Carriers - Not So Game-breaking After All? [Images]

Great post, and more or less lines up with my experience with carriers. Especially since the AI update, some of their fleet compositions have made my carrier balls much less effective. 

Great job on all the detail here!

As someone who does not frequently reach late-game, (and I haven't got there yet in 1.4 - I have a "new game" habit, okay?) does the AI build balanced ships at that point?

That are some nice tests but I could have told you just as much without doing a single test, just from understanding the how the combat mechanics work.

And the most important part of these mechanics would be that one weapon type of a ship can only attack one ship at a time.

So three specialised ships, one beam, one missles and one kinetic with 100k attack each are equaly effective as a single balanced ship with 100k attack of each type against a large force of small ships.

As in equaly bad because if each small ship takes 1k damage to destroy, then each 100k shot will waste 99k potential damage and only do effective damage of 1k(so 3k over all weapon types).

If you now take 10 balanced ships with 10k damage in each type which will total to the same 100k in each type for the fleet so equal in stats to the two previous fleets but only wasting 9k damage per shot and doing an effective damage of 30k over the whole fleet.

Now let's look at the other side of the coin a carrier that spawns lot's of fighters overkills very little because when x number of fighters have rolled enough damage to destroy something the rest of them will target something else. So the maximum overkill for fighters is always below the damage that a single one can do.

Their downside is that their damage is reduced faster as fighters get destroyed compared to a fleet of large ships, where it takes longer to destroy a single ship thus longer till their damage is reduced.

Now if we take all of this I've come to the conclusion that tinys and smalls beat huges because the huges with eqaul total stats because the huges massivly overkill them, thus their effective damage is a lot lower. BUT bigger thing beat smaller ones when they have just enough damage to destroy them in one salvo because then their effective damage is very close to their statistic damage. But this is limited by logistics because a huge that has just enough damage to kill tinys is a waste of logistic points, so it's limited to around two size classes lower after that it becomes inefficent. So with equal stats in each class we get something like this:

The problem why this system doesn't work that well is because of the large production available late game causes larges/huges and carriers to be produced in 1 turn, so it's inefficient to produce mediums to counter tinys and smalls. So either production needs to be nerfed or costs of the bigger classes need to be increased or another way to efficiently produce the smaller classes needs to be introduced. For this a have the suggestion of being able to produce batches of ships, so you can also produce 2 larges or 2/3 mediums or 2/3/5 smalls or 2/3/5/10 tinys at the same time. So for a 10 tinys you select a tiny and in the production queue you have 5 checkboxes next to the tiny to select how many you want to produce(can be changed until production on it has been started)

Very Very cool.  This looks like a LOT of work, how long did it take you?   Too bad the videos didnt work.  Great post, thanks

Technically, you can already batch produce ships by having more than one shipyard. It balances out somewhat because since smaller ships costs less, you can have one shipyard per planet producing smaller hulls every turn. Larger hulls like the one in my experiments require two to three specialized manufacturing planets to produce in one turn. If you split that up, you could have two or three times more smaller ships in the same amount of time. Alternatively, you could focus on Wealth and just buy all of your ships.

Nevertheless, even though you know this, it still seems like there is a slight misconception on how powerful carriers are amongst a good sized portion of the players. I see people saying that the game goes to crap once carriers come out all the time, so I also wanted to prove it to myself if what they are saying is true. It turns out that carriers are not the problem. 

Only about an hour or two to actually stage the battles and capture and crop the screen caps and write up the post. However, it is a bit frustrating that the images are not resizing like usual. I got it working before. Altough posting videos would really take forever.

Agree, never understood this as it has never been my in game experience.  Are carriers powerful- yes, are they "game goes to crap" over powered-no.  Must admit i had a Homer Simpson Dhooo moment wrt using a balanced weapons load out.

I only play on insane maps, where tech come relatively quick (even on very slow) and carriers are in a special place that do break many of the games rules.

