Doomstacks

Hello Vivisector 9999,

I also feel strongly about doomstacks, more than that, I hate the propensity of the AI stacking all its ships in a single stack throughout the game, regardless if it qualifies as a doomstack or not. Wipe out their sole fleet and they are done. It leads to single battle wars.

The below is an excerpt of a post in Command Issues some months ago, I am re-posting it here both for entertainment purposes and because it is relevant.

Two issues suffered by most (TBS space) 4x games are the number of ships one hopes to effectively control in battle and the "Doom Stack"

As we have not actually had any conflicts in space (yet), I will use naval warfare as an analogy of to illustrate the problems of these two issues in todays games.

2,500 years ago there raged a war in the eastern Mediterranean between Athens and Sparta. This war was fought on land and at sea as that part of the world is dotted with islands and offers tremendous coastlines to defend or attack. Deep into the war the Athenians suffered a catastrophic defeat at the bay of Syracuse and their navy was decimated. Had this been a 4x game the war would have ended shortly thereafter. The war did not end there however as the Athenians fought on for years after this event.

Fast forward to roughly 75 years ago where the IJN was soundly defeated at Midway by the USN. Again, had this been a 4x game the war would have ended quickly after that, instead not only did the war continue for years but the IJN actually retained quite a lot of offensive capability as well.

For me, it is fairly obvious that the reason real life did not imitate a 4x game is the absence of the "doom stack"! Let's face it, if you are here reading this and participating in this effort you know exactly what I am talking about. Moo2 and my personal favorite IG2, both suffered from AI factions that would lump every single ship they had into a single stack. Defeat the doom stack and the game was pretty much over as all you needed to do was mop up the faction that was suddenly completely and totally without a fleet.

In the analogies I use above this did not happen, why?

The Athenians had 100 ships in reserve that were released after the defeat at Syracuse, this allowed them to continue the contest and be effective. In other words there was no "doom stack" for the Spartans to defeat at Syracuse bay and end the war, just a sizable portion of the Athenian fleet

The Japanese were also able to continue the fight as there was no "doom stack" for the Americans to defeat at midway, just a sizable portion of the IJN with the rest scattered all over the pacific.

Personally, I would love to see this game avoid the "doom stacks" as it is a strategy that is usually only employed by a much weaker force attempting to isolate and defeat a portion of a larger force, otherwise no sane commander would amass his nation's entire navy (or army) into a single unit, if they did, then wars could be a single battle affair. (In space 4x games, where one end of your nation is many turns away from the other end, it is strategic madness to concentrate everything you have at one point, barring late game "star gate" technology of course)


ChrisKonstantine, Nov 4, 2016

 

No worries V,
This thread is more focused and just the tittle alone will draw more members into the discussion.

I am glad you created it,

 

This would solve the problem, no need for other measures. Putting all eggs in one basket is intrinsically a risk and in my opinion sufficient remedy would be to give other side means to make that risk backfire. In RTS games area of effect attack can seriously punish fielding large army in one place (remember psi storm in Starcraft 1?) and in Civilization series a single unit succeeding an attack roll would kill all units in a tile. In MoO 2 you can simply pass over doom stack, attack now lightly defended staging colony, reduce attackers range and force them to retreat.

BTW, what kind of doom stacks are we talking about, 5 battleships in a mid-game or 50+ titans in end-game?

 

ChrisKonstantine, I like your examples, and i think i see something underlying them, one of which is logistics. The ancient Greeks didn't wage war constantly, after a battle they would go home and tend to their fields, and come back the next year to wage more war. Logistics also plays in to the USN in WW2, it would have been difficult to support with food and fuel the entire US navy all sitting in 1 spot. The other thing i would surmise is that it would be rather difficult for all the ships in a hypothetical WW2 doom stack to bring their guns to bear in a single mega battle. Only the ships on the front line would be in combat, with the ships in back waiting for the ships in front to die (sink) before they could enter the battle. That wouldn't really make much sense. Fleets are a certain size because they have a limited practicality, guns can only shoot so far, and ships have to be spread out. Even the D-day fleet was spread out over a hugely vast area, and it only attacked land targets.

