There was already discussion about scoring of DNS, DSQ, DNC and so for high point systems previously.
It was clear that the default settings were not working properly, and codes must be defined manually.
High point systems are used for series lasting longer than a regatta.
In this case,
DNC scores normally zero
Points are given according to the number of starters S
code for DNF does not score properly with score for S+1. Often scores better with score for S
other codes (OCS, DSQ, DNS) should score as DNF (default)
As long as the defaukts for these systems are not comonly defined and automaticaly adjusted when one selects a scoring system, the procedure should be included in a help file…
I’ve decided that p >s in RInderle and p > s+1 in CHIPS are configuration error conditions and scoring is aborted with a message to sort out the codes - and some tips on how to do it… Silently tweaking p is ‘dangerous’ in a way and should be avoided…
PS: What is the location of your latest paper(s) on CHIPS etc - I’d like to link to them form the hewlp file if that’s OK.
COlin
Sailwave
Geoff Burrell wrote:
Just to clarify, the Rinderle B figures are derived from a look-up table whereas CHIPS is derived from a formula with no scope for going negative. A few years ago I devised a formula designed to be a reasonable fit to Rinderle B, or at least to behave similarly to it, but because the increments are a bit irregular it was not easy to get a good fit. That was one of the several reasons why I devised the CHIPS formula and, having done so, there was no longer any great merit in persevering to get a better equation for Rinderle B. I beleive at least one club in the US is using my "modified Rinderle B" formula which I recall putting in the Sailwave Files area if it is really needed.
Colin - now that I have seen the logic, as well as the formula I took
the time to read up more on this. It seems that for assigned points
situations (esp things like DNC / DNS and DNF) there is a possibility
for a negative calculation. I also noticed that some implementations
seem to assign fixed number of points (really low) instead of the
calculations. Given that Rinderle B (and CHIPS I suspect) were
originally designed as table look-up I suppose it is a potential that
this (negative number) situation can occur.
I can't find any pattern for assigning points for scoring codes.
Is there some way to make sure that the result is always positive?
In my series results that I was using to test (nothing like real data
huh?) it looks like the correct number of points were being assigned,
but sometimes a negative number.
Does this help? (or hinder?)
Bill
---
No virus found in this incoming message.
Checked by AVG - [http://www.avg.com](http://www.avg.com)
Version: 8.0.175 / Virus Database: 270.8.2/1738 - Release Date: 21/10/2008 14:10
Good idea to abort when such a configuration erro might occur.
I have just checked and my paper “All about CHIPS” is already in the Saiwave Files area, dated 2nd April 2007. This is a detailed paper that includes the rationale as to how CHIPS was derived, together with information on Rinderle B as well as my “modified Rinderle B” formula.
I’ve decided that p >s in RInderle and p > s+1 in CHIPS are configuration error conditions and scoring is aborted with a message to sort out the codes - and some tips on how to do it… Silently tweaking p is ‘dangerous’ in a way and should be avoided…
PS: What is the location of your latest paper(s) on CHIPS etc - I’d like to link to them form the hewlp file if that’s OK.
COlin
Sailwave
Geoff Burrell wrote:
Just to clarify, the Rinderle B figures are derived from a look-up table whereas CHIPS is derived from a formula with no scope for going negative. A few years ago I devised a formula designed to be a reasonable fit to Rinderle B, or at least to behave similarly to it, but because the increments are a bit irregular it was not easy to get a good fit. That was one of the several reasons why I devised the CHIPS formula and, having done so, there was no longer any great merit in persevering to get a better equation for Rinderle B. I beleive at least one club in the US is using my "modified Rinderle B" formula which I recall putting in the Sailwave Files area if it is really needed.
Colin - now that I have seen the logic, as well as the formula I took
the time to read up more on this. It seems that for assigned points
situations (esp things like DNC / DNS and DNF) there is a possibility
for a negative calculation. I also noticed that some implementations
seem to assign fixed number of points (really low) instead of the
calculations. Given that Rinderle B (and CHIPS I suspect) were
originally designed as table look-up I suppose it is a potential that
this (negative number) situation can occur.
I can't find any pattern for assigning points for scoring codes.
Is there some way to make sure that the result is always positive?
In my series results that I was using to test (nothing like real data
huh?) it looks like the correct number of points were being assigned,
but sometimes a negative number.
Does this help? (or hinder?)
Bill
---
No virus found in this incoming message.
Checked by AVG - [http://www.avg.com](http://www.avg.com)
Version: 8.0.175 / Virus Database: 270.8.2/1738 - Release Date: 21/10/2008 14:10
Yes, thanks Phil. I’ve detected -ve points for
CHIPS/Rinderle and explained that in all likelyhood the user has codes
set up for low point. I’ve always tweaked the default high point codes
(selected with the hand) so they work with both - i.e. based on
starters.
CJ
Philippe DE TROY wrote:
···
sailwave@yahoogroups.commailto:sailwave@yahoogroups.comOn
Behalf Of Sent:
**To:**sailwave@yahoogroups.com Subject:
=Version 2.00 build 0.4 uploaded with checks for -ve points when using
CHIPS/RInderle and a tweaked set of default of high point codes based
on Starters. This slightly compromises CHIPS at present - I cant
automate the code stuff because users may have tweaked them prior to
setting the points system and at present I cant detect a default set -
I’ll do it - it’s just not pri 1 at the moment…
Colin
Geoff Burrell wrote:
···
http://www.sailwave.com/binary/sw200b04.exe
sailwave@yahoogroups.commailto:sailwave@yahoogroups.comOn Behalf Of Sent: To:sailwave@yahoogroups.com Subject:
sailwave@yahoogroups.commailto:sailwave@yahoogroups.comOn
Behalf Of Sent:
**To:**sailwave@yahoogroups.com Subject:
Thanks Tony. I’ve fixed thye scrolling (not
uploade dyet). Notes the window posiiton anomaly (its a generic thing
not just del comp) and the pick list seems fine here - its probably
because you have two versioins fighting eachother for it somehow…
First, thank you for Build 4 - yes it catches the situation nicely and
puts up a warning.
Second, the discussion certainly helped my understand better how I had
to set up SW. My thinking was that it should be variable - getting the
points for doing notyhing or having a sliding scale that changes week
by week. However, working through the logic as I now understand it the
methods now being used (DNC = 0; DNF, RAF etc = Race Starters + 0)
works nicely. Only one change, and it's not for silverware so it is
manageable.