Endurance indexes

Since 2006, the discipline of endurance has established performance indices so as to summarize a horse’s career. Moreover genetic indices have also been available since 2012. These indices are specific, and it is important to understand them so as to interpret and use them appropriately.

3

coup de chaleur
Table of contents

History of endurance index

In 2003, the endurance sector was consulted on their needs concerning research, so breeders of endurance horses were approached via the ACA ( Association du Cheval Arabe) and they requested that performance and genetic indices be calculated for endurance.

The scientific committee of the French Haras Nationaux (National stud), decided to work on this topic, and Ifce/-INRA (Institut National de la Recherche Agronomique) researcher Anne Ricard established indices for the discipline of endurance. A constructive exchange between Ifce, INRA and the ACA enabled the publication :

  • In 2006, of the first performance indices
  • In 2012, of genetic indices

Criteria used

Indices are historically calculated on the basis of results in affiliated competitions, published on  FFE Compet,and forwarded to the SIRE and to INRA.

Since indices 2002-2010, data from the FEI database has been taken into account, for French horses ridden by foreign riders in French competitions. The results taken into account are only from competitions run at free pace (90km and over). At present, results of young horse SHF classes are not included in the index calculation.

Corrections are implemented to take into account the following factors :

  • Age : Younger horses are not as fast , but they are placed more frequently than older horses. This is integrated into their index.
  • Gender : Statistically, entire horses are less competitive than geldings. However some horses remained registered as male in the FFE Compet database, even when they had been gelded. They were therefore wrongly attributed the gender factor. Thanks to the FEI database, and information collected by veterinarians during endurance races, these castrations have now been recorded and updated. However, there may still be a few « false males » competing, which is why when a horse is gelded, the SIRE needs to be informed.

Three criteria used to calculate the index and including corrective factors :

Speed

Speed is only of significance in « free pace » competitions. This is why « imposed speed » races are not taken into account. For each horse, it is not the actual speed which is used. It is the comparison between a particular horse’s speed and the speed of the other horses placed. Moreover, this speed is corrected with the average level of the other horses in the race. Thus the quality of the competition encountered is also taken into account.

Placings (or result state)

This does not mean the rank obtained in a competition, which is shown by the speed, but the fact that the horse completed the race or not (elimination, retirement….). Adjustment between causes of non-completion were at one stage considered, but then this idea was abandonned. However, two corrections are included : the difficulty of the race, revealed through the number of horses placed ; and once again, through the level of the competition encountered.

Distance

endurance
© Ifce
The actual distance of the race is used whenever it is available, now also including one star races. For 90km races, this distance was recorded as a flat rate basis up to 2008. For races over several days, an equivalence has been established. E.g an « 80km x 2 » is equivalent to a 120km race.

Moreover, so as to give more weight to international championships (European, open , and world Championships), the distance for these races has been set at a flat rate of 200km, instead of 160km, which has a positive incidence on the primary index « distance », and therefore on the combined index.

Performance indices

endurance
© A. Laurioux
All the indices published are established over the horse’s global career.

Each of the three criteria used for the calculation (Speed, placings, distance) gives rise to an primary index. An accuracy coefficient which takes into account the number of performances, and the criteria repeatability, is associated to each index.


An overall index, or career index is then calculated using these primary indices, the correlation between them, and by applying the adjustments to each of them. Different simulations were carried out by the INRA and proposed to the project team of ACA. The latter then chose the adjustments which best reflected the breeders perceptions. They have now been established at :

  • 35% for speed
  • 30% for being placed or not
  • 35% for distance

In 2012, a new calculation method was introduced. This takes into account the correlation between the criteria, which is essential for the calculation of a genetic index. Introducing the notion of correlations betwwen the criteria, meant a significant increase in the coefficient of determination (CD), and hence the accuracy of the indices. This has possibly led to significant moidifications to the value of the less accurate indices. It is therefore no surprise if, for some horses, the 2002-2011 index shows a marked difference to the 2002-2010 index, and a higher determination coefficient, even if the horse did not take part in any competitions in 2011.
The performance index, IRE (indices de performances en raid d'endurance), is calculated every year, but is based on the horse’s entire career. We also call it the Career index. The global career index (IRE) summarizes the horse’s qualities, by taking into account all the performances over the three criteria. Thus, for each horse competing, 4 different indices are calculated, but only the global career index (IRE) is published. These indices are presented to show approximately :

  • 50% of horses competing have an index of over 100
  • 30% are over 110
  • 17% are over 120
  • 7,5% are over 130
  • 2,9% are equal to or over 140

Over the 2002-2016 period, 11 008 horses were attributed an index. This represents 58751 competition participations, with 37 587 placings, with a placing being a horse completing a race.

