Dominant strategy in ATW: “Sniping groups of armies”

[Jurand of Cracow]

After having played a dozen games with ATW-Great Campaign and ATW-AoC (and earlier in R2TW) I’ve come up with the dominant strategy which makes the game really boring (not to mention un-historical).

Features of R2TW and ATW that enable this strategy:
· Limited number of armies per faction (depends on imperium level) – the number of armies a faction fights against depends on the number of factions it is at war with but not their strengths (ie one faction with 20+ provinces is much weaker than 3 factions with 3 provinces each).
· No manpower limits (like in M2TW units’ refill rates, DeI mod for R2TW alleviates the problems) – if your army is destroyed, you just set up a new one: the speed depends on the recruitment slots, which are usually not a serious barrier.
· Cheap units in comparison to the economy: most of the armies are 20-units armies. This is better in the AoC DLC, but still, the result is the same.
· Automatic refill of the battle loses (if not enemy territory) – if you inflict losses on an enemy, he will come again in full numbers soon.
· High movement ranges: long moves per turn possible what enables “catching” of enemy armies (I mean attacking twice: one attack may be avoided, but the second has to be fought to death).
· Close-to-perfect knowledge of the enemy: where it’s positioned, how strong it is (only sometimes restricted by armies hiding or by having not enough agents) – also enable catching of the enemy armies.
· Auto-resolve skewed to benefit the numerical superiority: even a small advantage in numbers produce effects when one army is wiped entirely, and the winner suffers very small losses.

Implementation of the strategy:
· You get several full-stacks (2+) and tries to initiate the battles with higher numbers (2-1, 3-1,3-2 or more). The AI will flee but can do it only once. You get your army group on it again (what's enabled by the movement ranges), then autoresolve. The result should be the total annihilation of the enemy army(ies) with very low loses on your side - you can continue sniping immediately.
· If the AI has more armies than you, you need to do it patiently – snipe one army, then another one, then you dominate.
· It is sometimes employed also by the AI: it's not bad at positioning several of its armies around your army and attacking you. So the essential issue is to keep your armies together so that the AI cannot snipe by itself (and also to enable sniping by yourself). Perfect knowledge on the map enables it.
· After you’ve annihilated the AI in one province you go to another one and do the same. It’s really a war of annihilation. To prevent AI from spawning again you’ve got to “raze” settlements in the ATW or take them in the AoC.
· The result: playing the battles by yourself is really a worse option. Autoresolve is better plus battles with higher numbers are boring. You've got always lower loses and you can continue sniping. The composition of the armies (units), the quality of the leaders - matter much less.

Summary:
With unlimited manpower and automatically replenishing armies what you need to beat the AI is to wipe completely his armies. It doesn't make any sense just to win a battle - the AI would replenish and come back soon. You really need to wipe the armies. You can do it with the numerical superiority (ie more armies to attack 2-1 or 3-2) and then attack two times in one turn, preferably on the autoresolve (otherwise they flee, now wiping). What follows, the superior tactics is "sniping the armies": you make a battle group of 2 full stacks and take enemy armies one-by-one. You also take and raze the AI settlements quickly as a follow up (especially as the WRE - you need to raze constantly the neighboring settlements, letting the AI to spend money on rebuilding them, and then razing them again).

The interplay between a few elements (1. relatively low prices for the units, both recruitment and upkeep; 2. no-limit of recruits (as it was in the M2TW, 3. automatic replenishment) makes it the superior tactics. Otherwise you face a constant flow of enemy armies. Thus for the ATW to be playable a mod would have to introduce a kind of strong mechanism counteracting this mechanics. How to do it, I don’t know.

Age of Charlemagne is slightly better, I admit. First: lower movement ranges combined with some terrain features (loses in the mountain passes, much slower movement there). Second: less money from the buildings. Third: no razing (unless you're Asturia). However, all this doesn't prevent this strategy to be effective and superior, to my mind. Interestingly, I've seen the AI in the AoC also using this tactics: kept 2-3 armies together and then attacked together in two-phased attack, wiping my armies.

What do you think about this strategy? Am I wrong at some details?
JoC

[Pathfinder]

I know this is exactly how I play, all TWs since Shogun 2 in fact. I never try to bleed the AI's money though, that's a lost cause. The early game is the most fun to me because I can't just strategically auto-resolve, and by the mid-game you're punished for doing anything else with multi-stack battles you can't control efficiently and require supercomputers.

[Theramines]

There's a Steam Achievement for manually fighting every single battle in a campaign. I tried it and lasted about 100 turns. So many settlement defence battles.

