Competition News
Kommentar zu Regeln
Martin Scheel, Manilla 3.3.2007
Gestern wurde die Windows Extention Time verwendet, obwohl das Window (Startfenster) nie geschlossen wurde. Dies ist in den Regeln nicht vorgesehen und am ersten Teamleader-Briefing erzählte God Godfrey, wie mit der Stoppuhr die Zeit gemessen werde, in der das Window geschlossen ist! Auf meine Nachfrage heute, warum die Extension Time verwendet wurde, obwohl das Window nie geschlossen wurde erhielt ich die Antwort: „Die Piloten brauchten Zeit, um den Startplatz zu wechseln“.
Der Protest der Italiener, den wir auch unterstützten, wurde von der Jury abgelehnt. Im ersten Absatz der Begründung hält die Jury klar fest, dass dies gegen das Reglement verstößt. Dann folgen zwei Seiten Erklärungen, warum der Protest doch abgelehnt wird.
Eine ähnliche Situation übrigens letztes Jahr an der EM in Frankreich: Ungeachtet der Regeln und gesagtem an den Briefings wurde der Protest „Wolkenflug“ entschieden. Richtig oder falsch, die Wolkenflieger zu bestrafen, das sei dahingestellt. Dass aber die Jury einfach eine willkürlich festgelegte Höhe als „max. in der Wolke“ festsetzt und vor allem, dass der Penalty absolut willkürlich auf 50% reduziert wird, das ist lächerlich.
Ja, wozu sind Regeln. Ich bin keineswegs ein Verfechter der harten Linie. Was aber an den FAI 1 Wettkämpfen regelmäßig passiert ist lächerlich:
Einerseits wird pingeligst auf jedem Detail herumgeritten. So mußte Chrigel zb eine Bestätigung bringen, dass seine FAI Lizenz (Text auf der Lizenz: „Valid 03-07“), bis Ende März gültig ist. Und in Stundenlangen Diskussionen wird gehandelt, ob auch Team-Fahrzeuge die Piloten nach einem Absaufer wieder hochfahren dürfen, wann und wo Medien (Fotografen und Filmer) fliegen dürfen und so weiter...
Auf der andern Seite werden Regeln gebogen, wies beliebt. Und dies noch mit Unterstützung der Jury.
Hier noch die Antwort auf den Italienische Protest:
10th WORLD PARAGLIDING CHAMPIONSHIP
JURY DECISION ON ITALIAN PROTEST ON TASK 1 RESULTS
The jury established that the protest was received in time and the protest fee had been received. It also wishes it to be noted that jury member Ágúst Gu_mundsson was consulted by the Meet Director about correcting a scoring error; this was in his capacity as Chairman of the CIVL Software and Scoring Working Group. He made it clear to the Meet Director that he would do this without passing any judgement on the propriety of making such a correction as it was possible that he might have to subsequently consider that aspect as a juror. At this stage he did not see any of the complaints or responses and merely gave technical information and advice on request.
The jury considered the protest that the Meet Director was wrong to change the Programs GAP 2000 and Race 2003 retrospectively. This basic element of the protest is stated in the first part of paragraph 1) and in paragraphs 3) and 4) of the protest.
The jury first considered the other elements of the protest, which states \"We know how GAP 2000 in conjunction with Race 2003 works in all his (sic) function, also when we use Dep/Arr points in switch off\". For the last 3 years (i.e. since Race 2003 was taken into use by CIVL) no Category 1 cross country events have been scored using the Race 2003/GAP2000 combination. This is the first category 1 event at which this combination of scoring software and scoring formula has been used. Race 2003 with GAP2002 handles the scoring of the type of task set on this occasion distinctly differently to Race2003 with GAP2000. CIVL were not aware of this error in the Race2003/GAP2002 combination when it was approved for use in this competition. If competition directors at Category 2 events have become aware of this behaviour they have not notified it to CIVL. Nor is their any evidence that, at the task briefing, the behaviour was well known or that it would not produce a task valued at 1000 points.
The second element of the protest is that at the Team Leader briefing on the day following the task the Meet Director was not clear about the problem and did not explain that a correction would produce a task score of 1000 points. Two members of the jury were present at that briefing and clearly remember the Meet Director explaining that all validity factors were met and that it should have scored as a 1000 point task. He explained that this was the issue that he was working to resolve. There were no objections at this point.
The jury then considered whether there was a programming error in Race 2003 when it utilises the GAP 2000 scoring formula to score an elapsed time speedrun task without departure or arrival points.