Having a single mega fleet in orbit could be a tremendous drain on the local resources. Moo2 combat is also not realistic (and I dont think it should be entirely, its a game after all), because the ships are too close together, which allows for a doom stack to be effective in battle because all the ships in the stack can actually be used.

In Moo2, if the enemy (lets assume both players are human, and both are totally equal) divided his fleet in half and attacked on 2 fronts, and you countered 1 of his fleets with your doom stack. Your first battle you'd outnumber him 2 to 1 and would decimate his fleet. You would then attack his 2nd fleet, again decimating it. You now still have a decent fleet left and he has none, the game is over for him, even if he did manage to capture/destroy a few of your worlds.

In real life, battles take a long time such as invading and capturing an island, and it also takes time to move fleets around and re-supply them (logistics). If you spread out your attack and hit the enemy at many locations all at once you could do a lot of damage as you would cut off their supply lines and it would take time for them to mount a response. This is why there were usually several simultaneous battles going on the pacific and in Europe.

I agree with the above poster about area effect weapons. Civ4 had catapults/artillery to do damage to many units at a time, allowing you to quickly decimate a doom stack. In Civ4 multiplayer you couldn't build a single doom stack or you would get it quickly destroyed. You needed several stacks moving together to a create a doom army.

 

And I like your analysis, it is very intuitive and quite correct.


Here is another excerpt from Command Issues


In real conflicts you could theoretically mass all your forces together until you consider two issues.


Logistics (makes me laugh to read some ancient claims stating that the Persians brought 3 million men with them for their conquest of Greece, really? how did they feed them? what about the sanitation required to keep them healthy? did they not need to drink water? etc.)


For a space 4x game however, keep logistics light (unless your target audience consists of accountants). Stay in range and you are automatically supplied works fine.


Command and control, same example as above, how did the Persians effectively command 3 million men 2,500 years ago? answer, they didn't. The actual number of Persians involved was about a tenth of the number suggested by some historians.


At the end of the day we could go into greater detail but unless someone actually requests a Naval operations 101 class let's keep it basic.

Some ideas

In port costs versus field costs. These are different in real life, it could be replicated here

Facilities, range could be dependent on additional structures, right now you conquer an alien planet and your range automatically increases. Really? So an alien culture just happens to have all the spare parts, logistical infrastructure and repair facilities for your ships? Doesn’t work like that.

I recall a Norwegian built ship coming into drydock and we were scrambling to find material and parts based on the metric system as we were equipped to handle American ships which didn’t use it.

And of course reserves and mothball fleets. It costs roughly one fifth to upgrade an existing ship versus building from scratch. If reserves exist then the “doomstack” becomes less of an issue as a faction might be able to recover from losing all their ships in a single battle, right now they’re toast.

Basically my friend, I do not have much faith that the doomstack will be cured in my lifetime, therefore another approach may be how to make it less of a factor by introducing other aspects into the game which will limit the impact it has, try and work with it.

In other words when life gives you lemons you may as well make some lemon margaritas.

 

Again, what kind of doom stack are we talking about? One that gets thrown at you after turtling for 400+ turns or half a dozen battleships on turn 200-. The first case is a degenerate game state where game should have ended long time ago and therefore I wouldn't consider it for further discussion. The other case doesn't sound so menacing but at that point in the game 5+ battleships can pack a lot of fire power, take down any planetary defense and take a lot of beating.

In a normal case he would retreat and wreck a havoc on the other front. It would be quite rare for a warp dissipater be in effect because you'd would either have to skip two very useful techs or be creative (which good players don't pick). If you can wipe him out in 1 turn with mere 2:1 advantage then he is either having poor defense (no ECM or PD if you favors missiles or no inertia stabilizer or augmented engines if it is beam fight), game is in degenerate state or there are very specific circumstances (ion cannon vs unshielded). At that point in game missiles can't reach target in 1 turn and beams are not that powerful, especially if target has some beam defense. At best you'd get him decimated in true meaning of the word which means 10% killed. While you are busy winning a battle you'd be loosing a war, he'd probably attack with 2 fleets simultaneously so while you are handling his first fleet, the second one glasses one of your planets, you try to catch them, they pull Hannibal on you and move to other system, bomb it, in the mean time first fleet returns and gradually your economy starts falling behind and you star loosing fuel range needed to mount offense. No, given 2 equal and skilled players and non-degenerate game doom stacks are not an issue.