1: carriers get to auto upgrade their weapons and armor (in the form of their fighters) with each tech level at no cost - no other ship or unit gets this ability afaik

2: carrier task forces are like the zombie horde - if you don't kill the brain (the actual carrier) they just keep coming back - as the tiny ship just regenerate after a few  turns.

If I went to build a bunch of tiny ships to escort my huge ships, or medium, or whatever; they do not have the auto (free) upgrade feature, and they do not auto come back to life after a few turns when they are killed either.

That is my personal view as to why carriers are overpowered, and to me do not make sense. A carrier is both a wealth generating device and a mobile shipyard as well as a warship.

This is odd what you've discovered about a ship maximum in a fleet battle, and if true a find that is well worth your experiment!

I haven't yet reached the late game after the wheel removal so the amount of available production might be less of a problem now.

Yes, you could use shipyards at every planet but in my opinion that's just too much to organise, even more so with the removal of the social/military slider per planet.

EDIT: @Syntax_VI: Those points are valid points, but if you had to load the tinys manualy they would both be fixed, plus some other things:

The fighters wouldn't auto upgrade, since they are not spawned by the carrier anymore, they would not respawn from nothing, instead maybe boost their hull repair when inside a carrier, with my batch production idea filling them wouldn't be too much annoyance and you could actualy decide the fighter design yourself.

If you are playing on Insane maps, though, there is usually plenty of room between factions for you to avoid needing to build ships for a long time, and you have plenty of time to build new ships and decom old ones. Although, I almost never upgrade ships because they cost money and time, which I can just as easily put into building another ship. True, they get free upgrades instantly, but they will still be nowhere as powerful as a fully customized tiny hull, which should cost nothing to produce by the time you get carriers.

As for your second point, this is also a major weakness. If you destroy the carrier, then no matter how many fighters survive, you lose all of them after the battle. So instead of focusing on clearing out the fighters, you can focus on a fast missile ship that can get close and unload a salvo of missiles on the carriers before getting destroyed themselves.

I see your point there. I really do not like the mechanics of one production per turn no matter how much resources you have. I think overflow should go into producing the next. For example if you put 5K into building a 2.5K ship, you should have the option of getting two, especially since the lack of the wheel makes it harder to avoid waste.

ah but distances are immense on an insane map, and a battle group may be very far from home for a very long time when on the offense. Being able to regen losses and auto (free) upgrade are very powerful abilities not matched by any other ship class. On defense this may not be such a big issue, but If you have several hundred carriers and upgrade to the next weapon type they are all that much more powerful on that exact turn that your research completes. No need to scrap and rebuild. This may not matter as much on smaller maps with small fleets, but on a big map it makes a huge difference.

It is especially poignant in the hands of the AI, because of its playstyle on higher difficulties.

I think this just comes down to personal playstyle. I play defensively until I have all weapons and defenses unlocked. Even when I do move out and on higher difficulties like Incredible or Godlike, the AI does not research weapons fast enough for me to have to worry about upgrading ships. I just decomm the old ones once their tech becomes obsolete. This has never been a concern for me.

Production already overflows to the next project, but that doesn't solve the problem of producing small ships because 1 small per turn doesn't compare to 1 medium per turn.

The problem is more the amount of turns needed to produce multiple smaller ships because of 1 ship per shipyard per turn.

And managing a lot of shipyards to offset that is annoying, even with very few shipyards I send ships in the wrong direction because I forgot to change the rallypoint. I don't want to imagine that with a lot more ship yards.

Forgive me, but I disagree with your interpretation of your results.

What you've actually shown here is that carriers are worth at least twice as much as other units. You have reported 11 wins for the carriers, 3 wins for the targets and 5 ties. That's a 16% defeat rate, against ships which cost twice as much. The carriers even survived in over 50% of cases, winning battles against fleets that cost anything upto 90% more than themselves. A proper test would involve using equal costings, and I suspect the carriers will wipe the floor with their opponents every time til you hit the ship cap with that methodology. When a hard cap is the only thing keeping stuff balanced, it ain't balanced.

