An innovative system to evaluate and train dressage judges will be unveiled at the Global Dressage Forum (GDF) next month (26-27 October).
Maarten van der Heijden, sport director for the Dutch federation (KNHS), will explain the “judges’ dashboard”, developed over the past year in collaboration with the esteemed British scientist Dr David Stickland.
The dashboard has been trialled at around four Dutch national shows a week, in classes just below small tour.
Using programs provided by Dr Stickland’s company Global Dressage Analytics (GDA), every point is recorded. Within three days of each show, judges can study their analysis at home. Non-explainable differences are then evaluated with a personal coach.
GDA already offers a riders’ score analysis service, taken up en bloc by the Dutch federation in 2013.
Mr Van der Heijden said: “We decided to use this same IT from a different perspective.
“To be a good judge, you need to ‘look in the mirror’. We have already made a strong assessment of where some judges need help.”
Dr Stickland is a senior research physicist with Princeton University, currently seconded to the European Organisation for Nuclear Research (CERN), the
institute that discovered the Higgs boson.
Dr Stickland started analysing dressage scores “for my own amusement”, later undertaking consultancy work for the FEI. He was an early advocate of half-points and seven-judge juries.
He co-founded GDA with Akiko Yamazaki, the long-term supporter of US rider Steffen Peters, in 2009.
“There is much more information in dressage scores than people make use of,” said Dr Stickland.
He believes, for instance, that the two-point deduction for an error of course has more effect on the overall percentage than most riders appreciate.
The dashboard took an enormous amount of prior organisation because of forming the habit to send in sheets. The number of judges per show then provided a challenge.
“With seven judges, when one is very different you can hypothesise that the other six were right. But what about when you only have two?” he said.
He resorted to a basic colour-coding technique to identify judging trends of individuals against those of co-judges over the course of a year before being able to assert who was “miles off” with their marks.
The dashboard can analyse such subtleties as marking too high or low, even when overall rankings are compatible. It can detect judges failing to use the full range of marks for, say, collected walk.
“There is a lesson that someone is having trouble judging that particular movement,” he said.
Wayne Channon, secretary general of the International Dressage Riders Club, warmly welcomed the initiative.
He told H&H: “So often one judge can be just 0.5 points lower than their colleagues for a whole test and the result is 5% difference in the total test.
“Similarly, judges often end up with close matching scores, but the way they get there is not the same; one judge may mark the trot lower but compensate by having a higher canter mark.
“For the rider, this is confusing, but for the judge it looks like a good result. The GDA approach will enable judges to see where they are right and where they are seeing things differently.”
Dr Stickland stressed: “It’s important this is seen as education rather than reaction. We all depend on a vast number of volunteers to be judges and must give them the best feedback possible, not make them feel they are being painted into a corner.
“Judges don’t have hours to spare to compare sheets. The dashboard does this for them, in a format they can use when and how they wish.”
Ref: H&H 24/09/2015