Doomstacks: the death of strategy

[Cruxador]

Using a doomstack is a simple matter of sending all your troops (or, in Stellaris, ships) for the thing you want most immediately. There is no real strategy to this. Just take your pre-prepared large horde of ships, all designed for the same kind of engagement since there's only one kind, and throw them at the problem. That's not to say that there can't be strategy involved in related systems, but the doomstack itself is strategically dismal. There's no question of how to divide your forces (don't) or where to deploy them (where the most damaging enemy is) and no swing to wars (doomstack is dead? The rest is cleanup). Compare it to actually strategic games...

Well, why am I talking about this now? Stellaris is like this, and so are a lot of 4x games. Stellaris has only one mechanic that really works against doomstacks - wormholes, which take longer for longer fleets and therefore take so long to move a doomstack that it may, situationally, be more useful to not do that. But now that's being scrapped.
And while I'm not against changing the FTL in theory, a lot of the changes described so far are very much encouraging doomstacks.

I mean, let's take a look at from both sides.

The big feature of the latest dev-log is defensible chokepoints. So... What's that good for? First of all, from a defensive perspective, it means you've got defense taken care of in a static manner, so that's one less force deployment question. No (or at least, greatly reduced) need to dedicate forces to defense. And on the offense, you've got fewer avenues of attack and unless you've got a jump drive you've not really got any way to deal with the ambush laid for you except to forcibly break through. And that means you need your whole fleet. Because there's only two ways to balance defense: As a significant challenge for everything that can be thrown at it by an equivalent-strength empire (necessitating doomstacks) or as something less than that (meaning it becomes easy to just roll right over it if you do built a doomstack) and it's pretty obvious what Paradox is leaning towards.

So, defensively, there's no need to have multiple forces, and offensively, there's far less reasonable ability to. On top of that, there's also less need, because of star bases. With the new starbases, you can make shipyards all in relatively few systems, which reduces the number of targets that are militarily valuable to destroy.

So... With the announced changes, there's less reason and less ability to meaningfully divide your forces, and doomstacks, already prevalent, are aimed to be the law of the land. Now, there's a lot of changes that we haven't been apprised of - maybe the size delay from wormholes will be added to hyperlanes and movement will be slowed in general so deployment position matters despite all this. Unifying FTL does open up some design space for that kind of thing. But so far the thrust of the change is pretty clear. And I gotta say, making wars less deep is really not what I think Stellaris needs.

tl;dr: Announced changes for Cherryh encourage doomstacks which decrease strategic depth of the game.

 

[zizard]

Doomstacking is the historically proven technique for avoiding defeat in detail on a strategic scale. Deviations from doomstacking occur at smaller scales. i.e. control (direct or indirect) of parts of fleets in an engagement. Overtly promoting this leads to an allergic reaction of fanboys decrying "micro in my GSG". So people talk about it through the euphemism of "the doomstack problem". But fixing the doomstack on the strategic scale means somehow inverting the fundamental principle of warfare of strength against weakness, while simultaneously preventing the even less desirable strategies of extreme-fleet-splitting micro-hell, and conga-line-effective-doomstack. I don't think this is possible, so they will keep quarantining the doomstack threads forever.

 

[Summin Cool]

It's possible. The issue is that in 1.8 almost everything in the game is biased towards doomstacking, from the ship losses, to the ability to catch a fleet so easily.

 

[Dr. B]

I am pretty sure the developers are very aware of this problem, them developimg *strategy* games for many years.

We have to assume that coming up with mechanics to change how fleets work is part of the coming update, the chance is now, and they have more freedom in design now with the new FTL model.

 

[Tim_Ward]

I've been playing Warhammer: Total War Total Warhammer Total War: Hammer Total War: Warhammer recently, and I don't feel the slightest urge to doomstack in the game.

Partly, that's mitigated because of the 40 unit per battle limit, but even then I'd think you could just roll around the map with three or four stacks clumped together smashing things with auto-resolve, but if I did that I'd very quickly lose a lot of my of cities and go bankrupt soon after, which would result in attrition and the evaporation of my doomstack.

So, yeah, I think the thing that really discourages doomstacks is having things you must protect in order to keep your fighting forces effective.

That and the wild-overkill of the doomstack also means a lot of combat power is essentially wasted.