In the GAP 2000 User Guide for both GAP2000 and GAP2002 the technical descriptions of the formulas are identical except that the lead coefficient is used instead of departure points in GAP2002. In neither guide is there any reference to the effect of disabling the departure and/or arrival points and there is certainly no indication that the total points available would be less as a result of ticking these options and this is not the case when using the later formula where the flaw in functionality is corrected. Nor is there any available Race documentation explaining this and it is not provided via the programme\'s Help menu.
The way the time points are reduced by the error in the code for Race2003/GAP2000 is unfair in that distance points for all pilots are (in the case of this task) calculated against a base of 1000 points yet the speed points for those who made goal are only calculated against a base of 862.3
It is clear from the task Scoring Formula Data (on the last page of the Task Results (GAP) report) that the total points available for the task are 1000 with 541 of these allocated to distance yet the error in the code only makes a further 321.3 points available for time. The later GAP2002 formula makes the entire 459 time points available. It is clearly the previously undiscovered coding error that removes 137.7 of these points from the task winner. It is not considered either fair or correct for 44 of the pilots to have a portion of their total score calculated against a lower base than the remaining pilots.
There are check boxes to disable both departure and arrival points in both Race2000 and Race2003 when using the GAP 2000 formula but the functionality of these check boxes produces different results in GAP 2000 and GAP 2002 even if their description is the same in this respect.
Having satisfied themselves that a previously undiscovered error exists in the scoring of tasks with no departure or arrival points when using the Race2003/GAP2000 combination the jury went on to consider whether or not it was appropriate for the Meet Director to apply a correction to the program.
The principle of correcting scoring errors is already firmly established and regularly applied but these usually only affect a single or small number of pilots and are dealt with manually.
Although there is no previous record of scoring programmes being altered to produce the correct result during a competition there are regular examples of data within programmes being altered manually rather than automatically to produce the correct results. These manual fixes occur regularly in data for both the flight verification software and in Race itself and have in fact been used for other aspects of scoring this particular task without complaint. It has never been thought necessary to stipulate in the Local Regulations that errors may be corrected in this way.
On this occasion 44 pilots were affected by the error and it was possible to correct the error in the scoring module of Race2003 thus automating the work and avoiding the possibility of additional errors occurring when making multiple manual corrections. This correction to the scoring module was supplied by the programmer currently maintaining CIVL\'s Race software and the jury therefore consider it appropriate for the Meet Director to use it.
Manilla, NSW, 2 March 2007 ________________________________
John Aldridge – Jury President
________________________________
Ágúst Gu_mundsson – Jury Member
________________________________
Mansoo Chae - Jury Member
Gestern wurde die Windows Extention Time verwendet, obwohl das Window (Startfenster) nie geschlossen wurde. Dies ist in den Regeln nicht vorgesehen und am ersten Teamleader-Briefing erzählte God Godfrey, wie mit der Stoppuhr die Zeit gemessen werde, in der das Window geschlossen ist! Auf meine Nachfrage heute, warum die Extension Time verwendet wurde, obwohl das Window nie geschlossen wurde erhielt ich die Antwort: „Die Piloten brauchten Zeit, um den Startplatz zu wechseln“.
Der Protest der Italiener, den wir auch unterstützten, wurde von der Jury abgelehnt. Im ersten Absatz der Begründung hält die Jury klar fest, dass dies gegen das Reglement verstößt. Dann folgen zwei Seiten Erklärungen, warum der Protest doch abgelehnt wird.
Eine ähnliche Situation übrigens letztes Jahr an der EM in Frankreich: Ungeachtet der Regeln und gesagtem an den Briefings wurde der Protest „Wolkenflug“ entschieden. Richtig oder falsch, die Wolkenflieger zu bestrafen, das sei dahingestellt. Dass aber die Jury einfach eine willkürlich festgelegte Höhe als „max. in der Wolke“ festsetzt und vor allem, dass der Penalty absolut willkürlich auf 50% reduziert wird, das ist lächerlich.
Ja, wozu sind Regeln. Ich bin keineswegs ein Verfechter der harten Linie. Was aber an den FAI 1 Wettkämpfen regelmäßig passiert ist lächerlich:
Einerseits wird pingeligst auf jedem Detail herumgeritten. So mußte Chrigel zb eine Bestätigung bringen, dass seine FAI Lizenz (Text auf der Lizenz: „Valid 03-07“), bis Ende März gültig ist. Und in Stundenlangen Diskussionen wird gehandelt, ob auch Team-Fahrzeuge die Piloten nach einem Absaufer wieder hochfahren dürfen, wann und wo Medien (Fotografen und Filmer) fliegen dürfen und so weiter...
Auf der andern Seite werden Regeln gebogen, wies beliebt. Und dies noch mit Unterstützung der Jury.
