Project in support of the degree of Master of Engineering

Prediction of Formula One Results


Full title of investigation:

Prediction of Formula One Results Using Driver Characteristics

Author: Ian Horlock         Supervisor: Dr. John Terry




Formula One is the pinnacle of motorsport and as such budgets of over $400million are invested by manufacturer teams each year, with cars of today having more in common with a jet fighter than a normal road car. These budgets are used to build the 1000's of components which make up a Formula One car, ship the cars all around the world and pay up to 600 personnel. Factors such as aerodynamic efficiency, brakes, engines, gearbox, fuel and more recently kinetic energy recovery systems are often seen as the difference between a race winning car and a backmarker car.


Predicting results of a Grand Prix have long remained one of the hardest possible predictions in sport. In Football, Boxing, Rugby, to name a few, there are often clear favorites and the nature of many one on one sports makes it easier to predict. Formula One however contains many more variables due to the combination of machinery and driver as well as the presence of at least 20 drivers.


Being able to predict race times of rivals is an integral part of producing a winning Formula One team as this presents the best opportunity to outwit your opponents before the race has even begun. The last 20 years has seen a dramatic increase in the level of computations performed by teams to predict the results, mainly due to the technology era we live in. There are many different ways of modelling driver performance, some of which include inspecting previous race results, previous season results or potentially more complicated models.

Onto the abstract of the investigation.