I'd also suggest that using actual defenses on the carriers rather than hull buffs would tilt things in the carrier's favour, even when splitting between all three types. 

Syntax is also quite right in pointing out that the advantages of carriers are far more than their (enormous) potency in battle. Free upgrades mean that, while my battleships need to be retired every 50 turns or so, the carriers don't. That carrier I built on turn 120? Still viable on turn 500. It's smaller than a turn 500 carrier, but it still has just as much of a punch because it's using the top-level weapons, always. This is not something you can really sim into a test scenario like this.

Carriers aren't total unkillable God-ships, but then I don't think anyone ever claimed that they simply cannot ever be beaten. What you've really shown here is that carrier modules need to be roughly doubled in price just to bring them into the same ball park as everyone else, and even then they'll kill an equal-weight opponent in over 75% of cases, but only die in 50%.

I definitely pointed that out in my first conclusion that they are too powerful for their price. However, it is impossible to balance out the costs for a test without significantly reducing the number of Carrier modules. High Capacity Modules cost 67.5 capacity, which does not leave much room for anything else. The more you try to balance the price, the less of a carrier your carrier will actually become. To achieve the same cost as the rest of the ships in this experiment, you would only be able to have about 3 carrier modules max, which actually makes the ship worst, not better. My conclusion do acknowledge that the modules need to cost more.

I also agree that the carrier would benefit more from some defenses, but like I just mentioned, the more you strengthen your carrier, the less fighters you will have. You will not be able to mount any significant balanced defenses against endgame attack levels without dropping several carrier modules. Your carriers will still get destroyed because you have less fighters dealing damage, which gives the enemy more time to take down your defenses. You can try it yourself by enabling god mode if you have doubts and post your findings here.

On the other hand, I do not agree with your assessment that your turn 120 carrier will still be relevant for that long. Your fighters may get stronger, but unless you had unlocked all techs already at that time, your carrier itself will be obsolete. First of all, you would probably not have the capacity to equip that many modules early on. Fighter capabilities scale with their number. 3-9 fighters is nothing, even at endgame tech. They will still be weaker than tiny hulls, and the fighters and your carrier will be wiped out in an instant against later built hulls. Additionally, I have shown that buffs and debuffs make a large difference in fleet battle. Unless your carrier still had the capacity for fleet support modules even after equipping your carrier, and this is assuming you even have the tech unlocked at that time, then your opponents can easily make up the difference with a support ship.

By overflow going into the next I mean getting more than just one per turn. Bad wording I guess. If I destroy your ship yard, all your overflow will go to waste. In any case, This game seems to lean on the side of giving you more incentives to upgrade to larger ships than to diversify your fleet. It would be a pain indeed to put together a huge fleet of tiny hulls.

Your ships also lacked engines and life support, which is worse than useless at end-game. This is an artifact. The conditions that you've set the experiment up under have determined the result, but those cannot be generalized to normal play; it proves nothing about the relative strength of carriers and normal ships, but merely that the specific carrier designs that you created are basically a match for specific other, much more expensive ships you designed.

If you stack engines and life support, you would naturally lose even more capacity to stack carrier modules; therefore, your carriers would naturally be even weaker. These settings were designed so that the carriers can maximize what they are good at, and that is number of fighters. I do agree, however, that this is not truly reflective of normal gameplay. In a normal game, you are not going to get carriers anywhere near that many fighters. They will be significantly weaker because capacity will be spread among engines, defenses, and various other things depending on play style. Even I never stack more than 3 carrier modules on my normal carriers.

Nevertheless, the point here is not to emulate true gameplay because there are too many variable to consider, such as map size, race traits, opponent ship style, your total production, researched capacity, and various other things, but this experiment is to test these ships at the best of what they do. The best way that I see to do that is to omit components that are unnecessary in actual battles and redistribute that capacity to other more useful parts. If you can suggest a better design other than maximizing carrier modules, then I can try it for you to quell any arguments that you have.