Stellaris discourages this in the follow ways:

1) Ground combat is so ephemeral that it's really not a big deal to have your planets occupied. You can just take them back in like three minutes. Realistically ("""realistically"""), having a massive army occupying and dug in on one of your core worlds should be a massive military problem
2) It is essentially impossible for a smaller fleet to escape from a larger fleet before it is annihilated because of the ridiculous 1 month cool-down on emergency FTL, so ships have to clump together for safety.
3) Defenses might as well not exist so you can't split your fleet in half and then have the other half made up by space stations, except in the early game.
4) The fffffffffffffffffffffucking UI for managing fleets and moving ships between fleets is such an unbelievable pain in the ass that I wouldn't do it even if it was a good idea.

I don't really see how 'everything must be hyperlanes' enters into the picture.

 

[Summin Cool]

I've been playing Warhammer: Total War Total Warhammer Total War: Hammer Total War: Warhammer recently, and I don't feel the slightest urge to doomstack in the game.

Partly, that's mitigated because of the 40 unit per battle limit, but even then I'd think you could just roll around the map with three or four stacks clumped together smashing things with auto-resolve, but if I did that I'd very quickly lose a lot of my of cities and go bankrupt soon after, which would result in attrition and the evaporation of my doomstack.

So, yeah, I think the thing that really discourages doomstacks is having things you must protect in order to keep your fighting forces effective.

That and the wild-overkill of the doomstack also means a lot of combat power is essentially wasted.

Stellaris discourages this in the follow ways:

1) Ground combat is so ephemeral that it's really not a big deal to have your planets occupied. You can just take them back in like three minutes. Realistically ("""realistically"""), having a massive arm occupying and dug in on one of your core worlds should be a massive military problem
2) It is essentially impossible for a smaller fleet to escape from a larger fleet before it is annihilated because of the ridiculous 1 month cool-down on emergency FTL, so ships have to clump together for safety.
3) Defenses might as well not exist so you can't split your fleet in half and then have the other half made up by space stations, except in the early game.
4) The fffffffffffffffffffffucking UI for managing fleets and moving ships between fleets is such an unbelievable pain in the ass that I wouldn't do it even if it was a good idea.

I don't really see how 'everything must be hyperlanes' enters into the picture.

This.

 

[B3ndolf]

i agree with everything tim ward said.

given that the new system of claiming systems with outposts and being able to take control of those outposts and gain the income from any mining/research stations in the system during war is both going to give you something to protect and potentially push the game toward having lots of little raiding fleets and micro hell i wonder if the fleet cap that everyone is assuming will be a limit to the number of ships in a fleet will actually be a limit on the number of fleets you can have.......or maybe both

 

[zizard]

I am pretty sure the developers are very aware of this problem, them developimg *strategy* games for many years.

We have to assume that coming up with mechanics to change how fleets work is part of the coming update, the chance is now, and they have more freedom in design now with the new FTL model.

What happened in the last 1.5 years that finally lets them fix the problem that their previous years of experience didn't?

 

[Cruxador]

Yeah, Tim Ward knows what's up. The things that make other choices beside doomstacking necessary are things you need to defend, and insufficient mobility to defend them if you've only got one stack. That much I mentioned, but the over-kill thing is important too. If the AI was rebalanced so that you rarely needed doomstacks (to whit, if they would take advantage of your doomstacking by using smaller cheaper forces to be strong where you're weak) then the player couldn't get away with doomstacks... But that can't work with the current high mobility in the game.

And regarding that particular comparison, Total Warhammer only ends you up with two or three stacks, which is less than I'd call ideal.

I am pretty sure the developers are very aware of this problem, them developimg *strategy* games for many years.

We have to assume that coming up with mechanics to change how fleets work is part of the coming update, the chance is now, and they have more freedom in design now with the new FTL model.

Grand strategy isn't the same as strategy. What game has Paradox made where doomstacking wasn't a problem?

 

[Dr. B]

What happened in the last 1.5 years that finally lets them fix the problem that their previous years of experience didn't?

Well - looking at EU IV, there are combat width, attrition, morale etc that limit the effectiveness of one mega-army, as well as the need to siege fortifications.
Obviously, somewhat similar mechanics for Stellaris would help.

But it is the new tactical options that can really help. Examples:

-Starbase roadblocks.
With heavy starbases that you need to conquer to advance (like a fort), you can tie up a fleet for a long time. An FTL interdictor there, only to be deactivated after conquest of the system, would mean your doomstack is held in place while your enemy can run around freely, taking over your less defended systems, now instantly gaining your monthly resources. This means you may have to split your fleet.

-Area of effect weaponry.
Now that you can not so easily skip systems, mine fields (like Artemis fields from MOO2), are bad for over-large fleets. Also clain-lightning-like weaponry could be a thing, in limited numbers. You still win by sending all you fleet at once, but if it is a pyrrhic victory where you take a great deal of casualties, it is better to conserve ships/split the fleet.