Hier noch die Antwort auf den Italienische Protest:
10th WORLD PARAGLIDING CHAMPIONSHIP
JURY DECISION ON ITALIAN PROTEST ON TASK 1 RESULTS
The jury established that the protest was received in time and the protest fee had been received. It also wishes it to be noted that jury member Ágúst Gu_mundsson was consulted by the Meet Director about correcting a scoring error; this was in his capacity as Chairman of the CIVL Software and Scoring Working Group. He made it clear to the Meet Director that he would do this without passing any judgement on the propriety of making such a correction as it was possible that he might have to subsequently consider that aspect as a juror. At this stage he did not see any of the complaints or responses and merely gave technical information and advice on request.
The jury considered the protest that the Meet Director was wrong to change the Programs GAP 2000 and Race 2003 retrospectively. This basic element of the protest is stated in the first part of paragraph 1) and in paragraphs 3) and 4) of the protest.
The jury first considered the other elements of the protest, which states \"We know how GAP 2000 in conjunction with Race 2003 works in all his (sic) function, also when we use Dep/Arr points in switch off\". For the last 3 years (i.e. since Race 2003 was taken into use by CIVL) no Category 1 cross country events have been scored using the Race 2003/GAP2000 combination. This is the first category 1 event at which this combination of scoring software and scoring formula has been used. Race 2003 with GAP2002 handles the scoring of the type of task set on this occasion distinctly differently to Race2003 with GAP2000. CIVL were not aware of this error in the Race2003/GAP2002 combination when it was approved for use in this competition. If competition directors at Category 2 events have become aware of this behaviour they have not notified it to CIVL. Nor is their any evidence that, at the task briefing, the behaviour was well known or that it would not produce a task valued at 1000 points.
The second element of the protest is that at the Team Leader briefing on the day following the task the Meet Director was not clear about the problem and did not explain that a correction would produce a task score of 1000 points. Two members of the jury were present at that briefing and clearly remember the Meet Director explaining that all validity factors were met and that it should have scored as a 1000 point task. He explained that this was the issue that he was working to resolve. There were no objections at this point.
The jury then considered whether there was a programming error in Race 2003 when it utilises the GAP 2000 scoring formula to score an elapsed time speedrun task without departure or arrival points.
In the GAP 2000 User Guide for both GAP2000 and GAP2002 the technical descriptions of the formulas are identical except that the lead coefficient is used instead of departure points in GAP2002. In neither guide is there any reference to the effect of disabling the departure and/or arrival points and there is certainly no indication that the total points available would be less as a result of ticking these options and this is not the case when using the later formula where the flaw in functionality is corrected. Nor is there any available Race documentation explaining this and it is not provided via the programme\'s Help menu.
The way the time points are reduced by the error in the code for Race2003/GAP2000 is unfair in that distance points for all pilots are (in the case of this task) calculated against a base of 1000 points yet the speed points for those who made goal are only calculated against a base of 862.3
It is clear from the task Scoring Formula Data (on the last page of the Task Results (GAP) report) that the total points available for the task are 1000 with 541 of these allocated to distance yet the error in the code only makes a further 321.3 points available for time. The later GAP2002 formula makes the entire 459 time points available. It is clearly the previously undiscovered coding error that removes 137.7 of these points from the task winner. It is not considered either fair or correct for 44 of the pilots to have a portion of their total score calculated against a lower base than the remaining pilots.
There are check boxes to disable both departure and arrival points in both Race2000 and Race2003 when using the GAP 2000 formula but the functionality of these check boxes produces different results in GAP 2000 and GAP 2002 even if their description is the same in this respect.
Having satisfied themselves that a previously undiscovered error exists in the scoring of tasks with no departure or arrival points when using the Race2003/GAP2000 combination the jury went on to consider whether or not it was appropriate for the Meet Director to apply a correction to the program.
The principle of correcting scoring errors is already firmly established and regularly applied but these usually only affect a single or small number of pilots and are dealt with manually.
Although there is no previous record of scoring programmes being altered to produce the correct result during a competition there are regular examples of data within programmes being altered manually rather than automatically to produce the correct results. These manual fixes occur regularly in data for both the flight verification software and in Race itself and have in fact been used for other aspects of scoring this particular task without complaint. It has never been thought necessary to stipulate in the Local Regulations that errors may be corrected in this way.
On this occasion 44 pilots were affected by the error and it was possible to correct the error in the scoring module of Race2003 thus automating the work and avoiding the possibility of additional errors occurring when making multiple manual corrections. This correction to the scoring module was supplied by the programmer currently maintaining CIVL\'s Race software and the jury therefore consider it appropriate for the Meet Director to use it.
Manilla, NSW, 2 March 2007 ________________________________
John Aldridge – Jury President
________________________________
Ágúst Gu_mundsson – Jury Member
________________________________
Mansoo Chae - Jury Member
Comments
New Comment