The following lists updates and any errors reported regarding this analysis in reverse chronological order.
- October 16, 2005
Resolved some difficulties with the Nationals Data
Note: Nationals Data is missing the penalties. That may influence some of the rankings.
Added data about multiple solutions where they occur, demonstrating how we resolve these into ties.
- May 3, 2005
-
Improved presentation of category and flight results.
Merged pilots and judges with duplicate identifiers into one, eliminated multiple entries due to
alternate spellings in different contest data sets.
Added nationals data
- March 24, 2005
-
Corrected consensus ranking for flights in which a judge ties two or more pilots.
Added who flew with whom.
This data needs grooming to merge pilots with multiple entries in the pilot table
due to alternate spellings in different contest data sets.
- February 1, 2005
-
Adjusted the Spearman critical values to agree with those in the given reference.
Added sibling and cousin links.
Added links to explanations.
Removed flights optimistically entered into the scoring program but flown by no pilots.
- January 20, 2005
-
This was the initial release.
The eTBLP computation does not resort to the mean when there are too few pilots.
See for example, the 2004 Copperstate Power Primary Flight 1.
The CR computation does not properly account for ties in rankings from a judge.
