The Science Of Subjective Selection

Dec 25, 2017 - 2:40 PM

Past performances, large strings of horses, coolness under pressure—our columnist isn’t sure our team selectors are giving each ingredient the right weight or being properly transparent.

Finally, after all of the year’s results are in, I can say goodbye to 2017 and engage in one of my fondest guilty pleasures: trying to predict who will make the coming year’s U.S. Eventing team. It makes me happy in a way that New Year’s resolutions, with their inevitable failure and disappointment, never could. I’ve been doing this since the late 1970s, but I cannot say for certain that I’ve gotten any better at it.

Forty years ago, any reasonably knowledgeable eventing spectator could pick the top five or six horses and riders for a team championship, even if they didn’t know the inside scoop. We only had a couple dozen serious candidates for any championship, and many of those you could dismiss for cause without much thought. Those riders rode every bit as well as today’s top riders; we just had fewer to pick from. The riders didn’t always get along. But they won medals, both individually and as teams.

In the modern sport the U.S. team selectors and managers have a somewhat hairier task. For 2018 they have to select five horse-and-rider combinations (four team members and one individual) for the FEI World Equestrian Games in September in Tryon, N.C.

Recently, Erik Duvander, the new technical advisor for U.S. eventing, remarked to me how promising it seemed to him that we have so many good horses and riders from which to choose. Indeed, I believe only England has more horses and riders already qualified to compete at next year’s WEG. That fact changes the selection game in a way we have yet to really figure out.

A committee of five knowledgeable equestrians names riders to the various tiered semi-annual training lists, awards competition and travel grants, and selects teams for major competitions and championships. Each member of that committee brings years of experience producing event horses and competing at the highest levels of the sport. They all take this responsibility very seriously, and every one has the U.S. team’s success as their highest priority.

Unlike in other sports, we do not choose our national teams based purely on objective results at selection trials. Instead we trust our selectors to utilize a blended mix of subjective judgment and objective results to select teams in a way that should, if done well, produce medals. If we didn’t allow this sort of subjectivity, a computer could pick the team. Indeed, we’ve already tried that. It does not work.

Prior to making a decision, the committee identifies each horse and rider’s strengths and weaknesses given past performance. They form opinions about the quality of each horse, whether it has the athletic ability, temperament and heart to compete at top levels. They assess the rider’s technical skills and competitive character in a sometimes vain attempt to predict what each rider’s efforts will result in under the rarified pressure of an FEI championship. They watch competitions leading up to the championship closely, scrutinizing the quality of every performance to discern who is peaking and who is not. And they will have traveled to Tryon to assess whether the venue and the expected level of difficulty will play any significant part in shaping the competition.

Fundamental to the process is the notion that only the best get picked; you earn your place on the team by winning. Each selector understands that they must respect past results in competition as the best indicator of merit. But because the strategy behind winning a team medal requires that at least three team members complete the event, intangibles like consistency, experience and competitive character have great value in this context. How much? Well, that’s a matter of debate.

Never Going To Be More Wrong

Once upon a time, you couldn’t make a championship team for the WEG or an Olympics without some sort of team experience leading up to it. But today, team experience seems almost hard to get since the U.S. Equestrian Federation hasn’t made informal team competitions much of a priority. Further, with so many U.S. horses and riders competing at Fédération Equestre Internationale competitions here and abroad we tend to see a different name in the winner’s circle on any given Sunday. This makes team experience look less and less relevant as a predictor of quality.

Within the pool of qualified candidates, certainly we have five sound horses and riders that can produce a competent dressage test, finish the cross-country without adding dozens of penalties, and be healthy enough on the third day to jog sound and leave rails in place. If the selectors choose wisely, and if we have a little luck, our team can easily win a medal. Nobody gets it right 100 percent of the time; horses are reliably unreliable beasts. But what the selectors wouldn’t give right now for a crystal ball to answer the question: Which five are the right five?

I recently read Michael Lewis’ Moneyball: The Art Of Winning An Unfair Game. If you haven’t read it, it’s the story of how Billy Beane and his staff used targeted statistical analysis and game theory to position the Oakland Athletics as one of the most competitive baseball franchises in the country using one of the smallest budgets in the league. As Lewis explains it, the radical application of statistical analysis led to some important changes in how the Oakland A’s acquired and trained their players as well as the strategies employed on the field of play. When questioned about why he wanted to try such a new and radical approach in drafting players, Beane said, “You know what, however we do it, we’re never going to be more wrong than the way we did it before.” I think we’re in a similar spot with eventing right now, especially where it concerns team selection.

