1. Scoring: 15(j) and incorrect submissions
Should rule 15 (j) be changed to exclude incorrect submissions from
the speed bonus?
Or should I submit 10 bogus (but fast) programs so no-one effectively
gets any points for performance? 
)
Just asking,
Pete
2. Re: Scoring: 15(j) and incorrect submissions
Pete Lomax wrote:
>
> Should rule 15 (j) be changed to exclude incorrect submissions from
> the speed bonus?
>
> Or should I submit 10 bogus (but fast) programs so no-one effectively
> gets any points for performance? )
>
In the final placing, I think all submissions by the same person should be
considered an "entry" so that the same person can't win first & second place (for
instance). This would affect the ranking points given for speed as well -- maybe
only a person's current best submission should count for points in the list to
keep it from getting too confusing?
3. Re: Scoring: 15(j) and incorrect submissions
Andy Serpa wrote:
>
> Pete Lomax wrote:
> >
> > Should rule 15 (j) be changed to exclude incorrect submissions from
> > the speed bonus?
> >
> > Or should I submit 10 bogus (but fast) programs so no-one effectively
> > gets any points for performance? )
> >
>
> In the final placing, I think all submissions by the same person should be
> considered an
> "entry" so that the same person can't win first & second place (for instance).
> This
> would affect the ranking points given for speed as well -- maybe only a
> person's current
> best submission should count for points in the list to keep it from getting
> too confusing?
Both are good points. As for the first one, here is a proposed addendum
to the rules ...
---
Note: Any program that does not make a valid attempt to find and
accumulate token statistics, will be awarded a zero speed score. This is
to prevent spurious but fast programs being submitted.
---
As for the second point, would it sufficent to organise the rankings
so that any given person can only have one ranking at anyone time? And
that would be for their best effort so far.
I'll give this a day or two to get some feedback before I implement them.
--
Derek Parnell
Melbourne, Australia
4. Re: Scoring: 15(j) and incorrect submissions
Derek Parnell wrote:
>
> ...would it sufficent to organise the rankings
> so that any given person can only have one ranking at anyone time? And
> that would be for their best effort so far.
Yes.
-=ck
"Programming in a state of EUPHORIA."
http://www.cklester.com/euphoria/
5. Re: Scoring: 15(j) and incorrect submissions
> Derek Parnell wrote:
> > ...would it sufficent to organise the rankings
> > so that any given person can only have one ranking at anyone time? And
> > that would be for their best effort so far.
I think that's fair. It's more than likely that each person's
subsequent entry is an optimisation of the last, with slightly
different approaches taken to components of the task.
--
MrTrick
6. Re: Scoring: 15(j) and incorrect submissions
Derek Parnell wrote:
<snip>
> Both are good points. As for the first one, here is a proposed addendum
> to the rules ...
>
> ---
> Note: Any program that does not make a valid attempt to find and
> accumulate token statistics, will be awarded a zero speed score. This is
> to prevent spurious but fast programs being submitted.
> ---
>
> As for the second point, would it sufficent to organise the rankings
> so that any given person can only have one ranking at anyone time? And
> that would be for their best effort so far.
>
>
> I'll give this a day or two to get some feedback before I implement them.
I think that addendum is a good improvement of the rules.
Regards,
Juergen
7. Re: Scoring: 15(j) and incorrect submissions
Derek Parnell wrote:
> Note: Any program that does not make a valid attempt to find and
> accumulate token statistics, will be awarded a zero speed score. This is
> to prevent spurious but fast programs being submitted.
> ---
>
> As for the second point, would it sufficent to organise the rankings
> so that any given person can only have one ranking at anyone time? And
> that would be for their best effort so far.
>
>
> I'll give this a day or two to get some feedback before I implement them.
>
> --
> Derek Parnell
> Melbourne, Australia
It sounds ok, Derek, but it may be easier, sounder and safer, if the speed bonus
was available only to those entries that get all 350 points for the first five
test files.
jiri
8. Re: Scoring: 15(j) and incorrect submissions
jiri babor wrote:
>
> It sounds ok, Derek, but it may be easier, sounder and safer, if the speed
> bonus was
> available only to those entries that get all 350 points for the first five
> test files.
>
Sounds good. Be correct first, then be fast...
9. Re: Scoring: 15(j) and incorrect submissions
- Posted by jiri babor <jbabor at paradise.net.nz>
Nov 07, 2004
-
Last edited Nov 08, 2004
I wrote:
>
> Derek Parnell wrote:
>
> > Note: Any program that does not make a valid attempt to find and
> > accumulate token statistics, will be awarded a zero speed score. This is
> > to prevent spurious but fast programs being submitted.
> > ---
> >
> > As for the second point, would it sufficent to organise the rankings
> > so that any given person can only have one ranking at anyone time? And
> > that would be for their best effort so far.
> >
> >
> > I'll give this a day or two to get some feedback before I implement them.
> >
> > --
> > Derek Parnell
> > Melbourne, Australia
>
> It sounds ok, Derek, but it may be easier, sounder and safer, if the speed
> bonus was
> available only to those entries that get all 350 points for the first five
> test files.
>
> jiri
I really meant all 300 points from the first six columns for the first five test
files. The seventh column, 'Faster than Benchmark', is patently silly anyway.
jiri
