Page 1 of 1

build time vs. number of engineers

Posted: Wed May 03, 2017 10:55 am
by rift
I wanted to research Ammo Conservation (Foundry), which according to Ufopedia has a default build time of 14 days given 60 engineers. I have 28 engineers (almost half), so I expected the build time to be maybe 2-3 times longer than the default, but instead I get a whopping 120 days, which is 9 times longer. Clearly, the formula used in calculating build time vs. engineers is not the same as research time vs. scientists which is a linear relationship, e.g. half the scientists double the research time.

Beta 10 changelog just states:

"There are no engineer minimums to build items; instead, having too few relative to the item requirement significantly increases the time to build."

Does anyone know the exact relationship between engineers vs. build time?

Re: build time vs. number of engineers

Posted: Thu May 04, 2017 1:04 am
by EmptyPepsiCan
I don't know, and this isn't a terribly active forum. Hell, it might just be the two of us. With that in mind, I did a bit of googling. Click here for some code, numbers, and speculation.

Re: build time vs. number of engineers

Posted: Thu May 04, 2017 3:47 pm
by szmind
Oh man, 28 engineers is so few that you should not be eligible to research it that early :) You seem to have rushed certain projects. Well, congratulations, but you must wait. As for the formula - I doubt we can dig it out.

Re: build time vs. number of engineers

Posted: Sat May 06, 2017 9:59 am
by rift
Actually, the link provided by EmptyPepsiCan has some useful info. Based on Amineri's post, here is an approximate formula:

build_time = base_time * 0.5 .* (1 + (OptimalEngineers ./ ActualEngineers) ^ (0.06 * OptimalEngineers));

So, item build time (or foundry project duration) varies exponentially with number of engineers.

In my case of Ammo conservation with
ActualEngineers = 28
OptimalEngineers = 60
base_time = 14

the result is 115 days which is a little short of the actual 120 days I got in the game, but it's still a good approximation. The closer the number of engineers is to the optimal number, the more accurate the estimate becomes from this formula. Most Foundry projects have lower engineer requirement so this formula predicts build time pretty well (in most cases it may be off by 1 or 2 days).

The other thing to keep in mind is that even though I started the project with only 28 out of 60 engineers and 120 days to complete, simply adding 1-2 engineers here and there shortens the project duration significantly, because the relationship #engineers vs. build time is exponential. For example, just going from 28 to 29 engineers reduced the project duration by about 15 days! In a very short time the initial 120 days was cut in half.

So, rushing some essential Foundry projects is definitely worth it, e.g. Ammo Con, Alien grenades, Improved Medkit, etc.. It doesn't take nearly as long as it shows initially, given that you keep adding engineers, but also you can develop multiple Foundry projects simultaneously unlike research projects.