research

[acrashb]

As much as this has been discussed, I've just gotten high-level bombers as German in 1941. So walking across the USSR is a matter of a)rubble the city and therefore destroy the supply wipeout the units c) rinse and repeat (yes, I'm complaining about too much luck with research )

So, how about:

Rt=Ta+(Ta*Tv*rnd)-Ta*Tv*(SQRT(Ch-1))

Where

Rt=Research Time, weeks

Ta=base Time to research a given technology Area

Ch=number of chits invested

Tv=Time Variability, which I propose to set at .2 (one-fifth) - this is easily tweakable.

rnd=pseudo-random variable ranging from 0.0-1.0 in a normal distribution.

What this does is put some initial variability into the research time, then reduces the time consistently with additional research effort (chits), but with diminishing effect for each new chit (hence the square root). Key: don't tell the player the outcome or the Tv.

So, if a thing takes a base of 52 weeks to research, the median case (where rnd is .5), looks like:

Chits Weeks

1 Chit 57

2 Chits 47

3 Chits 42

4 Chits 39

5 Chits 36

If the player was very unlucky (rnd of 1.0), it would look like:

1 Chit 62

2 Chits 52

3 Chits 48

4 Chits 44

5 Chits 42

If very lucky (rnd of 0.0), it would look like:

1 Chit 52

2 Chits 42

3 Chits 37

4 Chits 34

5 Chits 31

Adjustments to the Tv and rnd can quickly produce desired results that would work in the game. Could also adjust the Sqrt to a cube root to flatten out the investment improvements.

For the tough stuff that should not come quickly, just limit the number of chits allowed in that research area.

Hubert, you're free to use this.

Most research proceeds in a relatively predictable manner: the problems to be solved are understood, and it's a matter of digging into it and allocating resources to do the tests to validate hypothesis. Fundamental research and bleeding-edge may be a different matter (e.g., energy from fusion), but I don't think that generally applied in WWII.

So some predictability is in order for SC3 or an SC2 patch.

[John DiFool the 2nd]

Amen to that (as I've said before ad nauseum). And put the progress percentages on the tech page too (0-100), so I can see how close I am. My latest game with Al's mod had the Germans with 2 chits in Naval Warfare (I was building up the Plan Z fleet) for over two years, and not one hit. The chit lottery has got to go! No R & D company in the world, past or present, would live with (or even survive) such a random system.

[Hubert Cater]

Thanks Acrashb and John, some really good ideas here and we are discussing this a bit more internally so there might be some changes to research down the road

[acrashb]

That's good news!