Hey folks :)
First let me say that I'm fully aware of Adalbert's engineer formulas. In fact thats where
all the problems started: my intention was (and is) to calculate the point where my
overspeed would increase by 1 knot while base speed stays the same.

So I played around with Adalbert's formulas and soon got to the point where I noticed that
their must be either a mistake or some kind of a "misunderstanding" because Adalbert's
formulas can't explain the phenomenon of losing 2 knots of overheat when losing 1 knot of
base speed.

1. Example:
Ship RN Dunkerque
Displacement = 21779
CV IV Heavy Engine
2 Engis on board, combined true ability: 2751948
Base Speed = 27
Overheat Speed = 39

2. Example (same as 1. example but only one ton more displacement):
Ship RN Dunkerque
Displacement = 21780
CV IV Heavy Engine
2 Engis on board, combined true ability: 2751948
Base Speed = 26
Overheat Speed = 37

One ton more displacement, loss of one knot base speed and two knots overheat speed.

Even with all the rounding Adalbert describes in his formula posts, you cant reproduce
this with calculations (even if you knew the c value for Dunkerque).

So I started to do some own formula work, which lead me to two conclusions:
1.: you always calculate with a rounded down base speed (or even easier: just use the base
speed you actually got ingame...and dont calculate base speed at all)
2.: "It still remains, that you have to round down your result and also only use two
'after decimal digits' of (OHR_Ship * OHR_engine)."
This seems to be the main mistake in Adalbert's formulas. In fact it seems like you only
use two after decimal digits for the whole overheat ratio term (including the engineers
ability), not only for OHR_Ship * OHR_engine.

I can't prove these conclusions yet. My engis need to level up twice then I got hard
evidence :P. But with these two additional assumptions the calculations seem to work even
for the loss of two knots over speed while losing one knot base speed.

While I try to level my engis: any1 else able to confirm this...or prove me wrong?