Fighter factories vs shipyards

[nakorkren (OP)]

Has anyone looked at the math behind building fighters in shipyards, rather than fighter factories?

I started my analysis on the assumption that you have a 1000 ton, 20 slipway shipyard. At an equivalent invest in research of 20,000RP, you get 1000 BP per slipway * an efficiency of 0.6 for a 1000 ton slipway (BP for a Naval shipyard is = 1+(((tonnage/5000)-1)/2)) or 20BP per fighter factory. Hence your 20 slipways produce 600 BP each for a total of 12,000 BP, which divided by 20BP per fighter factory means you need 600 fighter factories to achieve the same BP.

600 fighter factories requires 50k*1000=30M pop to work them. One 1000ton shipyard with 20 slipways requires 1M+100*(total tonnage of 20,000)=3M pop to run the shipyard. That's a HUGE benefit early game when pop tends to be a limiting factor
600 fighter factories requires 120 Vendarite each for a total of 72k Vendarite. One 1000ton shipyard costs 1200 Duranium and 1200 Neutronium, plus as far as I can tell about 100 D and 100 N for each slipway, for a total of 3200 D and 3200 N.

So by the numbers, a 1000ton shipyard with 20 slipways is strictly superior in population cost and mineral cost to the number of fighter factories required to achieve an equivalent BP, at the same research level. Not to mention that shipyard BP per research scales up faster than fighter factory BP per research. The ratio of Shipyard BP to fighter factory goes from 40 at start to 80 at the end of the research chain. Further, that 1000ton shipyard can also be used to build FACs for boarding or other purposes, and if you avoid fighter factories all together in favor of shipyards, you can skip researching Fighter Factory BP tech, whereas you will always need the Shipyard BP tech either way.

My conclusion is that there is no reason to build fighter factories, ever. Did I make a math error somewhere, or a bad assumption, that makes fighter factories more useful than they appear?

Update: I originally neglected the fact that the build rate is standardized on a shipyard of 5000 tons. A 1000 ton slipway, the smallest you can build a new shipyard, only gets 0.6 times the BP.

[nuclearslurpee]

The huge, huge benefit to using fighter factories is that you never have to retool them. With fighter designs being as tight as they are, changing a single component usually eliminates the possibility for cross-building, so every time you want to build a new fighter class (or more of an old one) you must retool. Retooling 20 slipways constantly is expensive and time-consuming. Fighter factories require you to select a different class from the list and it works. You can work around this to a degree by having several shipyards with 5-10 slipways each at 1,000 tons, but then your initial mineral expenditure increases considerably.

I'm assuming your numbers are right, they look in the ballpark and I don't want to dig in the DB for tech values tonight. However, I feel obligated to note that Not All Minerals are Created Equal™. 72k Vendarite may be a lot compared to 3.2k Duranium and Neutronium each (discounting the cost of retooling), but Vendarite is usually not a mineral one has a shortage of even when running the ground force construction facilities at 100%. 72k is an amount one might mine incidentally from one or two large asteroids, comets, or mineral-rich moons given time.

[Vandermeer]

The huge, huge benefit to using fighter factories is that you never have to retool them. With fighter designs being as tight as they are, changing a single component usually eliminates the possibility for cross-building, so every time you want to build a new fighter class (or more of an old one) you must retool. Retooling 20 slipways constantly is expensive and time-consuming. Fighter factories require you to select a different class from the list and it works. You can work around this to a degree by having several shipyards with 5-10 slipways each at 1,000 tons, but then your initial mineral expenditure increases considerably.

On the other hand, fighters being built in shipyards gives you the also pretty huge advantage of being able to refit them to newer designs. This is especially useful in C# where there is currently no easy way to train them other than passive fleet training experience, so you would really want to preserve what you can get.

I have also been thinking about replacing the fighter factory approach to a hybrid version, where the most common combat design gets its own shipyard/s, while all the extra utility fighters, sensor drones, satellites and general non-combatants can still be built in (much fewer) factories.

One minor thing nakorkren neglected in the analysis is however also the setup time. Though building 600 fighter factories might or might not take some time too, -depending on available construction capacity ofc-, there is no real way to speed up getting 20 slipways on a shipyard. (research is x2 at maximum) I can't check right now, but it seems like a task that can take you 20 years on its own to do, which is too long unless you invest in said shipyard buildup some tech-generations before actually designing the first plane already.
It is rewarding in the end, but many games can be quite time sensitive, especially when you actually depend on the fighter formula. (="I need some extremely high efficiency combat ships in short order")

[Jorgen_CAB]

You probably would never rely on Shipyards for producing fighters only. Having a singe 1000t yard with 20 slipways make it very restrictive in terms of building different kinds of designs. In general I often have two or three yards at 1000t with five to ten slipways. Their role is to build fighters and FAC including upgrading fighters occasionally or scrap them.

I also guess it depends on how heavily you are into fighters and how many models of fighters or small crafts that you have.

You also don't get 600BP when building something smaller than 1000t in said yards. A 250t fighter only benefits from around 525BP a year for example. Not a huge deal but still the size of the thing you produce decide the final build rating.

In general when it come to population I quite rarely find it that restrictive in terms of military construction capacity, most population that I need is for other purposes. I also think that you need to consider the minerals used as equally important overall. As said above, Vendarite are usually much less of an issue than Neutronium. Once population start to become an issue I might consider reducing my fighter factories for more slipways to free up population for more construction factories, but that is more of a strategic decision at that point. I rarely end up with population limits at game start but rather at mid game or later. In early game then I want my population to work as that give me more wealth income.

[Black]

I sometimes use shipyards to refit fighters, but I always modify them with SM to fit the size of the design, so if I have 250 ton fighters, I change the 1000 tons shipyard to 4 slipways of 250 tons.

