Debbie Minahan, who coaches skater Will Annis at the Yarmouth (Mass.) Ice Club, said she often uses video to analyze skaters' techniques but was excited to have access to this level of analysis.

"I thought this was an incredible opportunity that wouldn’t be offered anywhere else," she said. "For Will - working on all of his triples now - it seemed that it would be valuable to have that feedback to get a more consistent repetition out of the jumps. Some of them are more stabilized than others. The hope is that the information that we receive helps stabilize all of them and then he'll be able to move into triple-triples."

Will, who won the 2017 U.S. juvenile boys' championship last January and was second in the intermediate men's competition at the Eastern Sectional Championships in November, said he has a tendency to go too far to the left side on some jumps.

"This will help my air position, so it should help solve that problem," he said.

He was one of several young national-level competitors who visited Newark this fall to have their jumps analyzed in a visual 3-D computer program Richards developed with biomechanics expert Tom Kepple, chief science officer at C-Motion Inc. and a former instructor at UD.

For the test, small markers covered with retro-reflective tape are attached to the skater.

An array of 10 cameras are mounted to the ceiling of the rink, over the testing area, and those cameras capture light as it is reflected from the skater during their test.

"The camera just sees a grouping of red dots," Richards said.

Those dots provide a map for the computer program - relaying data about position, speed, rotation for the computer to plug into the model.

Richards and his research team use the data to produce a pair of three-dimensional computer models - avatars - one showing the skater's actual jump, the other a mathematics-based model that shows what the jump would be like if the skater made specific modifications.

They can see, for example, that if the skater tucks that right elbow in a bit more or doesn't lean quite so much, enough velocity could be gained to complete that triple or quadruple jump.

Nathan Chen, the defending U.S. men's singles champion, became the first skater to land five quadruple jumps in competition last January, raising the bar yet again.

Plenty of skaters have the potential to accomplish this, Richards said, and his model suggests physiologically realistic changes in every joint to show a skater how it can be done.

"There are two parts to jumping successfully," Richards said. "You have to leave the ice with enough energy to rotate at the speed you need to get the number of rotations you want. It's angular momentum. And almost every skater we test has enough of that for most or all of their jumps."

The speed and momentum required varies, of course, depending on whether the skater is doing a double jump, a triple or a quadruple.

Sometimes a skater is missing a jump by a half turn, by a quarter turn, sometimes less. A change in arm position may be all that's necessary to land the jump correctly.

But the adjustments aren't easy for skaters, Richards said. Most tend to conserve their momentum, often without even knowing it, he said. They get comfortable doing a jump at a certain speed and get nervous if they need to ramp it up.

"It feels to some that if they spin faster than 'this,' they will be out of control," he said. "They don't know how to land it."

The biomechanical science - as shown in the model - is quite persuasive, though, Richards said.

"Usually coaches know what they're looking for - but they're sending skaters here because they want them to see it for themselves," he said.