Convincing BLubmuz about Espionage, with a few cool "Top Espionage Tricks" buried in my paragraphs of text 
Sorry to continue, but i still do NOT understand, despite Dhoom's detailed post.
Please correct where it's wrong:
we meet the first AI1 and we assign 1 EP to him.
Essentially. We assign a WEIGHT of 1 to the first AI that we meet, not 1 Espionage Point to that AI. The way that Weights work is that they are relative to each other.
If I assign Weights as follows:
AI1 = Weight 2
AI2 = Weight 4
AI3 = Weight 0
AI4 = Weight 2
Then our Palace's Espionage Points will be distributed as follows, each turn:
AI1 = 1 Espionage Points added to the total accumulated against AI1
AI2 = 2 Espionage Points added to the total accumulated against AI2
AI3 = 0 Espionage Points added to the total accumulated against AI3
AI4 = 1 Espionage Points added to the total accumulated against AI4
doing so or do nothing changes nothing, since he will do the same until he meets another AI.
Correct. If we are the first competitor met by an AI, we will have all of their Espionage Points assigned to us initially. This behaviour will occur whether it is the first AI that we meet or the last AI that we meet.
However, consider that we are not making it a priority to meet the other AIs, but instead are making it a priority to find nearby locations to settle. Therefore, most AIs will meet another AI before they meet us, meaning that they will have already started assigning Espionage Points to a different AI instead of to us initially.
Then we meet a AI2.
We do NOT assign EP to him, so all our EPs go to AI1.
Correct.
a) we're the first opponent this AI will meet, he automatically assigns to us all his EPs.
If that is true. I'm under the belief that this situation won't be the case for most of the AIs, due to our "don't find them, let them find us" current Foreign Policy.
b) If and only IF he already met AI3, we can avoid to have AI2 assigninig the EPs to us.
We can't 100% avoid it, even if he has already met AI3. Shaka proved this fact, since he was Pleased with the other AI that he had met and thus wasn't assigning Espionage Points to that other AI.
Pros i can see:
we accumulate many EPs versus AI1
That Pro is true, but I think that there are a couple more Pros.
I admit that my knowledge comes from playing, not from looking at the SDK (Software Development Kit), but my anecdotal knowledge tells me the following:
2. An AI to whom we assign no Espionage Points tends to assign less Espionage Points to us. Often that can mean 0 Espionage Points assigned to us.
3. The AIs like to "compete with" or "one up" the human player in terms of Espionage Points. If we assign 0 to them, but they still assign some to us after having met at least 1 other AI, they are more likely to assign a smaller percentage to us. On the other hand, if we even assign a small amount over time (1 per AI, for example), they will assign a big percentage to us, just to stay "far ahead" of our Espionage Points' spending.
4. Knowing what tech a single AI is researching at all times is far more valuable information that knowing the Demographics of all of the AIs. Following the AI tech rate, you can often predict what similarly-sized AIs will be researching. You can also piggyback techs properly--for example: "oh, they are learning The Wheel on turn 75--if we wait until they are done, researching it ourselves will be cheaper." You can also plan tech trades a lot more appropriately, for example "oh, that AI is researching Aesthetics, let's go and grab Math/Iron Working/Currency instead, since one other AI knows Aesthetics and once they both know it, we can get one of them to trade it to us."
Cons i can see:
we know nothing about AI2.
We will still know a lot.
We will know how many cities they have once we can trade with them.
We will know when they learn Bronze Working and Monotheism, since AIs will almost always switch to the corresponding Civics IMMEDIATELY AFTER LEARNING THE CORRESPONDING TECH if they learn those techs early on.
If they learn Writing first, we will know about it (Open Borders agreement becomes available). Similarly, if they learn Alphabet first, we will know about it.
We can know who they like and dislike from the F4 Glance screen. We can also use this screen (or the act of repeatedly talking to them and then not making a trade) to find out who their Worst Enemy is.
We can find out what their best military unit is by repeatedly talking to them and then not making a trade.
We will know what Wonders they have built (even National Wonders with BUFFY).
This can be worst the later we meet AI2.
To be honest, even if the AIs did not behave the way that I have observed when you assign 0 Espionage Points to them, it takes a REALLY LONG NUMBER OF TURNS for an increment of 1 Espionage Point per turn to even give you Demographics info. In my example screenshot where 4 different AIs were receiving 1 Espionage Point per turn from us, only 2 of those AIs had their Demographics Info revealed--the other two did not. Over time, as those other 2 keep assigning Espionage Points to us, the threshold that we'll need to surpass in order to see their Demographics Info (as well as for any other Passive or non-Passive Espionage Info-gathering or Espionage Mission) WILL GO UP! So, it might take a great number of turns before we can reach that threshold and we might never reach it if said AI assigns 4 Espionage Points to us per turn.