Anyway, in response to OP: Yes, that is pretty much how it goes. It's pointless trying anything other than an aggressive war of annihilation against settled nations as the AI only comes back stronger if you wipe out individual stacks. Against hordes it's a different matter; they do suffer the effects of lost battles, except the Huns of course. So as it goes, the real danger in ATW are the 2-4 region nations, such as the Celts and the Franks in the west and infuriating, persistent bu*gers like Mazun in the east who can field full stacks because of the magic income bonus, turn up out of nowhere and snipe settlements, and then disappear into the fog.

I've tried various things to remedy this, with varying degrees of success. The removal of forced march is a qualified success because at least you can catch and destroy raiding stacks. Increased garrisons can be 50 50 because the campaign becomes a little too easy and AI reluctant to attack. A small increase in garrisons simply means that the AI attacks, but you can win most defence battles if you fight them manually: too repetitive. Reduced magic income leads to fighting spear levies and slingers over and over. However, it's a bit like swatting away flies as it doesn't really deal with the problem of losses ultimately having no effect on the AI.

A couple of solutions perhaps lie outside a population mechanic: Firstly, increasing the value of trade and agriculture to the point where magic money can be removed entirely, making economies more turbulent and by extension recruitment. It's something I've wanted to try but the amount of work and balancing required puts me off.

Secondly, utilising the mercenaries' pool replacement mechanism in some way, whether making all units 'mercenaries' or somehow tying it to faction recruitment pools. Currently, mercenaries pools are replenished in a gradual manner and a replication of that for factional units would make for a Medieval 2 type replacement system for destroyed units. Losses would matter. The problem of magic replenishment of damaged units would still be there though. And you'd have the problem of raiding Ebdani recruiting Comitatenses Spears in Lusitania. Again, it's a lot of work (every single unit would have to be pasted into the startpos) and likely outside my current modding skill set.

Finally, and bizarrely, I considered automatic attrition which balances replenishment rates, necessitating recruitment of mercenaries and new units, and merging damaged units to replace battle losses. Of course, the AI would simply collapse. I don't think it could handle that.

[Pathfinder]

You could remove reinforcement, that is, two armies joining the same fight. Ie. make everything a night battle. I don't know about you but I can't control 20 units individually let alone 40. I don't know if it's possible to mod reinforcement range, and it would make 1 stack vs 1 stack + garrison fights a nightmare but it would be an improvement imo.

[bigdaddy1204]

What about slowing down the auto replenishment? In some cases it's so slow already it makes more sense to disband the unit.

I agree with limiting the recruitment pool like in medieval 2. I also very much agree with removing the magic income from all factions. They should be entirely dependant on their real income from lands and trade.

As for sniping stacks, I do this too. But this has always been possible since the earlier games. I quite like that a stack which loses is totally destroyed since it means battles matter more.

[AbuKariba]

AI in Attila drews food from nowhere, AI factions manage to raise and upkeep armies in provinces with negative amount of food. It doesn't make the game much harder (at least on Normal difficulty), but is very illogical. Is cheating AI really the only way to make the game more challenging?

[alhoon]

Frankly, I was using "Sniping armies" in all total war games. I don't remember the days of MTW as well, but I think it was entirely possible to throw in 2 armies in a province. From the moment RTW showed up, of course I relied on number superiority and hitting the enemy with much more people. Numbers count; I had 20 stack armies even if I had to fill half the stack with peasants in M2TW. At the very least, the AI would spend time chasing fleeing peasants and tire their forces.


If I may, R2TW's and ATW's limit on armies is what makes the game more interesting IMO. In M2TW and ETW when you had 15+ regions you could afford to take out everything. When you got to 25+ regions? The game was practically over since even if you were at war with all your neighbors (and you wouldn't have many) you could easily defend your lands and have 2-3 stacks left to easily press the attack. I have played several campaigns where I didn't lose a single settlement.

In R2TW and ATW, things are harder. Sure, you have 8 armies to defend your 25 regions but if you're at war with 2-3 neighbors, even if they are 2-3 regions enemies you would be hard pressed and you would have to take risks. At most you could field 5 armies on the front-lines. If you're attacked from 2 sides you simply don't have the armies to defend-and-attack at both sides so you have to concentrate on defending in one front and press the attack in the other front trying to quickly gain victory.


Replenishment: I have in many R2TW campaigns reduced it. And then I go and I change it back the original because I don't want to sit and wait for my armies to replenish... :-)

[Jurand of Cracow]

Frankly, I was using "Sniping armies" in all total war games. I don't remember the days of MTW as well, but I think it was entirely possible to throw in 2 armies in a province.

