As someone very new to the sport of Eventing I’m always trying to better my understanding, and intuition around the importance of each phase, how they sum together, and what would be considered a “good” score in each of them. In this pursuit, I have very real advantage over most newcomers to the sport, in that I can bounce all of my questions off our in-house international athlete and Eventing data guru, Sam Watson. Even with this advantage, I’m increasingly realising that the answers to these questions are actually quite complex, and are very dependent on horse, level, venue and a myriad of other factors.
Even as a newcomer, I have picked up on the feeling and conversations around the sport becoming increasingly dressage focused. People with more experience than me can debate whether this is a positive or negative thing for the sport. For potential solutions, take a look at an article by Sam on “The Z Line”. I will tackle an altogether simpler question, how have dressage scores in Eventing behaved over the past decade?
In the following analysis I have used data from all international competitions since 2010, and in order to include 2018 in the results, we have adjusted all dressage scores to the current system (removing the multiplier for all results pre-2018).
Averaging all international tests by year since 2010 paints a compelling picture. Scores have been decreasing almost linearly over the past nine years. The average test score has gone from 38.1 to 35.0, dropping over 3 whole marks.
What was perhaps more surprising is how this trend holds across all (international) levels of the sport. While the average score is quite different per level, most notably in the CCI5 star competitions, each level has decreased in its average score by roughly three marks over the same time period.
Across almost all sports, we have gained better understanding of how to get better results by better training, better preparation, better nutrition, better selection of athletes, better tools, materials and designs. Using this understanding has led to consistently better results.
Is that what we are seeing here? One way to measure this is to compare dressage improvements to the other two phases, to see if they have improved similarly. If we look at the difference in average phase penalties per year vs the average in 2010 we see that most of the improvement in the sport is accounted for in the decrease in dressage score. XC Penalties have dropped by approx 1.3 marks (but up and down each year, see graph below) and SJ penalties have stayed roughly the same over the past 9 years. Dressage scores are the only phase that has consistently dropped year on year, and have dropped a full 3 marks over the past 9 years. Interestingly, the trend is showing no sign of levelling yet – which begs a question for the sport to consider, how low will it go?
Clearly, dressage scores are getting lower. Next, I wanted to see within which performance group the improvement was coming. For this, we measure the maximum dressage score of the top 1% ,10% , 50% and 90% of results per year.
I expected the biggest gains to be coming from the middle performance bands, the places where there was the biggest marginal gains available. Interestingly, from the graphs we can see that dressage scores at every level of the sport, have decreased across all performance bands, with the top 1% of results decreasing by approx. 2 marks at each competition level since 2010.
The decrease in dressage score is not limited to just the average performer, but is seen across all performance bands at all competition levels.
Dressage scores have consistently decreased over the past 9 years. That has to be taken into account by riders and owners as you plan your seasons and targets. If you are standing still, you are going backwards. Scores have decreased by a similar amount (approx 3 marks) at each international competition level. Inside competition levels, scores have also decreased among all performance bands, with the top 1% of scores falling ~2 marks in all competition levels. Better dressage scores have accounted for most of the improvement in final scores over the past 9 years. While XC penalties have shown some decrease (~1.35 marks), SJ penalties have stayed roughly the same since 2010.
Sean Murray is the Chief Data Scientist at EquiRatings.
He graduated from Trinity College Dublin with degrees in
Theoretical Physics and High Performance Computing.
He is currently writing a model to predict win and podium
placings across the CCI4 S&L and CCI5 competitions in 2019.
Share this story: