I had a situation where two boats with the same PHRF handicap finished one second apart and were shown as tied for first place.
The correction formula is CF = 566.431 / (401.431 + PHRF). These two boats had handicaps of 189, so CF = 0.959352.
Corrected times (in seconds) were 3409.537... and 3410.49636.... respectively. Still roughly a second apart,
but if you round to the nearest significant digit (the whole second) they end up the same.
If that's the explanation, how about adding a check for an elapsed time difference when corrected time is the same for two boats
with the same handicap?
This problem was previously raised (last year) by Mark Townsend - see attached email.
The reply by Art Engel explaining the situation is also attached.
regards,
Malcolm Osborne
Sedgefield, South Africa
Re_ [sailwave] Problem with Corrected Time Precision.eml (14.2 KB)
[sailwave] Problem with Corrected Time Precision.eml (97.7 KB)
···
----- Original Message -----
From:
jhuus
Sent: Friday, July 15, 2011 7:33 PM
Subject: [sailwave] Boats one-second apart incorrectly shown as tied
I had a situation where two boats with the same PHRF handicap finished one second apart and were shown as tied for first place.
The correction formula is CF = 566.431 / (401.431 + PHRF). These two boats had handicaps of 189, so CF = 0.959352.
Corrected times (in seconds) were 3409.537… and 3410.49636… respectively. Still roughly a second apart,
but if you round to the nearest significant digit (the whole second) they end up the same.
If that’s the explanation, how about adding a check for an elapsed time difference when corrected time is the same for two boats
with the same handicap?
In my humble opinion this is not a sufficient answer. The RRS say the corrected time should determine the position but are silent on the precision of the corrected time. However, the definition of finishing clearly anticipates finishes determined by a fraction of a boat length, i.e. well under a second. Therefore any scoring system must be able to differentiate such finishes and not simply declare a tie.
Where elapsed times are measured to a second, it is clearly impractical to declare a result between two boats of different handicaps where the corrected times differ by a fraction of a second. That is a timing limitation, and a handicapping limitation (handicaps do not differentiate to such accuracy), not a scoring system feature. However, much handicap racing includes significant numbers of same-class boats and their results must be accurately represented. It is simple to resolve as little as a one foot difference in finishing position, so the system should record this.
Simply recording finishing boats in order, as would be done in class racing, should be sufficient.
Ian
···
On 17 July 2011 18:17, Malcolm Osborne malcolmo@telkomsa.net wrote:
This problem was previously raised (last year) by Mark Townsend - see attached email.
The reply by Art Engel explaining the situation is also attached.
regards,
Malcolm Osborne
Sedgefield, South Africa
----- Original Message -----
From:
jhuus
Sent: Friday, July 15, 2011 7:33 PM
Subject: [sailwave] Boats one-second apart incorrectly shown as tied
I had a situation where two boats with the same PHRF handicap finished one second apart and were shown as tied for first place.
The correction formula is CF = 566.431 / (401.431 + PHRF). These two boats had handicaps of 189, so CF = 0.959352.
Corrected times (in seconds) were 3409.537… and 3410.49636… respectively. Still roughly a second apart,
but if you round to the nearest significant digit (the whole second) they end up the same.
If that’s the explanation, how about adding a check for an elapsed time difference when corrected time is the same for two boats
with the same handicap?
I had ths problem once with a TMF system. It can happen when TMF’s are lower than 1.000.
It is just nonsense to change accuracy to 1/10 of a second of even 1/100 (prescriptions of the French Sailing Federation) as you can introduce a difference of ranking for boats that should be tied or even rank them in the reverse order. And this is worse than tying boats that shouldn’t.
The protest about the case I’ve known was solved by the Protest committe by adding anoter second to the elapsed time of the slowest boat, after having controlled that this was not creating another undu tye.
Phil
···
----- Original Message -----
From:
Malcolm Osborne
Sent: Sunday, July 17, 2011 7:17 PM
Subject: Re: [sailwave] Boats one-second apart incorrectly shown as tied
This problem was previously raised (last year) by Mark Townsend - see attached email.