 

I don't know about anyone else but for me it doesn't really matter if it is a mid game doomstack or late game.

In blunt terms (a little uncharacteristic for me)

AI factions keep their entire fleet concentrated in a single stack throughout most of the game, resulting in single battle wars.
Very exciting (sarcasm)
Very realistic (double sarcasm)

and quite possibly the most annoying aspect in 4X games. Even my favorite IG2 did this but...
that AI knew when it was getting it's ass kicked and withdrew. due to factors that allowed some ships greater speed than others some ships could escape and if they got outside your radar range they were safe. (That game understood how to use strategic depth, Moo2 did not)

If the issue can't be fixed then at least reduce the impact it has. The first is difficult to do, I know that, the second is easy and you do not need to rewrite the game to do it.

But I've said enough on this, time to withdraw from this conversation and observe instead.

 

But placing penalties on doom stacks or giving bonuses for spreading forces without improving AI would only create more issues and not solve this particular problem.

 

One last post then,
No penalties are needed on doomstacks
No bonuses needed for spreading ships out.

I'm just saying that an alternative approach to trying to "cure" them is to live with them.

Just minimise their impact through the use of reserves and how quickly range is extended once a planet is conquered. They would still exist in this case but losing a doomstack or facing one would no longer result in a single battle war. If doomstacks are so difficult to remove in a natural manner then one could look at reducing their impact instead as an alternative solution.

These two approaches are not all that can be done either but even by themselves they could be very effective.

 

In my generic scenario, I was referring to the fact that if you have numerical advantage in a single battle (because he has half his fleet and you have the whole fleet), you will destroy more of his ships then he will of yours, regardless of whether or not they retreat. Then any damaged ships you may have will be instantly repaired the next turn. Since you took less losses and are now fully repaired, your fleet will still be larger then their second fleet (or if they retreated, your combined total fleet will be larger then their combined total fleet) and now you will have numerical advantage for all remaining battles, where as at the start of the war, you were both equal. How you play out from there, separating your own fleet and going on the attack, or separating your fleet and hunting down his fleet is irrelevant, you have the numerical advantage so all else being equal, you should win. Thus the attacker would have been better off with 1 mega fleet and hoping he wins, or can catch a few planets and glass them before the defender gets his fleet to you.

 

I really agree this, and its why i've always said conquering a planet should take many turns, as should colonizing a planet. In Stars in Shadow, i never bothered to conquer, i would just glass and colonize the same turn. It was far faster and more effective to just build colony ships and bring them with the fleet instead of a hundred military transports. That way i could hop from planet to planet every turn. But this tactic should not be allowed either, maybe by making it take many turns to wipe out a planet by bombing, and then the planet is uninhabitable for several turns so it cant be immediately colonized. And after colonizing it should take some time before that planet extends your range (it takes time for the infrastructure to be built).

One way to counter doom stacks would be to have increasing support costs per ship based based on fleet size. This cost representing fleet logistics would increase exponentially, because of the larger burden a huge fleet places on local infrastructure (think fuel production, food/agriculture production ect... needed to support a large fleet). As the fleet gets larger, it puts more strain (more demand) on the local supplies, and with supply and demand economics, with high demand and static supply, costs are going greatly increase. With local supplies exhausted, resources have to be shipped in from further away, which again will increase their costs.

Like the 3 million man Persian army, it would quickly exhaust all the food in the area, and thus food would have to be shipped in from further away, which would just be unrealistic in 500 BC without refrigeration. Many smaller armies spread out would have easier access to food and clean water.

Lets consider exponential support costs per fleet size, for example:
A ship of hull size 1 (a fighter) would be 1 hull point.
A ship of hull size 5 (battle ship) would be 5 hull points.
For every 5 hull points contained in a fleet, the fleet has an extra support cost of $1 per 5 hull points then squared.

