ENGLISH TEXTS
In a team sport like rugby, two main uses of AI can be identified.“ First, it can assist with physical preparation: planning, predicting workloads. There’ s a lot of talk about injury prevention with AI, but injury, by nature, doesn’ t depend on AI. So, we incorporate certain algorithms that may be useful. We’ ve been using AI in this way for three or four years. Using it effectively is actually very complex, the field is vast, so we test.” Many companies offer socalled“ revolutionary” systems, but most are not.“ AI can help detect trends, but predicting an injury down to a day or a week is impossible. We’ re still cross-referencing more and more data.” In the field of health and performance, certain parameters still escape the algorithms.“ Lifestyle, family context, emotional well-being- those are hard to feed into models. No algorithm can process that.” AI is therefore one more tool serving sport, but the machine still doesn’ t make the decisions.“ For example, we might try to get information from a chatbot, but since we don’ t fully trust it, we always double-check manually. That’ s when we see if things line up or not. We try to humanize the review process, and that’ s where the beauty of performance lies. We’ re monitoring more than ever, but performance is still measured by victory, and we still have no access to a player’ s mental state at a given moment. Humans remain central to the process and the athlete’ s support system.” The second aspect involves collaborative work between data specialists, video analysts, and coaches, who request targeted insights. These relate to team strategy, but also to individual player contributions within that system.“ Our job is to cross-reference all this information, to filter the constant back-and-forth between coaches, video analysts, and data experts, and deliver the most refined possible response to the coaches.” In fact, the three perspectives on the same issue are usually complementary. Either they align, or there are nuances, and in the end, it’ s always the coach who makes the call.“ And the staff doesn’ t show everything to the players. You can’ t overload them with data. We keep it to one, two, maybe three insights at most. Imagine: we process about 300 parameters per match. Multiply that by the number of games in a season, and in the end, only the most relevant conclusions emerge. We do the filtering. Coaches never focus on more than two or three key points. That’ s why having a multidisciplinary staff is crucial, to filter and provide a precise answer.” Are players interested in AI and what it can bring them personally?“ All players are now familiar with AI. Naturally, they’ re interested in it in terms of their performance.
They like getting data feedback after a match, but the point isn’ t just to see who ran more than the others. It’ s to understand how a player contributed to the team. That’ s our job. Otherwise, we’ d miss the point. Above all, players mustn’ t interpret the data themselves, they might think they performed well physically, when in fact they were ineffective in the game. What interests us is collective performance.”
� PAGE 56
MORGANE MOLLE
BACK TO THE ARTICLE �
By Serge Okey
“ DATA ANALYSIS, A POTENTIAL 1-2 HOUR A DAY JOB”
A rider and engineer specialized in genetics, Morgane Molle is one of the top specialists in new technologies within horse racing, based in France. Formerly the“ 2.0 right-hand” of Nicolas Clément, she is now back in France with Amy Murphy’ s new stable.
Galorama. You’ ve traveled the world, including Australia. How prominent are new technologies in horse racing globally?
Morgane Molle. I’ ve seen things in genetics that go beyond AI. Predictive tools and models are tackling the DNA of
147