The reply by Art Engel explaining the situation is also attached.
regards,
Malcolm Osborne
Sedgefield, South Africa
----- Original Message -----
From:
jhuus
Sent: Friday, July 15, 2011 7:33 PM
Subject: [sailwave] Boats one-second apart incorrectly shown as tied
I had a situation where two boats with the same PHRF handicap finished one second apart and were shown as tied for first place. The correction formula is CF = 566.431 / (401.431 + PHRF). These two boats had handicaps of 189, so CF = 0.959352. Corrected times (in seconds) were 3409.537... and 3410.49636.... respectively. Still roughly a second apart, but if you round to the nearest significant digit (the whole second) they end up the same. If that's the explanation, how about adding a check for an elapsed time difference when corrected time is the same for two boats with the same handicap?
Long ago before computers, there used to be tables for calculating the corrected time based on the number of minutes sailed, the seconds were just added, without adjustment, to the total.
That would overcome this issue but no doubt upset others not benefiting from the adjustment of the number of seconds.
Times change and boundaries move.
William Carruthers
(_(__(_
/)/)/)
···
Sent from my iPhone
On 18 Jul 2011, at 06:46, “Philippe DE TROY” philippe@detroy.org wrote:
I had ths problem once with a TMF system. It can happen when TMF’s are lower than 1.000.
It is just nonsense to change accuracy to 1/10 of a second of even 1/100 (prescriptions of the French Sailing Federation) as you can introduce a difference of ranking for boats that should be tied or even rank them in the reverse order. And this is worse than tying boats that shouldn’t.
The protest about the case I’ve known was solved by the Protest committe by adding anoter second to the elapsed time of the slowest boat, after having controlled that this was not creating another undu tye.
Phil
----- Original Message -----
From:
Malcolm Osborne
Sent: Sunday, July 17, 2011 7:17 PM
Subject: Re: [sailwave] Boats one-second apart incorrectly shown as tied
This problem was previously raised (last year) by Mark Townsend - see attached email.
The reply by Art Engel explaining the situation is also attached.
regards,
Malcolm Osborne
Sedgefield, South Africa
----- Original Message -----
From:
jhuus
Sent: Friday, July 15, 2011 7:33 PM
Subject: [sailwave] Boats one-second apart incorrectly shown as tied
I had a situation where two boats with the same PHRF handicap finished one second apart and were shown as tied for first place. The correction formula is CF = 566.431 / (401.431 + PHRF). These two boats had handicaps of 189, so CF = 0.959352. Corrected times (in seconds) were 3409.537... and 3410.49636.... respectively. Still roughly a second apart, but if you round to the nearest significant digit (the whole second) they end up the same. If that's the explanation, how about adding a check for an elapsed time difference when corrected time is the same for two boats with the same handicap?
Remember the issue is positioning boats with the same handicap, not the handicap system itself.
When lookng at boats with differening handicaps it is important to recognise the inherent precision of the handicapping system itself. Considering PYR, many years ago the figures were routinely calculated using “Langstone Tables” which handicapped only minutes - the seconds were added on afterwards un-handicapped. That inaccuracy was fine because the PYRs themselves were on a base of 100 - i.e. 1% accuracy, a second every 1.6 minutes. Now we have computers and PYRs on a base of 1000, but in a one hour race that is still 3.6 seconds each hour. Now consider if that 0.1% is true precision. Clearly not. Recently the RS200 handicap changed from 1059 to1057 and the Laser from 1078 to 1082. That is a total relative change of 6 points or 18 seconds in each hour. Frequently, most boats in our small class fleets finish within 18 seconds, yet in handicap racing gaps of several minutes are common.
So leave corrected times for boats in different classes at 1 second or more granularity and declare ties, but make sure boats of the same class are scored in finish order, or there would be no point scrapping for that small advantage at the line. I would also suggest that boats in a tight finish be given their true elapsed time, as adding extra time to ensure the handicap system splits them could give cause for redress should a boat in another class then split them on handicap.
Handicap racing is a fudge to allow differing boats to race at all, it can never be a precise science.
Ian
···
On 18 July 2011 06:46, Philippe DE TROY philippe@detroy.org wrote:
I had ths problem once with a TMF system. It can happen when TMF’s are lower than 1.000.