A fleet consisting of 10 fighters would have 10 hull points in it, so it would cost ((10*1)/5)^2=$4 per turn
A fleet consisting of 7 battle ships would cost ((7*5)/5)^2= $49 per turn.
A fleet consisting of 10 fighters and 7 battle ships would cost ((10*1+7*5)/5)^2= $81 per turn.
The 5 divider could be increased with technology advancements that would make larger fleets a little cheaper as the game progresses.

 

But you can generalize freely. 2 frigates vs 1 frigate on turn 50 is nowhere near comparable to 10 destroyers vs 5 on turn 100. You can't just postulate numerically bigger fleet would suffer less casualties either. It's quite possible that both fleets would get obliterated due to ship explosions and sheer number of nukes going off. And the larger context is important. If opponent attacks you with half fleet and defends with other half then you have ever right to gain advantage unless MAD happens. If attacker splits fleet to strike multiple targets then you may end up with Pyrrhic vicotry.

That reminds me of Deadlock, a clunky ground based TBS. There a map was divided in regions and you would build a separate base in each. If one base lacked some goods it could import them from other regions but the transportation costs would raise by a square of moved quantity. It worked quite well there.

 

I personally am in favor of training the AI to aggressively build doomstacks.

The reason why is because the alternative, as demonstrated in Civ 5 is a "carpet of doom", something the AI is even less prepared to fight back against. The only real solution is to have the AI build doom stacks in response to the human player.

 

Have to interject here,

This is the approach that I'm favoring myself though for slightly different reasons. Let's keep the Doomstacks, yes I think the AI will handle itself better as well. But let's go further.

You can have a game where the AI chooses to doomstack but you are free not to and you can have a game where the loss of a doomstack can be recovered from.

Quickly released strategic reserves will help. It would also address some of the concerns possibility has raised.

 

A doomstack is only a valid tactics if your fleet has no other use than fighting main battles. In history we find both types. Doomstack (spanish armada - which destruction ended the spanish superpower) and dispersed fleet (english fleet which protects colonies and trade routes overseas etc.) As long as a 4x game is so simple that it does not provide any other use for your fleet than just wait till the invasion begins it makes no sense to divide it into subfleets.

What I want to say is that it is the job of the devs to provide for more uses of a fleet than just big battles if they want to avoid doom stacks.

 

Here's my point to the story.
I'm gonna use the game Stellaris to illustrate my opinion:

In Stellaris you are encouraged to doomstack if you can win against the other fleet (i.e. you have more fleet power than them) but you are discouraged to do it if you can't and thus have to break apart your fleets and try to lure their main fleet away by attacking their mining and research outposts.
I'm not a huge fan of space combat in Stellaris. It's a very binary strategy to adopt and an unexciting one.
The main problem is that there is no way you can win if you don't have enough fleet power to rival your enemy. BUT ALSO there is no way you will inflict significant damage to the other fleet if your fleet power is too low. To the point where you might not even destroy a single ship if your fleet power is twice as low as your enemy's making it pointless to make a stand to defend your assets. MOREOVER, if one of your fleets gets engaged into combat, you cannot retreat until a certain amount of time has passed and thus most of the time your fleet gets destroyed or suffer too much damage to be of any use later on.

Thus to come back to PSS, in order to prevent doomstack strategy from happening you would need:
1/ To make smaller fleets more efficient and able to avoid and evade overwhelming forces.
2/ To make smaller fleets being able to dish out some damage if caught by overwhelming forces (Space Monsters i'm looking at you).
3/ To ensure that there is some cost in managing doomstacked fleets (increased maintenance, less speed, requires an admiral, etc).
4/ Reduce the micromanagement of smaller fleets (auto-fleet building, AI lieutenants to whom you can give tasks like raiding or patrolling, etc).

 

There should be bonuses for small fleets - maybe make them good at raiding?

 

I think this issue comes a lot from just static A.I that is not opportunist that doesn't see an empty sector ripe for the taking when you're at war with someone else.I also think traval is too fast in all these games so you can reinforce very quickly.Third point is losing planets should be big deal so just mopping up lost planets after you destroy their fleet like in most of these games in just very easy to defending planets is not a big deal.

Moo3 while crap had the best ground combat system I have played.An indepth almost wargaming system with no fiddly micromanagement.