R&D factory repair rate?

[AirGriff]

Would some veteran JFB's throw out some rough idea of how fast R&D factories repair? I know they repair faster as they
get closer to the release date, but I'm trying to figure out what factories I should change based on release dates and
expected repair rates. I know it's somewhat random, but just looking for some experienced estimates.

How fast would, say, a factory of 10 repair with a release date in 6 months, 12 months, 2 years, 3 and 4 years?

Assuming a fairly rapid repair rate for 6 months out, would it be wise to switch factories with distant releases of ac
you're not going to use to an earlier released model, let them repair, and then switch again(of course preferably keep in the
switched ac line)? That seems to me to be an economical way to do it, but maybe I'm missing something?

Any help would be much appreciated before I go button pushing.

[Amoral]

It isn't a linear relationship. Factories more than about 14 months off effectively do not repair, while factories less than 6 months off repair at about .5 / day.

Choosing an early model in the line, repairing the factories, and then upgrading the factories is a very strong tactic. For example, you can switch several factories to a6m2-N and wait for them to repair, and then upgrade the factories to a6m5, and they will stay fully repaired.

[Lokasenna]

You should expand all R&D factories to 30, as total factory size has a lot to do with how quickly the factory will repair. Notice how the 0(55) Ki-84a Frank factory repairs several points before any 0(1) factories do.

I mention size 30 because it is the optimal size for R&D, providing you with a guaranteed 1% of research per day (2% with the engine bonus).

[Spidery]

A factory will, on average, repair from 0 to complete size in roughly 63% of the time remaining before the model being researched is available.

It is roughly "full factory size" / "days remaining to aircraft availability" chance per day.

It is not true that a factory for a long time away will not repair - it repairs slowly. However, the time to total repair is proportional to the total time so it will get to produce R&D points for proportionally longer. So, ignoring randomness and rounding effects, you get the same effectiveness from an early or a late model (but if too soon until available then rounding effects become pronounced).

It is almost always best to develop the factory on the first R&D model of a line and then change to the model you want to produce. E.g. research Ki-44 and change to work on the Ki-44-IIc. As Lokasenna says, a size 30 factory gets optimal value for your R&D factory. You might do something smaller if you plan on rolling into production at less than 30 (e.g., you could have a size 20 factory for the H8K1).

If you use a factory to work on something mid-war (e.g., Ki-84a) and then change it for something late war (e.g., Ki-83) you will get approximately the same amount of months research benefit but at nearly twice the supply cost as were you to just use it for the late war aircraft.

My recommendation, for most of the R&D factories plan on using them only on a single model-line and having them turn into production. This avoids wasting supply changing between models.

Also, keep in mind that research is twice as effective when you can use the engine bonus (but with an HI cost). In a scenario 1 game, it is hard to get the engine bonus for all your research. For example, N1K1-J research may miss out on it whereas J2M2 is easy to get the bonus for.

There is also an increasing randomness in results when researching late models which actually makes them slightly more effective.

[Lokasenna]

That is one the big benefits to Scenario 2 - the extra engine factories. Much easier to get the engine bonus across the board.

The HI cost is nominal. For every additional month you accelerate an aircraft, you're out only 1800 HI. That's less than a single day's surplus. Over the course of all of your R&D, that might cost you an additional 180,000 HI than it would without the bonus, and you'll be that many months (100 total "R&D advancement months" of engine bonus there) early. Very, totally, extremely worth it, if you budget for it.


Edit - on budgeting for it...it helps to monitor the pools for your expensive, but low attrition airframes, like transports and your flying boats. I can't stand low rates of production of 2 or 8 per month, so will bump it to 15-20 in a lot of cases, but I will then keep an eye on the pool levels. I don't want to produce too many planes that cost 72 or 144 HI apiece. For transports, I like to keep enough in the pool to replace one entire Sentai in case something goes awry, like LRCAP over an objective. For the Mavis/Emily pools, it's harder to justify having 18-20 spare, but I like to have double digits...just in case. Judicious use of rest on these units will help keep ops losses lower. All of this helps you be able to afford the R&D cost.

[AirGriff]

Fantastic stuff, thanks guys. So, for a late war RD factory of just 2 or so for a plane I never intend to use I should convert it to an earlier model and expand it to 30 rather than 2 such factories expanded to 15, yes? I guess it might depend on how far out the ac becomes available. Will have to crunch some numbers.

[AirGriff]

Yeah, a mathematician I'm not. It appears you'd be better having one 30 factory than two 15 factories for chances of repair regardless of how many days out. Can't wait for my 2 hour nap between 1 and 2 o'clock today.