Horse racing is one of the most popular spectator sports in the world, and it has a rich heritage. The American Triple Crown – the Belmont Stakes, Preakness Stakes and Kentucky Derby – is among its biggest and most famous events. Other major international races include the Melbourne Cup, Prix de l’Arc de Triomphe and Grand National. Some of these are televised to millions of viewers worldwide. Whether you enjoy watching the sport or wagering on it, you should be aware of a few important facts before you place your bets.

The first and most obvious is that the sport is unequivocally unnatural. Horses are not born to run or compete, and a horse race bears no resemblance to how horses naturally hunt or play.

Breeding 1,000-pound thoroughbreds with massive torsos and spindly legs is a recipe for breakdowns, especially for animals still in adolescence at age 2 when they are thrust into intensive training. Racehorses often die from cardiovascular collapse or a failed heart, and many are killed by blows to the head or spinal fractures caused by collisions with other horses or the track. Other common causes of death are pulmonary hemorrhage, broken necks and severed spines.

Some of the most exciting and historic races occur when a longshot makes an epic last-minute surge to win a head-to-head duel. Secretariat’s 31-length demolition of a field in the 1973 Belmont Stakes is just such an example, as is Arkle’s six-length routing of a field at the 1964 Prix de l’Arc deTriomphe.

Many races erupt into thrilling battles between a horse with the best long-term prospects and a challenger that has better recent form. Those races are often called head-to-heads or star turns, and they are considered the pinnacle of thoroughbred racing.

The new research by Aftalion and Quentin Mercier, a mathematician at the EHESS, reveals that winners are determined by strategies that maximize energy output from muscles requiring both aerobic (which needs oxygen) and anaerobic (which produces waste products that lead to fatigue) fueling pathways. The model could eventually allow trainers to plug in data about individual horses — like their unique aerobic capacities — and get custom racing recommendations, from pacing suggestions to ideal racing distances.