Now, we have enough EPs versus AI1, but we need to at least see the Demographics of AI2.
Why do we NEED to see the Demographics Info? What info from the Demographics screen are you after? Can't getting this info wait until later in the game? It seems to me that by your strategy, this info WILL NOT COME SOON, ANYWAY!
Why do I make that claim?
Well, for starters, with only 4 Espionage Points per turn and no fractions being allowed to be assigned, the maximum number of AIs for whom you could have Demographic Info are 4 AIs.
Further, it will take you a VERY LONG TIME for enough Espionage Points to accumulate against all 4 of them in order for you to see all of their Demographic Info.
For a good part of the early game, the most that you'll see is the Demographics Info for 2 AIs. What about the Demographics Info for the other AIs? How useful is it really to have that small bit of info for one more AI, compared to the Pros of my approach?
It seems obvious to assign all our EPs to AI2 to recover, then start balance between the 2 AIs depending on needs.
If you assign Espionage Points to an AI at a small rate, they will assign Espionage Points to us. If you assign none to them, many will assign none to us. Later, once we've got about 16+ Espionage Points to assign per turn, we can switch our target. We'll "catch them with their pants down," getting a big leap up on relative Espionage Points out of the gate.
Please bear in mind that i use only passive espionage, i barely build 1 spy or 2 in all my games just to help avoid sabotages.
There is a reason for that: your Espionage Points allocation is hurting you. What matters the MOST in Espionage Costs is your relative Espionage Points accumulated against an AI compared to the relative Espionage Points that they have accumulated against you.
What matters for Spy Missions a bit are other factors, such as:
- How many Espionage Points TOTAL, AT ANY TIME, EVEN ACCUMULATED BEFORE MEETING A SINGLE OTHER PLAYER, each of you and the AI in question have both generated. This "hidden total Espionage Points earned" even applies to Espionage Points that were spent.
- Costs can be reduced if the city you perform the mission on has our State Religion.
- Costs are influenced (reduced or increased) by distance to our Palace.
Advanced Player Tactic: Build your Forbidden Palace (which costs more than your Palace) near your Palace's original location. Build your Palace again (a very cheap building) close to the AI that you want to run Spy Missions against. Then Spy Missions (such as City Revolt, used to capture cities) become cheaper. Plus, you'll have a nearby Palace to reduce the Maintenance of captured cities.
- Costs can be reduced if you have a Trade Route with that city (if you have Open Borders, there is a good chance that you will have such a Trade Route).
- Possibly other factors that I am forgetting.
But, out of all of those, the biggest influence on Espionage Cost is actually your Espionage Points spent on that AI vs that AI's Espionage Points spent on you.
If you had a way of reducing the Espionage Points spent on you, you'd have an amazing tactic for being able to run several mid-game successful Espionage Missions, instead of just being able to afford one or two, due to the AIs having spent many points on you. Such a tactic could let you "save up" our Espionage Spending time for a couple of the AIs until near the end of the game, when we might want to switch their Civics around.
Oh wait, we do have such a tactic! We're discussing it right now!
Now, if you want to steal a tech, what we'll probably do is build a few Courthouses so that we can generate many Espionage Points per turn. Then we'll target an AI who has spent 0 Espionage Points or only a few Espionage Points on us. We'll quickly catapult ahead of them--but you will also find that they will start spending Espionage Points on you, since you started spending on them. That's the nature of the beast--but the trick is to be able to spend more on them than they can spend on you, without letting them accumulate many points on you over time.
The Espionage Points on Zara will probably mostly end up being Passively used, or for limited Spy Missions, such as the odd City Revolt, as he'll spend a lot on us. But for other AIs, we can potentially use this tactic to steal techs, steal their treasury, influence their Civics at a key point in the game, etc.
Now, if I haven't convinced you of the value of "zero to sixty" acceleration on our Espionage Points (come on, a car reference has to work for you, right BLubmuz?), then there is one more very convincing point for doing so in a game where we care about Diplomatic Relations:
Again, this info is from my anecdotal experience, but it would appear that if your Espionage Points against an AI are about 20% or more higher than their Espionage Points on you, when your
Spy is CAUGHT (which can happen with reasonable frequency), there appears to be
no Diplomatic Penalty!!! That point is HUGE!
If your Espionage Points spent on an AI relative to theirs spent on you are close in value or theirs is larger, then when your Spies are caught, there is a (permanent, or at least long enough to appear to be permanent) Diplomatic Penalty of -1!!! That's huge in a Diplo game and even not so great to get in a Cultural game! So, any trick that we can do to avoid getting these negative modifiers, such as the "zero to sixty" approach, should be leveraged to the maximum in this game.
You are right, of course.