Sniping in RTW/M2TW is much less efficient for a number of reasons (lower movement points, easier evasion of an attack, small zones of control). To my taste, it's a much better solution than those in the later TW games.

From the moment RTW showed up, of course I relied on number superiority and hitting the enemy with much more people. Numbers count; I had 20 stack armies even if I had to fill half the stack with peasants in M2TW. At the very least, the AI would spend time chasing fleeing peasants and tire their forces.

One of the things I'm looking for in the mods is a constraint on such a behavior. In the M2TW mods you may find different solutions:
- adverse traits gained by the generals (Byg's Grim Reality solution) for fighting constantly weaker forces,
- very weak peasants and militia units (many mods) that flee instantly (actually, it's not true that numbers always count),
- slow re-fill rates for the new units (various SS submods),
- high prices and low availability on mercenaries.
If you have a good M2TW mod then this strategy is much reduced or impossible.

I may, R2TW's and ATW's limit on armies is what makes the game more interesting IMO. In M2TW and ETW when you had 15+ regions you could afford to take out everything. When you got to 25+ regions? The game was practically over since even if you were at war with all your neighbors (and you wouldn't have many) you could easily defend your lands and have 2-3 stacks left to easily press the attack. I have played several campaigns where I didn't lose a single settlement.
In R2TW and ATW, things are harder. Sure, you have 8 armies to defend your 25 regions but if you're at war with 2-3 neighbors, even if they are 2-3 regions enemies you would be hard pressed and you would have to take risks. At most you could field 5 armies on the front-lines. If you're attacked from 2 sides you simply don't have the armies to defend-and-attack at both sides so you have to concentrate on defending in one front and press the attack in the other front trying to quickly gain victory.

I agree that after passing like 10 regions the game can be very easy. But some submods tried to constrain it as well. Eg. Byg's Grim Reality introduced the "War Councillor" mechanics which limits the number of the attacking armies to 3. Or the DLV with exponential costs of additional settlements above 10. Or increased rebellion changes. The number or the armies in the M2TW can also be limited by the requirement "no-army-without-general" (due to the quirk in code that has to be a home-rule, unfortunately).
I think the M2TW has much more potential in modding that than the later TW titles.

Replenishment: I have in many R2TW campaigns reduced it. And then I go and I change it back the original because I don't want to sit and wait for my armies to replenish... :-)

Well, if you want to go on the conquest spree all the time, then we're looking for different things in the games.
And I think you'll be disappointed with Thrones of Brittania by that balancing mechanism of war weariness. Me, I'm actually looking forward with hope to the ToB as the number of units is going to be limited, and other balancing mechanism put in place. But let's see how they will fit together.

[alhoon]

Many of the modding changes you mention are possible in R2TW and ATW too, I think. There are mods that deal with manpower so you can't go on a recruitment spree and it's easy to lower the replenishment rate. Traits can also be scripted (I think) to appear if you fight constantly weaker forces.

The movement could be lowered but IMO, since the map is so large it shouldn't be. In M2TW the movement was shorter but so was the map... from my games I recall that to go from A to B was taking about as many turns in M2TW as in ATW (which I play now, I don't remember R2TW).
In all these games the movement is unrealistically low; it didn't take years to move through friendly, well mapped terrain (thus in my mod, movement in friendly terrain is faster). But of course, it's a matter of game balance.

As for the limit on armies in M2TW... well, it's built in in new games as is the city garrisons.

Expansion in ATW, R2TW and especially S2TW brings with it penalties. In the case of S2TW kinda crippling penalties... And also the games are still challenging even if you have 40+ regions so there's (IMO) no need for artificial penalties to limit one's number of provinces owned. However, it could be easily modded to cause crippling corruption if you expand too far, or put in large public order penalties etc. Just a few changes to the Database so that victories don't give Imperium but only regions do and then adding corruption and public order penalties for high Imperium.

[Basil II the B.S]

Unfortunately, the OP's way is indeed the way the game is played normally, which is why it was a letdown under many points (this is especially bad in R2TW, Attila can be alright). I made a long thread some time ago discussing how much this is influenced by the provinces system for a number of factors: http://www.twcenter.net/forums/showt...ement-systems)

I didn't include high replenishment rate and high moving points, but those are also contributing factors. Another element are agents: I don't how you guys use them, but for me, in both Attila and R2TW agents have very little meaning except for adding them to an army and letting them develop those traits that benefit OP's strategy the most, notably moving points. Any agent, regardless of whether it's a spy, diplomant or champion, can add a 18% extra moving points and you just have to let it stay in the army. In Shogun 2, you could have agents do that, but you'd be wasting slots and they could be more useful (or even necessary) doing something else, now you don't need them. So you'll have armies that can effectively walk through an entire province (meaning 3-4 regions) within a single turn.