I would like to see a rewrite of the fighter modules and make them again a 'special' class of  hull that is not available to the player. Call it Micro which would represent a single seat (non hyperdrive) fighter. Give it, its own set of rules and its power is based on what  you research within the tier of carriers. Also add MORE fighters to each module and remove any and all defenses from them but give them say 25 hps. This would give them some staying power but be fragile as well. 

Tie dmg output directly to a specialization that you do and it defaults to that tier, (laser,missile, railgun). Put 5 on the first module and make it cost a lot and its mass would preclude more than one on a large hull.  This gives players more tech choice that are hard to make..again making tech fun and difficult to decide on!

Just my +2 coppers. 

Remember the old joke about theoretical physicists:

Milk production at a dairy farm was low, so the farmer wrote to the local university, asking for help from academia. A multidisciplinary team of professors was assembled, headed by a theoretical physicist, and two weeks of intensive on-site investigation took place. The scholars then returned to the university, notebooks crammed with data, where the task of writing the report was left to the team leader. Shortly thereafter the physicist returned to the farm, saying to the farmer, "I have the solution, but it only works in the case of spherical cows in a vacuum".

You've made a spherical cow model. The level of simplification used to create the test here has created a scenario that is ultimately preposterous. 

Here's what I'd suggest.

Rip out a bunch of the fighter modules. Replace them with defenses and self-heal modules rather than mass-stacking precursor modules. Do not attach weapons to them, as that might change their role and push them up the enemy's target list.  Make every ship on both sides have 3 engines and a life support. Then set up the fleets so they match in cost and re-run the tests. That might actually give us something remotely approaching a gameplay scenario so we can get a reasonable measure of how OP carriers actually are in a genuine gameplay scenario. 

I don't think anyone ever suggested that carriers are completely invulnerable, or that they never, ever die (at least, no one who actually understands how roles work). Showing they can die doesn't mean that they aren't gamebreakingly OP. Showing that they can be defeated by an EQUAL fleet might actually prove something (though since they pretty much can't be, all we'd prove is that carriers actually are game-breakingly OP). Otherwise, why bother at all? You might as well just use huge 'target' ships vs Large carriers then report back that they won... great, yeah, no surprise. If you can only beat carriers 50% of the time when you're using ships that are twice as expensive, I honestly don't think it's realistic to interpret that as 'carriers not so game-breaking after all'; you'd basically be relying on other game-breaking features (i.e., massive overproduction) to produce more expensive ships to counter them.

If overproduction is ever sorted, then imagine what this test actually suggests. Brad has indicated that he'd like to see the most powerful ships take 4 or 5 turns to produce for even the best industry worlds. If your 14k manu ship is built in 5 turns, your 6k carrier is built in two. The carrier player has twice as many ships as his opponent, and still tends to destroy his enemy 75% of the time and only loses his own ship 50% of the time. That's still game breaking; my only realistic counter to the carriers is... more carriers. Nothing else can be built in sufficient numbers to matter.

Relative to what? Neither your carriers nor your targets have engines and life support, and in practical game conditions you'd expect both to have similar amounts of drives and life support. That's going to cost a fleet of battleships a lot, but it's not going to cost a fleet of carriers much at all; your carrier design has far more fighters than your own results suggest that they need (at minimum, you could drop ~2 carrier modules from each carrier and the single-battle performance of the 3-carrier fleet you tested would be largely unaffected as 21 fighters per carrier + 3 carriers = 66 vessels total, which means you get the same 61 fighters and 3 carriers in the engagement that your 27 fighters per carrier + 3 carriers group would have fielded but with far more space available for other purposes, whereas dropping a similar amount of space on the battleships will directly reduce the combat power of the various 3-ship groups you deployed, and 98 size worth of drives and life support is a lot of movement advantage to give up, especially at max-tech). And no, it basically does not matter that your battleships would win if the 3-carrier group with 98 capacity spent on drives rather than fighters came up against them, because under almost any practical circumstance a fleet that spent ~100 size on drives and life support cannot be engaged by a fleet that spent 0 size on drives and life support, and on top of that the faster fleet has a significant advantage in choosing where any battle will be fought - just try forcing ships with a base move of 31 to engage where you want them to when your ships have a base move of 1. You can't even really take away the strategic initiative by threatening shipyards and planets with things that are too expensive to engage because by the time you can reach shipyards and planets with your slow-moving juggernauts, the ships with a base movement of 31 could've killed most of the shipyards in your empire several times over with a significantly lower force commitment than it'd take for you to prevent that from happening.