It is just nonsense to change accuracy to 1/10 of a second of even 1/100 (prescriptions of the French Sailing Federation) as you can introduce a difference of ranking for boats that should be tied or even rank them in the reverse order. And this is worse than tying boats that shouldn’t.
The protest about the case I’ve known was solved by the Protest committe by adding anoter second to the elapsed time of the slowest boat, after having controlled that this was not creating another undu tye.
Phil
----- Original Message -----
From:
Malcolm Osborne
Sent: Sunday, July 17, 2011 7:17 PM
Subject: Re: [sailwave] Boats one-second apart incorrectly shown as tied
This problem was previously raised (last year) by Mark Townsend - see attached email.
The reply by Art Engel explaining the situation is also attached.
regards,
Malcolm Osborne
Sedgefield, South Africa
----- Original Message -----
From:
jhuus
Sent: Friday, July 15, 2011 7:33 PM
Subject: [sailwave] Boats one-second apart incorrectly shown as tied
I had a situation where two boats with the same PHRF handicap finished one second apart and were shown as tied for first place. The correction formula is CF = 566.431 / (401.431 + PHRF). These two boats had handicaps of 189, so CF = 0.959352. Corrected times (in seconds) were 3409.537... and 3410.49636.... respectively. Still roughly a second apart, but if you round to the nearest significant digit (the whole second) they end up the same. If that's the explanation, how about adding a check for an elapsed time difference when corrected time is the same for two boats with the same handicap?
Am I missing something here? Handicapping is artificially precise
yet you seem to be measuring to within a second - just because you
can. Perhaps in cases like this you should change your computers so
that finishes cannot be� recorded closer than two or three seconds
apart. You could introduce a little algorithm that disables the
‘finish’ button for x seconds after a press. Probably lead to
whining from deluded competitors. Meanwhile our tax system in Canada
neither charges nor refunds amounts less than one dollar and we
survive without worrying about the cents.
Donald�� On 18/07/2011 6:09 AM, Ian Savell wrote:
···
O:-)
Remember the issue is positioning boats with the samehandicap, not the handicap system itself.
When lookng at boats with differening handicaps it isimportant to recognise the inherent precision of the
handicapping system itself. Considering PYR, many years
ago the figures were routinely calculated using “Langstone
Tables” which handicapped only minutes - the seconds were
added on afterwards un-handicapped. That inaccuracy was
fine because the PYRs themselves were on a base of 100 -
i.e. 1% accuracy, a second every 1.6 minutes. Now we have
computers and PYRs on a base of 1000, but in a one hour
race that is still 3.6 seconds each hour. Now consider if
that 0.1% is true precision. Clearly not. Recently the
RS200 handicap changed from 1059 to1057 and the Laser from
1078 to 1082. That is a total relative change of 6 points
or 18 seconds in each hour. Frequently, most boats in our
small class fleets finish within 18 seconds, yet in
handicap racing gaps of several minutes are common.So leave corrected times for boats in different classes at1 second or more granularity and declare ties, but make
sure boats of the same class are scored in finish order,
or there would be no point scrapping for that small
advantage at the line. I would also suggest that boats in
a tight finish be given their true elapsed time, as adding
extra time to ensure the handicap system splits them could
give cause for redress should a boat in another class then
split them on handicap.Handicap racing is a fudge to allow differing boats torace at all, it can never be a precise science.
Ian
On 18 July 2011 06:46, Philippe DETROY philippe@detroy.org
wrote:�
I had ths problem once with a TMFsystem.� It can happen when TMF’s are lower
than 1.000.�It is just nonsense to change accuracy to1/10 of a second of even 1/100
(prescriptions of the French Sailing
Federation) as you can introduce a
difference of ranking for boats that should
be tied or even rank them in the reverse
order.� And this is worse than tying boats
that shouldn’t.�The protest about the case I've known wassolved by the Protest committe by adding
anoter second to the elapsed time of the
slowest boat, after having controlled that
this was not creating another undu tye.
�
Phil
�
----- OriginalMessage -----
From: Malcolm Osborne
Sent:
Sunday, July 17, 2011 7:17 PM
Subject:
Re: [sailwave] Boats one-second apart
incorrectly shown as tied
Thisproblem was previously raised (last
year) by Mark Townsend - see attached
email.