Definitive evidence: I finished the Parthian legendary campaign autoresolving every single battle. I fought none. And it's not like I'm an expert of R2TW or Attila campaigns. I barely know the buildings trees.

[Jurand of Cracow]

Many of the modding changes you mention are possible in R2TW and ATW too, I think. There are mods that deal with manpower so you can't go on a recruitment spree and it's easy to lower the replenishment rate. Traits can also be scripted (I think) to appear if you fight constantly weaker forces.

Maybe, but I've tried a few mods, and I haven't found any that provide for experience I get playing the recent M2TW mods, that are much, much better than vanilla or the mods of old). Even the DeI doesn't feel good to me - what is rather the fault of the R2TW, the mod itself is an ingenious try to salvage a bad game, IMO. Sniping is still the best way to play.

Another element are agents: I don't how you guys use them, but for me, in both Attila and R2TW agents have very little meaning except for adding them to an army and letting them develop those traits that benefit OP's strategy the most, notably moving points. Any agent, regardless of whether it's a spy, diplomant or champion, can add a 18% extra moving points and you just have to let it stay in the army. In Shogun 2, you could have agents do that, but you'd be wasting slots and they could be more useful (or even necessary) doing something else, now you don't need them. So you'll have armies that can effectively walk through an entire province (meaning 3-4 regions) within a single turn.

Yep, you're right: I kept veterans only in the armies. The priests were very stationary (in the provinces with low PO), and the spies actually were the only useful for the gameplay - for reconnaissance.
I think that it's another element of the newer TW titles that is inferior to the previous titles. I think the modded agents in M2TW (and I insist "modded") are much better.
As we've seen in the recent video, they won't appear in the Thrones of Britannia, so the CA came also concluded that they're not needed.

[Basil II the B.S]

Am I reading that correctly? Agents won't be in Thrones of Britannia?

[Jurand of Cracow]

Am I reading that correctly? Agents won't be in Thrones of Britannia?

If I was watching that video correctly - no, they won't be there.

[Basil II the B.S]

Wow, looks like it's going to be another let down.

[Jurand of Cracow]

Unfortunately, it seems that it keeps on being the dominant strategy also in the Thrones of Britannia. I've found this to be a very interesting video, just have a look.

[Jurand of Cracow]

In the recent R2TW-DEI the AI is perfect at doing exactly this strategy. Have a look at AI 4 armies (out of 6 in total) ganging up to my 3 gang ;-)
When one of my armies just strayed from the group, the AI attacked the rest. I'm in awe how precise the CAI is now.

[Basil II the B.S]

That's pretty good. At least it forces you to either actively use spies like in Shogun 2 or strategize around the map, unless you want to engage in massive battle, which is an interesting option as well, though in my experience the best the current system can handle is 2 armies for each side.

[Patronus86]

The manpower limits inherent to Medieval 2 were an awesome gameplay mechanic. The force pool replenishment of certain high-tier units was very slow, which meant that you had to use those units carefully and have reinforcements ready after major battles.

This mechanic was further developed in some of the Medieval 2 mods, where you had significant limits on recruitment: you could only recruit units in settlements that had high-enough culture or certain unit recruitments were restricted to certain specific settlements. Third Age Total War did an awesome job with this mechanic and it really forced you to be careful with how you employed your forces in big/bloody campaigns.

I’ve found that Empire and later games (and their corresponding mods) just don’t do as well with this mechanic. Even in DME Empire, I can start to recruit generic infantry units within a turn or two of conquering an enemy city (assuming the appropriate recruitment buildings are in place). I've also found that i can reinforce depleted units, even when afield in enemy territory (ETW).

The earlier games (Rome I and Medieval 2), especially when complimented with mods, really did a better job of making the player value his/her units as the recruitment was slow and, at times, restricted by location.

[Jurand of Cracow]

The sniping groups of armies is alive and well in the Thrones. It's easy (though tedious) to paint the map with fighting just a few battles at VH, perhaps possible entirely without.

Spoiler Alert, click show to read: 

[Advenas]

The resplenishment mechanic is hitting me hard as I'm going down on older TW games.

I love Attila and played (and is actually playing) a lot, but compared to my very short experience with older games, it looks like an indistriualized-reinforcements and a constant bloodbath.

Not saying that's a bad point, and I know too less of this period to say if it's totally inacurrate (I insist on totally).

But I'm much more picking my fights in M2TW, as early on, I can feel that a costly defeat may appear crushing for my campaign.