obtained, so for instance if you lose the information level for some statistic, the game should tell you (maybe by graying out the statistic) that it used to be such-and-such the last time we heard about it.
I know I may be committing sacrilege here by stating this, but I think the espionage system is the one area where MOO3 was definitely superior to GC2. For those of you who don't know how it was done in Master of Orion 3, let me bottom line it:
Spies in MOO3 were categorized at recruitment into economic, technological, political, diplomatic and military. Each had a different focus of work: Economic spies sabotaged production, technological is obvious, political spies assassinated leaders, diplomatic undermined relations between AI's and military spies damaged enemy ships. Each spy had characteristics like cloak (ability to avoid detection), dagger (ability to successfully complete missions), luck (the one thing he shouldn't run out of, but always does eventually), and loyalty. You could insert spies, but that always took some time and was not always successful, and you also couldn't simply withdraw spies instantly.
Each time a MOO3 spy spent in enemy territory, a random check was made against his discovery. Each time he attempted a mission, a more difficult check was made (I think at least it was more difficult). Now, how high the chances were to be discovered each turn depended on a statistic MOO3 called oppressometer (poetic, I always liked that name). The oppressometer was a number for all the injustices and inconveniences a government inflicted on its citizens in order to provide "security", and it had the logical tradeoff that it made the people unhappy.
The espionage system I wish for in GalCiv is one similar to the the MOO3 one. Basically, each planet should have a security (oppressometer) statistic that can be raised by counter esp centre and secret police (there would have to be a security building of the more-than-one-per-planet variety after this). I think buildings like stock exchanges and embassies should lower security somewhat. This security rating should determine the detection risk per turn for each enemy spy on the planet (maybe something like 5% for very high security, and around 1% for very low). There should also be a significantly higher risk of getting caught whenever the spy actually does something to harm the target empire, also based on the oppressometer.
There should also be an empire-wide oppressometer contribution, determined by the government (best in dictatorship, worst in federation) and a racial modifier (best for completely "alien" races like Yor and Thalan, worst for Korx and Krynn). Also, everything that raises security should have a tradeoff in at least one of morale, economy and influence. Additionally, getting in and out of a target destination should no longer be automatic, but take time based on travel and trade restrictions. I think the "getting there" should be easier the more trade is between the two relevant empires and the more tourism is going on. Basically, you should be able to make it rather hard for spies to get into your empire, as long as you kiss trade and tourism income goodbye.
Spies should also be differentiated by activity. For Galciv, I think a useful differentiation would be technological (chance to steal techs, chance to reset current research project for the target civ), military (chance to destroy the current starport construction project on a planet, chance to damage/destroy enemy ships and chance to gain sensor vision of select enemy starships), diplomatic (small chance to wreck alliances and incite wars between AI's, chance to lower inter-AI relations), and economic (constant money drain on target planet, chance to disrupt social construction project). All of those kinds of spies should accumulate general information on the target empire to raise the infiltration level. As an afterthought, spies should only be able to examine the building structure of foreign planets once they are actually there. Being able to view a planet just by inserting a spy there seems kind of cheesy to me.
This approach would solve all of the above issues:
1. You would no longer know where exactly a spy is. Once he's made, he's gone, it just takes time. It would mean a little more mystery for the game, that you don't know exactly how much you are spied on at any given moment.
2. Under the system I described, the cost for maintaining a high oppressometer is relative to empire size, not absolute. Spy swarming would no longer be a winning tactic for the big empires, and it would no longer be a useful tactic to hoard spies until finally "invading" the target empire by the dozens. Additionally, if the chance for a tech steal to occur is made proportional to the number of techs the target empire has and the spying empire doesn't, it would be an additional equalizing influence.
3. If there are different activities for spies, it would not only solve this issue, but also make the espionage system a lot more interesting as a whole.
4. Right now, each turn a spy stays inserted, he accumulates "intelligence points", and once those points reach certain thresholds, you reach a higher infiltration level. That is basically a good idea IMHO, but it would be even better if those intelligence points degrade over time (either they vanish after a fixed time like a year, or they decay exponentially over time). Either way, you could lose infiltration levels as well as gain them that way.