Sure, we can try this too. Only, if you do not arm the carriers, they will just become sitting ducks when their fleet of fighters die. It is probably better toa rm them, and then set their role as support. Give me like, a day to get it set up. I will post back with results.

Dup post.

As naselus suggested, I think it is extremely important that tests be done with "real-world" ship designs.  The rationale for that, in my opinion, is that one of the reasons carriers are powerful (it should be and is one of their strengths) is that their fighters don't have to worry about engines or life support or sensors.  This is a potentially big deal, because it means that carrier fighters are naturally more powerful per hull space than a capital ship.  By ignoring non-combat concerns, you put capital ships on an even playing field with carrier fighters - and the game isn't designed that way; fighters are supposed to have an advantage in this regard.

In my tests I noticed that the NoDefenceDamageOverflowBug is still not fixed, which means that damage doesn't overflow from the defences to the hp.

This makes this discussion useless anyway untill it's fixed because this favours the small ships greatly.

To give an analogy:

Deathstar vs a bunch of fighters, deathstar shoots the first fighter and the fighter is like: "Oh no my deflector shield has been destroyed". Then while the deathstar is recharging, those fighters proceed to bomb that exhaustport.

It takes a "Deathstar" with 1000k Beam Damage two shots to destroy a 1hp 1shield "9Vbattery powerd shieldgenerator".

And a BorgCube with 100k in each type still needs two shots of at least one type to desrtoy a 1hp 1amor 1chaff 1shield "9Vbattery powerd, tinfoill armoured and tinsel decorated shieldgenerator", so it might tank up to 4 shots...

In one word the situation is still: broken

Except the small ships in this test only had missile defense, and the carriers had no defense at all. So this defense problem only affected the carrier fleet's attacks for the majority of the testing.

So after running even more tests, the conclusions are clear to me that a fleet of well built carriers will handily stomp any other fleet of equal cost; however, one-on-one, these carriers can be beaten by ships of equal cost.

By well built, I would like to say that the golden number of carrier modules to have for a carrier is 5 because including the carrier itself, you have 16 ships. 16 x 4 = 64, the hard cap without wasted ships. This leaves enough capacity for you to stack better defenses while still having a good number of fighters.. Of course, this is assuming that you play like me and research all mass-related specializations.

In any case, since it is pretty hard to define what an actual gameplay ship really constitutes, considering all of the scenarios that can happen, I will be going with naselus suggestion. The ships in this experiment all have 3 engines and 1 life support component, the best that can be unlocked, and some decently thought out components.

Also, might I point out that self healing in this game is USELESS? You are probably going to take more damage than self healing can ever fix by the time you unlock it. The devs need to rethink this.

The main carrier for this test is below. I also built a similar carrier with fleet support modules and less defenses, but the differences in performance were mostly negligible. For example, One-on-one, a missile ship vs a carrier will still end up with mutual destruction, and the balanced ship's rate of win went down from 100% to about 80%, depending on your luck with jamming.

http://i101.photobucket.com/albums/m59/solgrieving1/TestCarrier3_zpstpmgrvlk.jpg~original

The targets were these ships. They cost less than the carrier because I was trying to average the price between my two carrier designs, even though I only mainly tested with one. Keep in mind that these are not perfectly optimized against carriers. For example, 200 of any offense is enough to kill most fighters two times over. I did optimized a bit against the fighters' missile weapons, but it could have been better. Remember, optimizing against your opponents is basic gameplay strategies.