�
The replyby Art Engel explaining the situation is
also attached.
�
regards,
Malcolm Osborne Sedgefield, South Africa
�
�
-----Original Message -----
From:
jhuus
Sent:
Friday, July 15, 2011 7:33 PM
Subject:
[sailwave] Boats one-second apart
incorrectly shown as tied
�
I had a situation where two boats
with the same PHRF handicap finished
one second apart and were shown as
tied for first place.
The correction formula is CF = 566.431
/ (401.431 + PHRF). These two boats
had handicaps of 189, so CF =
0.959352.
Corrected times (in seconds) were
3409.537… and 3410.49636…
respectively. Still roughly a second
apart,
but if you round to the nearest
significant digit (the whole second)
they end up the same.
If that’s the explanation, how about
adding a check for an elapsed time
difference when corrected time is the
same for two boats
with the same handicap?
I’m finding this thread fascinating on many levels, though I don’t
think it’s a Sailwave issue.� From a scoring tool perspective, I
think SW is doing exactly what it should be.� As a previous poster
mentioned, mathematical calculations are only valid to the least
precise input.� So if you’re entering hh:mm:ss, SW should give you
back hh:mm:ss (rounded as appropriate).
That said, it's pretty frustrating to be racing against a level
boat, finish a bit ahead only to result in a tie after the math.�
But that’s part of the fundamental problem with handicap systems.�
They are barely suited to the task they are given (trying to compare
apples and oranges in a fair way).� When you hit the corner cases,
the handicap system often fails.� That’s why I’ve never heard any
sailor utter the words: “I love the handicap system”. 
As with many things in the world of yacht racing, I don't know what
the “right” answer to this question is.� If you went into a protest
or redress hearing, the typical thing to do would be to collect the
facts, apply the rules and make a decision based on the application
of rules to facts.� In this case, a literal application of rules
(both RRS & handicapping system) would indicate that the boats
should be tied.� I don’t know of any rules that say you can
arbitrarily change a boats finishing place.� Perhaps there is a rule
that would allow this to happen?� I’m sure that some protest
committees would want to grant redress, but I can’t think of how
they could justify it (based on the rules, not “common sense”).
Regards,
Pat
Donald Wyllie wrote:
···
O:-)
Remember the issue is positioning boats with the samehandicap, not the handicap system itself.
When lookng at boats with differening handicaps it isimportant to recognise the inherent precision of the
handicapping system itself. Considering PYR, many years ago
the figures were routinely calculated using “Langstone
Tables” which handicapped only minutes - the seconds were
added on afterwards un-handicapped. That inaccuracy was fine
because the PYRs themselves were on a base of 100 - i.e. 1%
accuracy, a second every 1.6 minutes. Now we have computers
and PYRs on a base of 1000, but in a one hour race that is
still 3.6 seconds each hour. Now consider if that 0.1% is
true precision. Clearly not. Recently the RS200 handicap
changed from 1059 to1057 and the Laser from 1078 to 1082.
That is a total relative change of 6 points or 18 seconds in
each hour. Frequently, most boats in our small class fleets
finish within 18 seconds, yet in handicap racing gaps of
several minutes are common.So leave corrected times for boats in different classes at 1second or more granularity and declare ties, but make sure
boats of the same class are scored in finish order, or there
would be no point scrapping for that small advantage at the
line. I would also suggest that boats in a tight finish be
given their true elapsed time, as adding extra time to
ensure the handicap system splits them could give cause for
redress should a boat in another class then split them on
handicap.Handicap racing is a fudge to allow differing boats to raceat all, it can never be a precise science.
Ian
On 18 July 2011 06:46, Philippe DETROY philippe@detroy.org
wrote:�
Ihad ths problem once with a TMF system.� It
can happen when TMF’s are lower than 1.000.�It is just nonsense to change accuracy to 1/10of a second of even 1/100 (prescriptions of
the French Sailing Federation) as you can
introduce a difference of ranking for boats
that should be tied or even rank them in the
reverse order.� And this is worse than tying
boats that shouldn’t.�The protest about the case I've known wassolved by the Protest committe by adding
anoter second to the elapsed time of the
slowest boat, after having controlled that
this was not creating another undu tye.