Calculating the genetic index

endurance
© A. Laurioux
The genetic index or BRE (BLUP for endurance races) is established on the basis of the performance index. The genetic index is centered around 0, which represents the average value of the active population. In the same way as for the performance index, a genetic index is established as a value (the index itself) and a coefficient of determination (CD).

When the CD is very low, the corresponding index is not published. It usually indicated that the horse only has a very remote link to endurance.

Publication of the indices

Performance indices

The ACA requested not to publish an index for any horse that was not significant :

  • Horses never placed (190 horses) but indexed over 100 because they competed in long distances. The accuracy for these horses is low, and only 5 of them have a CD of more than 0,5.
  • Downgading for the 739 horses nver placed, but having competed only once : with an index of below 100 and a very low CD (below 0,39)

Indices are available for each horse on the website www.ifce.fr, section Info chevaux

Important note

Following a principle set up by the ACA since the first indices were published, it is the best yearly index which is used as a basis to qualify for a PACE (Bonus attributed for the capacity to compete in equestrian events,  « Prime d'Aptitude à la Compétition Equestre » in French).

Genetic indexes

Each horse’s genetic index is published along with the performance index on the website www.ifce.fr, in the section « Info chevaux » (Horse info). Thus the genetic index appears, but only if the horse is registered at the time of calculation, and when the CD for the index is equal to or over 0,12.

This concerns 68 742 horses since 1990. Some ancestors, such as DENOUSTE, an arab horse born in 1921 is indexed  (BRE +5 ; CD 0,22).

Prospects: Performances abroad included in the indexation for endurance horses

Performances of French horses abroad

For the past few years the issue of integrating performances from competitions abroad has been a matter for debate. Files concerning international performances abroad exist, for the main part recorded on the FEI website, and represents a large volume of information, in particular including international championships.

A first study concerning using this data in the index calculation demonstrated that integrating these results poses several serious issues due to :

  • First of all, it is necessary to correct the imperfections contained in the files.
  • Secondly, the sheer mass of data
  • Lastly the difficulty of assessing the level of the competition encountered

The work carried out brought to light significant gaps in the results used in different indexation simulations, particularly when races from the Gulf states (group 7) are taken into account (average speed in these races reaches 20,3 km/h, versus to 15,5 for France). It is important to note that over half the number of endurance races in the world take place in these countries.

To date, none of the solutions presented following the studies has been adopted by the project team. Therefore indexation for the moment remains based on races run in France.

Conclusion : the path towards more objective characterisation of endurance horses

The new model of performance index characterises horses much better than the previous star system, because each performance is put into context (quality of the field of competitors, difficulty of the race, the horse’s speed ….). Moreover the factors age and gender are taken into account. In addition, they integrate all the available performances, including counter performances.

The genetic index, which takes into account the performances of the lineage, as well as those of the horse itself, give a  much better view of the genetic worth of a stallion, than the previously used DRE. In particular, it enables a distinction of young stallions, even before they become performers. Thus, the ACA promotes any particularly worthy young stallions, which facilitates early testing of their descendants.
However it is natural to quote performances for stallions :

  • Having obtained remarkable performances before 2002
  • Having excellent performer descendants before 2002 or having obtained performances abroad.

Reminder : the performances obtained in the current year are only taken into account in the index at the beginning of the following year.

Above all, do not forget that each index, (performance index or genetic index), is adjusted by an accuracy coefficient (or coefficient of determination, CD), which is equal to a value situated between 0 and 1. The higher the CD the more reliable the index. If the CD is low, the index then only has a small informative value and can vary substantially, either following the horse’s own new performances, (or the performances of horses in his lineage, for the genetic index).

Know more about our authors
  • Translated from french by : Karen DUFFY Translator
  • Margot SABBAGH Development engineer IFCE
  • Sophie DANVY Development engineer IFCE
  • C. DEPUILLE
To find this document: www.equipedia.ifce.fr/en
Editing date: 20 05 2024

Was this information sheet useful ?