# Research aux hubs: what do they buy you?

I don’t have hard numbers. Hoping someone can help clarify, but this is where a stop watch and bad math has me so far.

Research ship 1 = total time
Research ship 2 = - 18% or 82% of total time
Research ship 3 = - 15% or 85% of remaining time (85%(82%*Time))
Research ship 4 = - 10% or 90% of remaining time (90%(85%(82%*Time)))
Research ship 5 = - 8.5% or etc.etc.
Research ship 6 = - 8% or etc.etc.

Here was my stop watch times I used to try and figure out the percentages.

Research 1 - Fighter = 25 seconds
Research 2 - Fighter = 20 seconds
Research 3 - Fighter = 17.5 seconds
Research 4 - Fighter = 15.75 seconds
Research 5 - Fighter = 14.5 seconds
Research 6 - Fighter = 13.5 seconds

R1 Corvette = 57 seconds
R2 Corvette = 47 seconds ?
R3 Corvette = 40 seconds
R4 Corvette = 36 seconds
R5 Corvette = 33 seconds
R6 Corvette = 30.5 seconds

R1 Cap = 1:25
R2 Cap = 1:10
R3 Cap = 1:00
R4 Cap = 0:54
R5 Cap = 0:49.5

@innociv - any insight from the data files Ogilvie?

A quick setup in Excel:

The most appropriate trendline I found was using a power series. The equation can be modeled as such:

Actual Research time = (Base research time) * (number of research ships ^ -0.343)

Using this equation, if the base research time for Corvette Drive is 57 seconds, then it takes approximately 31 seconds with 6 research ships. Fairly precise according to your recorded time of 30.5 seconds.

In HW1 classic, the effect of each additional research ship was greater but certain research tech took longer (Fighter Drive base time was 75 seconds instead of 25).

• Each research ship out of the total 6 you can build seem to be counted as separate ships by the game. Hence the names “Auxiliary research hub X”. This means the player can only build one research ship at a time as each research ship is unique and has a unit cap of 1.
• Building them from a mothership makes them come out of the capital ship hangar bay. Building them from a carrier makes them hyperspace in. There is no research lag or delay for either of these construction methods - the instant the build bar reaches 100% that research ship’s effects on research timings are immediately present.
• Whether or not multiple research ships are connected to each other is irrelevant to their effects on research timings.
1 Like

Interesting.
I heard this elsewhere, which this confirms isn’t right (as expected based on known timings of things).

Not sure where they got that formula from, as it’s not listed in tuning.lua. Perhaps they work for GBX, but that formula seems wrong.

See, by that formula it means that 1 research ship would be 29% faster than the listed amount. Which almost sounded acceptable, though by my math still put HW1 at a disadvantage.

But by your numbers, well I know the exponent used IS 0.5. (And the exponent is something that can be changed. Formula can’t)

So first off, that +1 in the formula is obviously wrong since base time is base time.

It appears that the actual formula is something like

t(new)=t/(n^(exponent/1.5))

which seems quite bad. Not much point in getting more than 1 or 2 additional research.

Explains why it takes around 15 minutes to get an HC out. With the other formula it’d be 10.5, similar to BCs.
Though, I’m not crazy about the other formula with that exponent either because it makes the listed time not the correct time with just 1 research vessel.

1 Like

Looks like the optimal number is 3, anything else is kinda of a waste. Very usefull info.

lol you math freaks

1 Like

Maybe research for HW1 races should be cheaper yet the payoff for multiple research hubs increased…

Yes, it certainly makes for more need for thought around when to build additional research hubs. The second one seems like a no brainer early game before your production is going given you have the available tracks, and having too many early could hurt your production due to increased spend on RU (researching faster) and the cost of the extra hub. The problem is if you wait until it’s effects are more useful say for destroyer research which could give you an extra 30 seconds back, you are tieing up a production line for over a minute when you probably need it most.

Much planning required. I like it

I found that the research.lua file for Kushan/Taiidan has larger base times for each tech for 0 research ships so I adjusted your really precise equation to take this into account.

The new equation is:

Where n is the number of research ships.

Results look pretty consistent with Ratamaq’s recorded times.

5 Likes

Yeah, I didn’t read into the numbers enough to realize what was listed as “total time” was with one research vessel time, so is much lower than what is listed in the files.

The formula I came up with is what you’d have in order to have the base time be what was in the file. (IE, 25s for fighter drive in this case, when really it’s 35 that gets reduced 29% from the first lab)

Would have been obvious if I looked up the actual base times in files. Was dumb of me not to.