Espionage vs. Siege

[iamnleth]

In a recent game I was playing, my war against Kublai Khan was basically foiled because I decided to use the spy's "Support City Revolt" mission rather than siege to keep up with my 2-move cuirassiers. In the second turn of the war, I had a stack of three spies in the city of Beshbalik (previously four, but one was caught) and my stack of about 20 cuirassiers outside the city. One of the spies completed the mission successfully and I take the city.

Following that, I moved toward Karakorum. Neither of my remaining two spies were caught this time around after spending a turn in the city, but both of them failed the mission to throw the city into revolt. I was left with no way to take the city because I didn't have siege weapons. I continued the war, waiting for more spies to arrive, but by the time they arrived, I was at tech parity with Kublai and the war was now in stagnation. I also noticed that those spies were only successful around 50% of the time, maybe even less.

Now, I think I know the reason for this. Please correct me if I'm wrong:

The fewer EP points that Kublai is generating, the easier the mission should be for me. He must have been generating a good amount. The "Support City Revolt" mission was costing me around 450 EP points each time I tried. Also, is it made easier if I'm generating more EP points?

I guess my general concern is when to use siege weapons. I used to think that spies were a good use for this, but now I'm not so sure. When should spies be used vs. siege, and is there a chance of success percentage that can be seen before I try this again?

 

[InFlux5]

I've had the most success using a combination of both. I bring fewer siege weapons than I did before espionage was added, but I still bring them, mostly for collateral damage. I think it's important to have a backup, because of scenarios like the one you mention.

If the spies succeed, just use your cats for collateral damage. If not, they can still bombard.

I must say, I really like the new dynamic. There are key choices to be made about how many hammers to invest in spies, and how much gold to invest in espionage. It's made war a bit more interesting.

 

[Joelius Ceasar]

I love the revolt method because I'm very impatient. I wish I could say more but, alas, I am but a noob.

 

[BalbanesBeoulve]

The game does tell you what the % is before you choose the mission. It's also much, much cheaper, 50% cheaper, if you have your spies fortify for 5 turns before pulling off the mission.

I use siege because I use rifles and maces more than mounted units, but if you want to go for a quick mounted attack, you basically just need to plan ahead. Get more spies built than you think you'll need.

 

[iamnleth]

The game does tell you what the % is before you choose the mission. It's also much, much cheaper, 50% cheaper, if you have your spies fortify for 5 turns before pulling off the mission.

I use siege because I use rifles and maces more than mounted units, but if you want to go for a quick mounted attack, you basically just need to plan ahead. Get more spies built than you think you'll need.

Isn't there a good chance that they will be caught if I leave them fortified in a city? Also, how do I increase my chances of success?

 

[BalbanesBeoulve]

Isn't there a good chance that they will be caught if I leave them fortified in a city? Also, how do I increase my chances of success?

Yes, but it's a risk worth taking for half price missions. I don't think it'spossible to increase the success rate, though.

 

[Joelius Ceasar]

How long do you have to wait for half price? can the price go any lower?

 

[iamnleth]

He said fortifying them for 5 turns will minimize the cost.

 

[Ibian]

But really, why bother? You cant do it very often unless you invest commerce or specielists, which means your tech rate goes down. And siege weaps are very useful even against unfortified cities anyway.

The only time i have spare spy points is if i burn a GSpy, but i dont tend to get very many of those if at all.

 

[iamnleth]

But really, why bother?

I mentioned that I used spies to keep up with my cuirassiers, something that no siege weapon until mobile artillery can do. I wanted this to be a quick war.

You must also note that spies are cheaper than siege weapons, especially since the needed number of spies is less than siege weapons. And siege weapons can take multiple turns to bombard a city, particularly when there are walls and castles. Edit: If you're fighting with a tech lead (which is the only way I like to fight my wars), you shouldn't need siege to sacrifice for collateral damage.

One more thing I'd like to add. In this thread, I revealed a strange scenario regarding siege (strange in my experience, at least). I could not discover why my siege weapons were bombarding so poorly against a city with no walls or castles. It remains a mystery to me, and before I do any more major siege warfare, I'd like to figure out what was going on behind the scenes.

 

[UncleJJ]

I replied in that older thread here that in fact you are probably mistaken about the walls and castle (despite what the game seems to display). In fact everything would be explained if walls and a castle were present.

 

[Zanttu]

You need more spies, if you want to use them properly. I usually build 5 spies per city I want to take (1-2 may get caught and 1-2 can fail), and fortify them to the target cities 5 turns before the attack. Spies work great for me, because I favor short wars. I declare war, take the cities I need and make peace. Most of my wars are shorter than 10 turns (or at least the battles don't take longer, some AI's refuse to talk for a long time even if they are losing).

 

[lilnev]

Spies are best during the treb vs. castle phase, when bombarding takes a long time. Before that, the relative cost of spies is high. Once you have cannons, you can blast down the walls in short order (although spies may continue to be more convenient, especially if you want to move fast).

Bring more spies! A rule of thumb is that one third of your spies will get caught (can be more if your target has a security bureau) and success rates are ~60%. So if you bring three spies, there's a decent chance (~16%) that you'll fail. For mission-critical tasks like city revolts, I bring at least five spies. More if I'm going to be rolling through multiple cities.

peace,
lilnev

 

[madscientist]

The spy city revolt works best during the castle era, ever notice how friggan long it takes to where down those castled cities? Spies are the answer there, just bring an excess. Also you should be bringing seige weapons anyway to suicide against the defenders and make life easier for your main host.

