As for letting attrition work for you, that's only in a defensive war. I'm talking more about an offensive war where you are trying to capture enemy territory.
No it isn't. Let's return to the Poland/Russia example. I declare war on Poland in late summer and Poland eventually moves their stack onto a Russian province and begins their seige ... just in time for winter. Now the month before spring is due I time my troops to arrive for an attack on the 29th or 30th. Attrition here easily
halves Polish troop totals, before serious battle ensues.
But what if the AI lands his forces on a province which actually can handle the supply? Well the AI rarely has more than two seiges going at a time, so I split my troops up (and it helps to have pure cavalry here) and seige 3 or 4 provinces to his one. Let's say the AI takes and actually completes his seige, odds are I've finished one of mine, possibly two, and the third is fairly far along. So either the AI will take and chase my armies (which means he isn't seiging me and against an opponent of considerable size means I can lead his army far away from another seige, possibly making an amphibious 'escape' or marching into third party land) or seiging a province at which point I keep up the 3 : 1 ratio. Of course on top of this I'm normally slaughtering his new recruits for cheap WS.
In the end the AI can only hope to be competitive in WS if it seiges far more valuable provinces ... in which case I can merely seige back control of my own territory. With large states getting 99% WS even when the AI holds some territory is quite doable. With colonial empires it becomes trivial.
If I'm small, OTOH, I'm always on the look-out for hostile armies, and try to be more "active" in my manouvring, since I can't afford to have to many (or any...) provinces captured.
If I'm a small state I can't afford to have my army destroyed. Even being forced to take a loan is cheaper than having to replace 30,000 cavalry.
In my experience usually both battles and sieges are needed, in some combination, in order to win a decisive victory. But the exact breakdown depends so heavily on the particular situation that I wouldn't even try to give a general rule.
Utterly false. As Nubia, as well as Tibet and several other minors, I've won decisive victories retreating from every battle on the first day. I'd have small forces run through unfortified territory, and a handful of seige armies to gain control of the real war targets I wanted from Spain/Portugal. Seizing every Iberian colony in South America, Africa, and North America is very good for WS (unfortified CoT are gimmes).
Seiges are required to gain anything except cores, battles are only dictated by tactical requirements. One can get 50%+ gains with no battles; indeed in one war I lost more troops fighting Spanish rebels to preserve the government than running away from the Spanish. Had I let the government fall I could have taken over 100% gains. Likewise with a pagan or one province minor you need only slip in behind his exiting army and take control of the province(s), thas was particularly true if the minor was already in another war. I'm sure it just slipped your mind about how easy it is to steal seiges and use othe AI nations as troop sinks and shields.