Lately, our teams haven’t fared so well at major championships, with no team podium finishes at WEG or Olympic competitions since winning the bronze medal in the Athens Olympics in 2004. I do not put responsibility for these failures solely on the selectors. But what would it say about U.S. event riders if we said that the selectors always got it right, but the riders repeatedly let us down? I’ve seen them ride. We have many excellent international riders, and they could easily win if we prepare them well with the right tools and support. So it seems fair to say that the selectors and the process they use share some of the blame for these disappointing results.

Sharp Double Edge

I’m not a selector, never have been and probably never will be. But I’m not stupid. I watch the decisions and statements that come from that committee like finance guys watch the Federal Reserve Bank chairman’s statements about interest rates. Every so often, the committee will do something that just cannot be explained logically without insider knowledge.

If these types of unexplainable decisions had consistently led to international success, we wouldn’t need to examine them. You would hear no complaint out of me. But they haven’t. Oh sure, we’ve had winners who carried the Stars and Stripes to individual medals. Amy Tryon, Gina Miles and Phillip Dutton all won individual medals at the WEG or Olympics in the last 12 years. We have a couple candidates with potential to do that right now. But to win a team medal, we only get one drop score, one weak link, and our current system tends to produce two or more of those on a regular basis.

While I will defend vigorously the use of subjectivity in selecting teams, as a tool, that sword has a sharp double edge. When used wisely it can make allowances for bad luck, minor lameness, or a temporary loss of form and can help tailor a horse’s training program to make sure it peaks at the championship. And I support increased flexibility regarding timing when naming a team. Horses and riders gain and lose form too quickly to safely name the team far in advance of the championships. But used unwisely, the selection process loses all integrity and creates an unleveled playing field for the candidates to navigate.

In the past, we’ve seen candidates make teams with mediocre competitive results over those with clearly superior records. We’ve seen selection criteria crafted or altered in ways that favored specific horse-and-rider combinations and disadvantaged others. Some years, the committee has written the selection criteria so loosely as to allow the committee full discretion to act as they wish without any real accountability. They’ve even gone so far as to demand that riders sign waivers stating they would accept the committee’s decisions without complaint.

Grants sometimes get awarded in ways that do not reflect thoughtful development policies, and training lists sometimes ignore obvious differences in soundness and competitive records between the candidates. Moves like these lead to charges that the process is biased, overly political, corrupt, or that the selectors just aren’t good at their jobs. Let’s hope none of that is true.

Challenging Assumptions

Because the selectors and eventing high performance managers have never felt the need to explain how or why they make these choices, we can only speculate about their motivations when they exercise subjective judgment. In some cases, for reasons known only to the selectors, a single poor performance will eliminate a rider from consideration, while in other cases they may choose to ignore it. I try to tell myself that such decisions arise because the selectors know more than I do about the candidates. But the more I know about each candidate, the more I study the statistics and watch them in public, the more I ask myself, “I wonder why they decided to do that?”

So what changes would I make to the process? Well, right off the bat I would willingly challenge some popular assumptions about what makes a rider valuable as a team member. Taking a cue from Moneyball, sometimes I wonder if the selectors aren’t being tricked by what they think they see in the performances they judge and biased in what they think they know about a sport that has changed drastically in the last decade. Just as Billy Beane challenged his scouts to rethink the way they drafted players based on performance data, I would challenge some of the longest held assumptions about team selection.

Right now, the committee seems to favor riders with large strings of horses. Having multiple rides at the upper levels allows riders to hone their skills in a variety of competitive situations. There is strength in numbers. But every strength exposes a weakness. With many horses in training a rider’s time gets split putting an emphasis on quantity over quality and does not promote a close bond between horse and rider. It’s true, riders with many horses get more practice riding events, but the data shows that two of the most successful riders in America right now have only one horse competing at CCI*** or higher.

Riders with one horse have to make sure every moment counts in training, so in this way, what looks like a weakness can become a strength. Further, riders with many horses have the added advantage of always looking like a winner on at least one of their mounts, whereas a rider with only one looks like a failure whenever they don’t win. Just because a rider looked useful to the team on her best horse, doesn’t mean they get the same credit on a lesser horse when their first horse goes lame.

Some riders inspire confidence and can get the selectors’ attention simply because they look like they should win, even if they don’t. A rider who rides well should be able to win and gets easily perceived as potentially better than their results. But potential doesn’t win medals. Unorthodox riders who repeatedly achieve consistently good results might get overlooked for training and support simply because the experts don’t appreciate their form. If you really believe only correct riding can be effective riding, I only need to point you to the great French show jumper Roger-Yves Bost who achieves unassailable results using a style that is as unorthodox as it is comic. I would take his clear round over a 4-fault equitation ride any day.

