Re: [Enigma-devel] Par-Times

[Ralf Westram]

Hi,

>> please send me your .enigmarc files since I'd like to update the par
>> times included in the index files.
> Be very careful to check to make sure that any level times only
> possible to development bugs are not included.

That's a bit difficult - but I'm trying to.

We should definitely also raise the revision number for several
levels, for which a very short par occurs. Such a par also means,
there has been a version out, which has been buggy and therefore was
to easy to solve. Raising the revision will reset the par times which
are generated by my script and it will encourage players to try the
level again.

It would be very useful for me, to get more enigmarc files from all of
you out there. I don't need it for better par times, but I would like
to do some statistics, e.g. calculate the hypothetical probabily for a
par time.
Currently I have 7 score files, which is definitely a bit less to get
meaningful results from such statistical analysis.

> Also we do want to make sure that par is always possible to at least
> meet, so the absolute best time ever performed would be too difficult
> for people to even approach. Do remember that even in golf where par
> does not mean 'average', it does mean approximately 'just a little
> bit above the average of experts'.

Nevertheless - in golf you'll really need years of hard training just
to reach par. Wikipedia says:

| Par is also the theoretical number of strokes that an expert golfer
| should require for playing the ball into any given hole. The expert
| golfer is expected to reach the green in two strokes under par (in
| regulation) and then use two putts to get the ball into the hole.

In other words: The _expert_ golfer is expected to exactly reach par,
if he doesn't make an error.


Regarding short levels, there's a tolerance span of nearly a whole
second - which is much compared with level par times of 2-5 seconds
(1 second is 20-50% of overall level time).

Regarding more continuous levels, our par times won't get so
optimized, since nearly nobody will try to time-optimize a level, for
which you'll need approx. 7 minutes..  or will you? :)


Ralf


-- 
VI VI VI - the editor of the beast.