Stationing the spies in the city reduces costs but to me speed is more valuable there as your target is going to be producing either alot less espionage or none.

 

[UncleJJ]

I like to use spies to reduce a city to revolt, but like other posters have said it is wise to back up the spies with siege for bombardment. I like to send enough seige to reduce a 100% castle defence in about 3 turns (needs about 140% of bombardment capacity) and then on the fourth turn use collateral damage to soften the defenders. To get 140% bombardment capacity you'd need about 10 seige with each basic catapult giving 8% and trebuchet 16% and the accuracy promotion another 8%. So using a spy to revolt the city can save 3 turns of bombardment against a castle. The nice thing about having both methods and a nice stack of seige is that lesser cities can be reduced in one turn.

I like to have a big enough SoD with enough defenders that after capturing one city I can send the seige ahead of the main assault troops that stay behind and heal with my medic 3 general. Splitting the SoD like this has to be done with care as it invites a counter attack. Alternatively you can move the wounded troops with the SoD and heal the CR2 maces while the seige bombards the next target's defences.

There is a problem with using cuirassiers and cavalry that is hard to match up with using seige so spies become indispensible. That means you really have to make a lot of spies and also spend a lot on espionage. The chances of a spy getting caught depends on how much you've spent on EPs and also how much your opponent has spent, so spending a lot really pays off. Espionage is usually a big part of my game anyway and I like to boost my normally high EP spending with a short burst of high %EP on the slider, sometime 2 or 3 turns at 100% EP just before the war.

Using an attacking force of entirely mounted troops presents problems that I don't like. It is fast but is vulnerable to counter attack and it is difficult to take well defended cities (even when walls or cultural defence is zeroed by a revolt). Usually I will send a fast moving mini SoD (mounted and spies) to pick off a weakly defended city while the main SoD with foot and seige (and some mounted) engages the enemy main army and better defended cities. Attacking mounted troops benefit greatly from seige collateral damage when against tough targets. So seige is needed, so I may as well have a balanced SoD including foot that give good defence and are more cost effective against cities anyway.

 

[Dier]

You have to make the most of a tech edge. I recently cleaned up a map with tanks, because I had them first. I was mass building tanks and spies, nothing else. If my spies failed, I'd attack anyway and lose more tanks but not momentum. Within 20 turns everyone had capitulated.

If you install spy specialists and focus spy points on your target, you'll hardly need to touch the slider. Spies are great, ships aren't half bad, but artillery is just too slow for me.

 

[LlamaCat]

you had a good strategy going there Trv, but the problem sounds like you didn't really invest enough in espionage to support it. you need to check the espionage screen to get the estimate base costs of creating a revolt in your intended targets, it's all right there for you to see - no mystery in what to expect. also, if the city shares your religion and shorter distance to your capital will lower the espionage costs. you really need to send the spies out earlier and let them sit for 4-5 turns first, to increase the odds and of course lower the costs up to 50% maximum (and you don't need open borders to do this all ahead of time before you declare war - big advantage!). But sending more than 1 spy into each city is the right thing to do, in case one gets caught.

and if you are going for this strategy, I would direct 100% of your espionage spending against that one civ leading up to and for the duration of this kind of war.

 

[OTAKUjbski]

The amount of Espionage vs. Siege in my wars reflects the amount of Production vs. Science/Gold cities I have.

If the land grants good to superb production, there's little reason not to leverage it with numerous siege units.

However, if the land is weak to mediocre in production, it makes sense to compensate by leveraging the economy towards .


-- my 2

 

[iamnleth]

I had all of my EP points directed at Kublai (even for about 20 turns prior to the war). I ran about 20% on the EP slider to gain some more EPs, and it came out to around 70 EPs per turn at Kublai; I had roughly 2000 accumulated by the time I was ready to declare.

The amount of Espionage vs. Siege in my wars reflects the amount of Production vs. Science/Gold cities I have.

If the land grants good to superb production, there's little reason not to leverage it with numerous siege units.

However, if the land is weak to mediocre in production, it makes sense to compensate by leveraging the economy towards .


-- my 2

Good point. This was also a factor in my game. I actually used Universal Suffrage to rush-buy those 20 cuirassiers. I really only had 2 production cities, and I figured that producing (or purchasing, in this case) siege would be much more costly.

Off-topic: is rush-buying units a popular choice when little production is available? I figured I might as well take advantage of a philosophical/National Epic capital and produce a few Great Merchants to use seeing as I had just reached democracy.

 

[InvisibleStalke]

Off-topic: is rush-buying units a popular choice when little production is available? I figured I might as well take advantage of a philosophical/National Epic capital and produce a few Great Merchants to use seeing as I had just reached democracy.

Rushbuying an army is definitely a good option. For me its usually either rushbuying ships (if my costal cities are low production) because my next opponent is overseas, or it is when I hit the point in time where I think I have enough technology to take anyone on the map and my best route to take them all out is to maximize production. At that point I drop science to zero and commit totally to military production. Its amazing how fast your military grows.

If I go nuclear for example, its not unusual for the Manhattan project to be built and two turns later have a dozen ICBM's raining down. And 5-6 per turn thereafter. That can spoil an AI's entire afternoon.

If you have just reached Democracy you should combine rushbuying (which you can probably only do from a GM - empires at that point in time don't normally produce enough cash) with drafting - which is incredibly productive and requires only food.