http://i101.photobucket.com/albums/m59/solgrieving1/TargetShips_zpsj8w39m6m.jpg~original

The results were as follows. I am only showing these four tests because the effectiveness of other ship types have been mentioned before. Furthermore, I am not showing 2v2 or 3v3 because the carriers gain a significant advantage the more ships you add, up to four ships, where the ship count hits their caps. When you add a fifth ship on each side, the carriers start losing their advantage as fast as they gained it adding up to four. 7v4 is when the fleet of carriers get stomped due to cap.

In any case, here confirms that 1v1 is no issue, but 4v4 and carriers roflstomp.

http://i101.photobucket.com/albums/m59/solgrieving1/Results_zpsp1xgqruv.jpg~original

So the results are obvious. Carriers are extremely powerful for their price, but before you start lining up those carriers for production, there are a few things to consider that this experiment has made me realize, besides the obvious number of ships in battle cap.

One, is the obvious cost. Sure, in a 4v4 situation of ships of equal costs, the carriers win every time, but the real question is, if you can build more expensive ships, why would you not? From another point of view, an effective carrier has a price cap of about 4-5K, but just because your opponent is building carriers at 4.8K each does not mean you have to build all your ships at the same price. In real game play, this is never going to be the case either unless it was intentional or coincidental. If you can produce 8 or 10K Military Manufacturing per turn, you really have no reason not to build more expensive ships. Even though military production does roll over to the next ship, you still can only ever get one ship per shipyard per turn, so it is more effective to build as expensively as you can produce in a turn. By endgame, it is not at all difficult for a single planet to produce over 5K production alone, and ships needing 10K+ can still be produced in one turn with a group of planets. In effect, cost is no object in endgame, and researched techs is what matters in mid to late game. Even without the wheel and all of my planets set to balanced (33/33/33+50/50), my best group of planet can still churn out over 5K military per turn.

Basically, Quality over Quantity is one of the ways that you can beat a capped fleet of carriers using the same number of non-fighter ships, and this game seems to really push for that because of the simple fact that shields completely recharge after every battle. Essentially, as long as your better quality fleet of ships does not sustain significant HP damage every battle, you can use one fleet to destroy countless carrier fleets; however, this is not the case for carriers, as they need a few turns to respawn their fighters. Therefore, you are not losing out by much by combining four planets to produce one expensive ship vs your opponent using one or two planets to produce one carrier. Of course, things like map settings do matter in this case, such as the map size, quality of planets you get, available resources, etc.

The second thing that I noticed is that despite getting free upgrades, fighters cannot be customized. This to me is a noteworthy weakness. This means that once you reach endgame, your fighters are going to have fixed specs. That makes your carriers easier to optimize against, like I did in this experiment. The Altarians had missile specialization (the most troublesome one to face due to its long range), so I was able to tip the odds in my favor by adding more missile defense and less kinetic defenses. The results worked out really well. Obviously in actual gameplay, ships are not going to so uniform like this, so your ship building skills for the situation that you are in does play a large role in your success in battles, without or without carriers.

I believe there was one other thing that I noticed, but I forgot it, so oh well. However, I can use this space to suggest that it would probably be pretty fun to use fighters as escort cannon fodder so your huge ships in the rear can do their job. I do this myself for my actual fleets, but against the AI, it is never necessary.

In any case, this only confirms the previous experiment: Carriers are extremely powerful for their price. This seems to me like their main selling point. Free upgrades are nice, but this only matters mid-game, and in anywhere but late/endgame, it is going to be the techs you have researched that matters the most. I do believe that carriers should be among the most powerful ships that you can have, so they do not seem to need nerfing. My fix for carriers would be to increase their cost to match their power, and to enable multi-ship targeting. It is pretty crappy that your 4 huge ships can only take out 4 fighters at a time despite how powerful they are.