> Bonus question. Can anyone come up with an EMG transformation that
preserves
> the idea that net luck and net skill expressed in EMG, exactly and
> consistently replicate the match outcome (as it should, I contend,
when
> expressed in MWC)?
If luck is computed at the same level than skill (e.g.
both 0ply or both 2ply) and that level of play evaluates luck perfectly,
you have that NetLuck+NetSkill=100% in MWC (or = to the acutal result
in points for money), if both luck and skill are raw, not normalized.
However, even if EMG transformation is linear, it
is so at a given score and cube value. When you sum luck or skill values
from different scores or from different positions with same score but different
cube values, the overall transformation is not linear, sine it
includes the effect of the MET. Hence it is not possible that the overall
sum of EMG values matches the (normalized) actual result.
> The suggestion for reporting "net luck" is interesting as
an addition. But I
> would not like to lose the reporting of raw luck in MWC. You can have
a +20%
> net luck, but it FEELS very different as a player when that is +80%
for you
> and +60% for the opponent versus -40% for you and -60% for the opponent!
We should indeed keep in the luck display normalized
and raw luck. I just proposed to use net luck in MWC/Points (instead
of the "Luck rate EMG") to compute the "Luck rating".
Notice that Net luck (in MWC or Points) is already
used by gnubg in the "Luck adjusted result".