�
Phil
�
----- OriginalMessage -----
From: Malcolm Osborne
Sent:
Sunday, July 17, 2011 7:17 PM
Subject:
Re: [sailwave] Boats one-second apart
incorrectly shown as tied
This problemwas previously raised (last year) by Mark
Townsend - see attached email.
�
The reply byArt Engel explaining the situation is also
attached.
�
regards,
Malcolm Osborne Sedgefield, South Africa
�
�
----- OriginalMessage -----
From: jhuus
Sent:
Friday, July 15, 2011 7:33 PM
Subject:
[sailwave] Boats one-second apart
incorrectly shown as tied
�
I had a situation where two boats with
the same PHRF handicap finished one
second apart and were shown as tied for
first place.
The correction formula is CF = 566.431 /
(401.431 + PHRF). These two boats had
handicaps of 189, so CF = 0.959352.
Corrected times (in seconds) were
3409.537… and 3410.49636…
respectively. Still roughly a second
apart,
but if you round to the nearest
significant digit (the whole second)
they end up the same.
If that’s the explanation, how about
adding a check for an elapsed time
difference when corrected time is the
same for two boats
with the same handicap?
As results transcriber (from on-paper calculations done live into Sailwave) I cross check the results for errors (on both sides). In the case of a handicap race fleet where identically handicapped boats with differing finish times were given a tie (in Sailwave), then I would add / subtract a second to re-produce the previously posted (on paper) result. Just a SMOP in SW?
Is anyone here seriously saying that, in such a case (the recorded results do show a difference in place), a tie is a fair result?
Mike
Lancing SC
Exactly Mike!
If Laser 167406 sneaks the overlap on Laser 189634 at the lee mark and sprints for the nearby line, gaining a place by just a few feet, he’s not going to be happy to be told afterward “it was a tie” because there was also a Topper in the race so “handicap rules apply”. No club I’ve ever sailed in would do that.
And the secondary point, if 189634 has 2 seconds added to his finish time (handicap > 1000 so 1 second won’t do) to force the result system to break the tie, and then said Topper beats him by 1 second, his bad day will get a whole lot worse, as will that of the race committee who falsified his time!
Ian.
···
On 18 July 2011 13:27, Mike mike.croker@phonecoop.coop wrote:
As results transcriber (from on-paper calculations done live into Sailwave) I cross check the results for errors (on both sides). In the case of a handicap race fleet where identically handicapped boats with differing finish times were given a tie (in Sailwave), then I would add / subtract a second to re-produce the previously posted (on paper) result. Just a SMOP in SW?
Is anyone here seriously saying that, in such a case (the recorded results do show a difference in place), a tie is a fair result?
Mike
Lancing SC
Guys
Lets make life easy.
Let the program work out ties, do not try to apply who was first on the water.
Go back to the rule book. Appendix A A7 boats are scored on corrected time.
The RYA PY instructions say
···
From: Ian Savell
Sent: Monday, July 18, 2011 2:25 PM
Subject: Re: [sailwave] Re: Boats one-second apart incorrectly shown as tied
Exactly Mike!
If Laser 167406 sneaks the overlap on Laser 189634 at the lee mark and sprints for the nearby line, gaining a place by just a few feet, he’s not going to be happy to be told afterward “it was a tie” because there was also a Topper in the race so “handicap rules apply”. No club I’ve ever sailed in would do that.
And the secondary point, if 189634 has 2 seconds added to his finish time (handicap > 1000 so 1 second won’t do) to force the result system to break the tie, and then said Topper beats him by 1 second, his bad day will get a whole lot worse, as will that of the race committee who falsified his time!
Ian.
On 18 July 2011 13:27, Mike mike.croker@phonecoop.coop wrote:
As results transcriber (from on-paper calculations done live into Sailwave) I cross check the results for errors (on both sides). In the case of a handicap race fleet where identically handicapped boats with differing finish times were given a tie (in Sailwave), then I would add / subtract a second to re-produce the previously posted (on paper) result. Just a SMOP in SW?
Is anyone here seriously saying that, in such a case (the recorded results do show a difference in place), a tie is a fair result?
Mike
Lancing SC