[misanthropope]

i have a strong preference for shipyards; it's the factory effort required that puts me off building fighter factories.  it's not really a 4x game, but it IS a 4x game *simulator*, and managing your factories would be a big thing, if managing your resources were a big thing.

usually 90% plus of my fighters will either be of a single combat design or a single scout design, and building them in dedicated shipyards you don't really miss the flexibility of fighter factories.  i have found that my starting endowment of fighter factories plus the FF i dig up or capture seems to always be sufficient for the niche fighter designs i come up with.

[xenoscepter]

 --- I haven't checked my math on this in a long time, but I sat down at one point and figured out which was better and how. I don't recall what my conclusion was on a mineral / BP /wealth basis, but I remember clearly that 4,000 Tons of Fighter Production was the point where Shipyards and Fighter Factories broke even in terms of Population costs. So anything less than 4,000 Tons of Shipyard is going to be more Population efficient than Fighter Factories, while anything over that is less.

 --- That said, the math was old, and probably very flawed. Case in point; in the matter of Shipyards vs. Factories, both use a different tech to increase output. Having a few 500 Ton, 2 slipway "Fighter Yards" can produce multiple designs concurrently and at full capacity, whereas Fighter Factories need to either que up and build batches one after another or else split capacity to build multiple batches at once. Adding slipways adds more parallel construction, while the ability to Use Components gives you even more speed if needs be.

 --- The savings in tech might well be worth it, while the build speed is of dubious utility. Refit can be quite nice though... a missile fighter can have it's M-FCS and sensors package replaced and extend it's useful life on the cheaply.

[skoormit]

This analysis is very thought-provoking.
I had not crunched the numbers before, and I am surprised at just how strong the economic advantages are for shipyard-based fighter production.

At first, the retooling time seems like the biggest downside.
But I'm not sure it's actually all that significant.
Retooling a 20-slip, 1kt yard for a design change involving a single component on a fighter can't take that long, can it?
Even if it's the most expensive component on the ship.
Heck, even a major redesign is probably never going to take a month.

Saving 27M workers (not population; actual workers--very significant difference) in exchange for 3.1k DUR/NEU plus ongoing occasional minor retooling costs seems like a trade I would like to make--at least when I reach the point in the game where labor is the bottleneck to expanding production.

Another advantage not mentioned: shipyard expansion occurs independently of your factory production.
The initial construction represents about 40% of the total cost for the 20-slip yard.
The other 60% does not require use of your construction factories.
Fighter factory production, of course, requires confacs for all of it.

Minor note:
Using yards also saves 72k VEN and ~69k wealth (plus occasional minor wealth costs when retooling).
I consider the value of saving 72k VEN to be essentially nil. I always have more than I need.
If your economic experience in game is anything like mine, saving ~69k wealth over the length of time involved in establishing this much fighter production capacity is also essentially nil. Thanks to taxes on civilian shipping, my economy always runs at a surplus (after I have a second non-trivial colony established), which means my wealth is almost always at the 2x racial income cap. I could probably spend this "extra" 69k (spread out over the production time) without ever running an annual budget in the red. Even if it did manage to dip into my savings, my treasury would recover to the cap very shortly after production was finished.

[Steve Walmsley]

I am running a fighter heavy campaign at the moment. I have 620 fighter factories, but also a shipyard with twelve slipways producing fighters.

[skoormit]

Update: I originally neglected the fact that the build rate is standardized on a shipyard of 5000 tons. A 1000 ton slipway, the smallest you can build a new shipyard, only gets 0.6 times the BP.

Minor correction: the build rate depends on the size of the ship being built, not the size of the yard.
A 500t ship gets a 0.55 multiplier.
A 250t ship gets a 0.525 multiplier.
A theoretical 0t ship gets a 0.5 multiplier.

[skoormit]

One 1000ton shipyard with 20 slipways requires 1M+100*(total tonnage of 20,000)=3M pop to run the shipyard.

Another minor correction:

This is the old VB6 logic.
In C#, it's just 250 workers per ton for naval yards.

[nuclearslurpee]

One 1000ton shipyard with 20 slipways requires 1M+100*(total tonnage of 20,000)=3M pop to run the shipyard.

Another minor correction:

This is the old VB6 logic.
In C#, it's just 250 workers per ton for naval yards.

For those curious, this means a 20x1,000 ton shipyard requires 5M workers instead of 3M given in the OP, so the general point in the OP about shipyards being more population-efficient still holds.

[Jorgen_CAB]

One thing that might or might not be important is the cost of retooling. To retool from one 250t design to a completely different 250t design could have a cost of roughly 5 times the produced crafts. The time for retooling usually is less than a month so not that long... in my experience.

[skoormit]

One thing that might or might not be important is the cost of retooling. To retool from one 250t design to a completely different 250t design could have a cost of roughly 5 times the produced crafts. The time for retooling usually is less than a month so not that long... in my experience.

Is the retooling cost significant at all?

How many fighters does a 20-slip yard produce in a year?
Depends on tech level, cost of the fighter, governor/sector shipbuilding bonus, etc. 
Ballpark--5 per slipway per year? That feels like it's on the low end of things.
So, for a 20-slip yard to pay a retool cost equal to 5 fighters, if you retool every year, represents at most a 5% annual tax on doing business via shipyards.

And that's if you retool to a completely new design.
More likely you are just changing out one or two components at a time.
Which means this tax is more like 1% of the annual shipbuilding cost for this yard.

[Agraelgrimm]

Is the retooling cost significant at all?

I think the bigger cost here is the time you are going to spent retooling for a different class of fighter. And since fighters are usually made on demand, is better to not have to wait months or a year for the retooling to complete. Especially when running with full tech values. By the time you are being able to finish them all they may already be outdated.