-Admiral effectiveness.
This is unrelated to the new update, but what if an admiral could only command a limited number of ships effectively (based on level, tradition, tech etc), losing bonuses/gaining penalties from increasing fleet (power) size? After all, commanding a small, united task force has to be a simpler task than trying to maneuver an enormous armada.


There really are many ways to help the doomstack problem, and the new system for fleets make it easier.

 

[Dr. B]

Grand strategy isn't the same as strategy. What game has Paradox made where doomstacking wasn't a problem?

Acually, none, and it took a hefty number of hard limitations to try to break up the doomstacks in EU IV.
Wich simply means that you need a strict set of rules and associated penalties to enforce a system where doomstacks are not the optimal way.
A stricter FTL system is a necessary part of this solution.

 

[Bersercker]

Gonna repost my suggestion from the other thread:

Imo the easiest way to nerf doomstacks is to make raiding meaningfull.
Like the current full bombardment stance gives 4% to kill a pop and 10% do destroy a building each month. If the devs changed it from "each month" to "each day" i'm pretty sure doomstacks would just dissapear. Gotta make it scale with fleet size (so one corvete wouldn't be able to glass someone's homeworld) and the percentage of planet that is already ruined (like the longer you bombard the harder it is to hit something). Add some techs and buldings for planetary shields and make the bombardment influence the warscore and its done.

Even the defence stations would suddenly become non-useless because they wouldn't have to fight against doomstacks.

Acually, none, and it took a hefty number of hard limitations to try to break up the doomstacks in EU IV.
Wich simply means that you need a strict set of rules and associated penalties to enforce a system where doomstacks are not the optimal way.
A stricter FTL system is a necessary part of this solution.

Em, what about hearts of iron? The frontlines might not be ideal for a space game like Stellaris (same as space roads with roadblocks ) but there are no doomstacks there.

 

[Dr. B]

Em, what about hearts of iron? The frontlines might not be ideal for a space game like Stellaris (same as space roads with roadblocks ) but there are no doomstacks there.

Ah, yes that is true but as you say perhaps a bit too restrictive for a 3-dimensional space game.

 

[monsterfurby]

Thing with breaking up doomstacks is that it increases the complexity and micromanagement of war massively. There are a few solutions to this:

1. The HoI-route, where you'd delegate the micromanagement to the AI (you still coordinate "doomstacks", i.e. theatres, but those are broken down into smaller units within the simulation).

2. Slowing the pace of the game overall to make it more manageable - i.e. longer battles, slower travel, better response times.

3. Making a distinction between "campaign armies" and "garrisons" while keeping both viable. This is the Total War solution and one that works pretty well there. In the endgame of Warhammer TW, you'd usually have four or five full stacks for maneuvering, with mostly garrisoned border cities and points of interest, with your expansion HEAVILY capped by your income (which also serves as a guideline for scaling your armies until you know what you're doing).

I'd probably be alright with either solution, though a combination of 1 and 2 might be more elegant than 3, which relies on a hard army/fleet size cap.

Edit: also the stuff I mentioned here.

 

[Bersercker]

Thing with breaking up doomstacks is that it increases the complexity and micromanagement of war massively. There are a few solutions to this:

1. The HoI-route, where you'd delegate the micromanagement to the AI (you still coordinate "doomstacks", i.e. theatres, but those are broken down into smaller units within the simulation).

Ah, you're right if you don't micromanage all the forces all the time then the HoI frontlines may be considered doomstacks too. And this micromanagement might not fit Stellaris anyway due to it not being a wargame.

Though i'm sure the correct balance for the number of fleets can be found and enforced by influencing the minimal required fleet needed to do some damage in a timely manner. Be it eu4 forts or planetary shields that require a certain amoung of firepower to overcome "regeneration" or reducing the damage done to a planet by a significant percentage scaling with tech, etc.

Edit: What about introducing NATO counters in space to simplify the micromanagement?

 

[ZomgK3tchup]

Well - looking at EU IV, there are combat width, attrition, morale etc that limit the effectiveness of one mega-army, as well as the need to siege fortifications.

In EU4, you have several armies that stay below the supply limit until a doomstack shows up; then they Voltron together into a doomstack. EU4 definitely has a doomstack problem; it's just that the doomstack doesn't show up until it's time for combat.

What Stellaris really needs from EU4 is the interface. Once a ship joins my doomstack in Stellaris, it's never leaving by virtue of how difficult it is to do anything that isn't splitting the fleet in half or splitting off individual ship types.