And don’t even get me started on how seductive a beautiful horse is to horsemen; we easily lose control when the animal looks like Secretariat or American Pharoah. Phillip Dutton’s Mighty Nice had me at hello. But his looks didn’t make him jump clear at Rio last year. Phillip did that. Performance statistics, when well understood, can adjust for this kind of faulty perception.

Just as Billy Beane refused to pay salaries based on what a player did five years ago, I do not believe we should get overly sentimental about past experience in the selection process. We aren’t selecting riders based on what we know they can do if well mounted or what we remember they did in the past. Their value as a team member is limited to what they can do right now on the horses they are riding right now. We don’t have to take an experienced rider on a bad horse when we have a dozen qualified riders who have better records waiting to prove themselves. There isn’t a single rider in the pool of candidates who should be guaranteed a spot just because they’ve been there before. Those riders need to prove themselves just like everyone else.

That said, we can identify certain riders who possess something I call Olympic character, an inner strength and drive that brings out their best performances when other riders would wilt under pressure. This counts for a lot but just how much I cannot say. It’s a matter of subjective judgment to be sure. But how this gets valued depends a great deal on their horses. We’ve put bad horses with experienced riders on teams many times, and it almost never works. We need to keep in mind that experience and determination may give a rider a high chance of completing, but if they do so with scores a dozen other candidates can beat, what’s the point?

Lastly, I believe that the current system of separating the pool of candidates into Elite, Tier 1 and Tier 2 actually sews the seeds of discord amongst the candidates in a negative way. By definition, this system creates an unlevel field from which each candidate must strive to make the team. Riders on the Elite list receive the lion’s share of team benefits, training time with the team advisor, as well as financial reimbursements for training fees paid to private coaches as they prepare their listed horses for the spring selection trials. They will most likely be first considered for Nations Cup teams, and if anyone receives permission to avoid a spring CCI, preventing them from truly being tested in head-to-head competition with their peers, it will be horses on this list. The USEF regards this practice as wisely investing in the candidates most likely to succeed in FEI championships. Because this practice maximizes the potential for mischief, I view it as fundamentally damaging to the integrity of the entire process.

Recently, without explanation, two of the most successful horses in this country in 2017—a horse that won the most competitive CCI*** in the world and the horse that won two CIC*** and had the best CCI**** finish of any American in 2017—were listed as Tier 1, while horses with lesser or no record at all in 2017 made the Elite list. These placements were entirely within the selector’s discretion.

While I can make arguments for why each horse deserves its place on that list (I’m a lawyer after all), there are just too many horses and riders with similarly deserving records to recommend them to allow for any meaningful separation in this manner. And when horses with no recent record due to lameness get placed ahead of horses currently winning, the implicit message sent to these riders is that it doesn’t matter how well they do; it’s never going to be good enough. This practice discourages candidates and their owners from trying to make teams, invariably leads to hurt feelings and jealousy amongst the riders, and allegations against the selectors of using unfair bias in the selection process.

Instead of a tiered system for the allocation of training support, I would encourage the USEF to find another way to provide continuing education for all eligible candidates that works to improve depth of qualified horses and riders, does not operate on a system of exclusivity, and that does not give one candidate an unfair leg up in preparing for major championships. Riders will more readily accept and support the final outcomes when they know they have a level playing field from which to start.

Right now, I think the whole selection process plays to a rider’s weaknesses in a way that sets them up for failure. By the end of the selection process our eventing teams tend to look like the final contestants on Survivor: exhausted, sore, hungry, extremely irritable, in need of showers, and trained by the experience to hate each other. That’s not exactly what we want going into a World Games.

If we look at the last 13 years of team results, we have to admit that the way we have done it hasn’t worked. Just as the riders must strive to improve to become more competitive, I think it is well past time for the selectors and USEF high performance managers to look at their own processes in order to better understand what goes into producing a team medal and what does not, where they go right, and where they go wrong in team selection.

Patrick McGaughan was rider in residence at the U.S. Equestrian Team from 1981 to 1982. He graduated from Duke University (N.C.) in 1987 and from the University of Maryland School of Law in 1991. A team gold medalist at the 1987 Pan American Games aboard Tanzer, he is now “a reformed lawyer spending my time teaching, riding and training horses” at his Banbury Cross Farm in Clarksburg, Md.


Social Bar

Join Mailing List

Shopping Cart

Like Box

Chronicle Headlines

Most Popular

Like Box

Rider Spotlight

Charity Spotlight

Horse Spotlight

Like Box

Trainer Spotlight

Like Box