 

[methegrate]

Thing with breaking up doomstacks is that it increases the complexity and micromanagement of war massively.

I saw this on the dev corner, and personally I completely disagree with it. Micro doesn't have to increase massively, and I think the difference between that and complexity is critical.

There seems to be this idea that the more clicks you have in the game, such as with more fleets, the more micro-intensive it is, but there's a huge difference between micromanagement and complexity/activity. In the former, you're just doing lots and lots of clicks as chores (for example, army attachments). It involves many points of contact, but relatively few decisions. The latter is about actually making and then executing upon decisions. That might still involve lots of clicks, but each point of contact represents actual judgment and engagement with the game.

So, there's nothing wrong with having more to keep track of during war. Increased engagement is good, as long as those extra clicks represent engagement and decision making. Ordering three or even a dozen fleets, or responding to that many, isn't micromanagement if each one represents an interesting gameplay decision. That's... well, gameplay.

Think about it in the context of Total War. In Total War you click a ton, but I wouldn't call it micro-heavy because most of those clicks involve an actual decision. Spend money here, build a unit there, send one of many armies or respond to someone else's. Lots of things happening, but few are rote chores, so there's not much micro. Chess would not be better served by collapsing it down to just two pieces, because even though you move 16 units around a board, you're thinking about your engagement with each piece.

Obviously we don't want the player to get overwhelmed, and Stellaris currently doesn't have many real, meaningful choices to make during warfare. So more clicks would probably feel like micromanagement right now. But that just means doomstacks are papering over a bigger problem. Without something interesting to do with those fleets, breaking doomstacks up would just lead to a tedious, micro-heavy experience, and that's probably what will happen if they put in fleet caps.

But the fault there comes from the fact that players don't have anything interesting to do with those units, not from the fact that there's anything wrong with doing multiple things at once. Fixing that by limiting fleet numbers is essentially conceding, "okay, yeah you don't have many interesting choices to make, but at least you don't have to make that many of them, right?"

 

[Jabby]

Its seems to me no one was watching the stream carefully because theres a new thing called fleet command limit as well as naval capacity. Wiz actually says its about have a big fleet when talking about the warfare tradition tree.

 

[methegrate]

Its seems to me no one was watching the stream carefully because theres a new thing called fleet command limit as well as naval capacity. Wiz actually says its about have a big fleet when talking about the warfare tradition tree.

I think everyone is expecting this next. They almost have to do it, now that they've gone towards chokepoints and starlanes.

It will be disappointing for two reasons:

First, it won't make Stellaris more engaging. Doomstacks are a symptom, not a root cause. Fleet command caps will still leave the basic problem that your best play is always to concentrate as many ships as possible. It'll just mean now you have to figure out other ways to do that, or now you right click as many ships as possible three times instead of once.

Still the same logic and gameplay, just with a few rules pushed on top. (This would be the micromanagement discussed above. In both cases the decision is the same, to throw the biggest blob you can at them. Just here, it will take 3 or 4 clicks instead of just one for a doomstack.)

Second, it's really, really derivative. Starlanes, wormholes, chokepoints and fleet caps... How many space 4x's work exactly that same way? Without trying to be offensive, it will feel like they just downloaded Endless Space and said "looks good to me!"

But yes... this looks like the next bold, groundbreaking announcement.

 

[Sinister2202]

I too have concerns with doomstacking. But it's not easy problem to solve, as trying to beat the other fleet eventually LEADS to doomstacking in the end, just to top one another. Because the opposite of doomstack combat involves smaller and divided fleets, doomstacking is always inevitable.

That's why I have suggested softcaps for fleet admirals to limit how much ships they can command in a fleet per their rank/level with severe penalties, should the fleet exceed the cap. And admiral level cap increase for supremacy tradition would also improve the fleet cap for a admiral as the admiral reaches the level cap. People felt somewhat mixed at my suggestion, but I am glad the developers are actually considering it. But some had pointed out that this would just result in multiple fleets follow one another, creating a doomstack with a work-around.

My other suggestion is to give peacekeeping missions to players across their empires, a sort of maintenance if you will... This would encourage players to divide fleets and station them across their empire, but war time will lead to doomstacking once again anyways, so that's more of a hassle than a solution.


Doomstack is way too over dramatic. I only see doomstacks in sci-fi tropes, in situations involving massive threats to their very existence, or a "final dramatic battle" against some great threat, but not in small skirmishes or a war against an empire you are very confident in beating in less than 10 minutes. Usually (in trope) that involves doomstack against a weakling is an evil empire bent on total domination. But that's just a flavor for rp.