By Mihika Desai
Often referred to as the pinnacle of motorsport, Formula 1 demands physical and mental prowess from its drivers. But at the same time, it involves efficient strategies, split-second decision making and immense pressure. During a 90-minute race, teams are essentially solving a series of high-stake economic problems: managing resources, navigating uncertainty and optimizing returns – all at 300km/h. And with advancements in telemetry, simulators, and machine learning, race strategy has become one of the most technological driven strategic environments in any sport.
At its base, Formula 1 strategy is essentially about maximizing positions under uncertainty, a textbook economic problem. Every undercut, tire-selection, and pit stop requires weighing trade-offs: track positions versus tire degradation, aggression versus conservation, safety-car gambles vs consistency. The wrong call can cost millions in lost points, prize money and sponsorship value- making the sport a real-time situation in applied decision theory. A Grand Prix is a continuous series of optimization problems solved by teams, engineers, drivers and race strategists.
Teams’ choices maximize the expected position given imperfect information. This is done by quantifying everything. As explained by Formula 1 strategist Neil Martin, the key behind race strategy is in preparation before the race, which includes stringing the main variables – tyre degradation over time, time lost in pit stops, which drivers perform better at different race circuits, and the underlying pace of the cars and teams. These combined allow the strategists to build a base strategy, allowing them to minimise race time by calculating the quickest way around the track, when to pit and what tyres to use. From there on they add more variables like the other cars, traffic patterns on track and the probability of a safety car.
The race engineers and teams use AWS’s (Amazon Web Services) broadest and deepest functionality to collect, analyse and leverage data and content to make decisions. Each race car has 300 sensors, generating more than 1.1 million data points per second transmitted from the cars to the pit. These monitor everything from the car’s speed and acceleration, to the steering angle, to the status of its drag reduction system. Teams analyse strategies and outcomes from past races, and debate which data points would be most valuable to extract and feed ingest into Machine Learning models, as well as which data points might be most valuable to include in real time during a race.
After running 20 – 30 million simulations to understand the race conditions, teams choose a strategy that minimises their weak points and matches their risk profiles. This involves risk-return tradeoffs – just as invested money can render higher profits only if the investor accepts a higher possibility of losses, a team can gain higher positions if they have a higher risk preference.
For instance, there might be two competing strategies: one gives the team exposure to the podium but risks finishing ninth or tenth if it fails, while the other might guarantee fourth but with no podium chance.
An example is in Miami 2024, when Lando Norris’ first win resulted when he stayed out during a safety car when Max Verstappen and Oscar Piastri pitted. The timing of the safety car meant Norris could pit under caution, losing less time than he would have under green-flag conditions, and rejoined the track in first place, ahead of the rest of the field.
Further, telemetry (real-time data collected) tells the team everything – vehicle speed, tyre pressure, energy consumption, engine performance, driver controls and more. But telemetry alone does not guarantee the right call. The abundance of information doesn’t imply clarity. Strategy desks must filter the million data points into a yes/no decision to be made within seconds.
Here the economics of error is highlighted. A sub-optimal tyre change or delayed pit stop does not cost merely a few positions, it can cost millions of dollars in sponsorship, commercial value and price money. With the winner of the Constructor’s Championship receiving $140m, the value of a good strategy team is quantifiable in simple economic terms: they produce returns.
Every strategic call is made in a competitive environment where rivals make their own optimalisations simultaneously. Teams not only choose their own strategy- they respond to what competitors might do. This is classic game theory.
If Mercedes pit early, Red Bull must decide whether to “cover” the stop, knowing that hesitation gives the rival the undercut. If McLaren stay out during a safety car, Ferrari may evaluate whether following them gives them a dominant position later in the race. The “best” move may not be the fastest on paper, but may prevent a rival from achieving a more favourable equilibrium.
One of the sport’s most important variables is tire degradation which can be understood through diminishing marginal returns. Managing tire degradation, or the gradual decline in tire performance each lap due to wear and heat, is crucial, as it influences lap times and determines when a pit stop should occur. A tyre may lose only a few seconds per lap initially but may drop once it enters a higher degradation phase. This means there is a precise point where staying out on track no longer yields a net benefit. Teams refer to this as the “cliff”, but economists may see it as the point where marginal cost exceeds marginal benefit.
Pit too early and one could lose track position that cannot be recovered. Pit too late and seconds could be lost in tyre wear. The optimal strategy lives between this slim choice between the two.
Although the algorithms run these situations and suggest strategies, each team sits at the centre of race strategy. Despite the sophistication of the computational tools, humans still sit at the centre of race strategy. Algorithms suggest options, but strategists interpret risk profiles and choose according to team preferences.
Different teams have different risk tolerances – how much uncertainty are they willing to accept to achieve a win. This is a concept which can be understood using behavioural economics. For example, Red Bull Racing favours high-reward strategies. With the team’s boss stating that risk taking has become a part of their DNA, they trust that they can recover from risk.
Time pressure amplifies behavioural biases: loss aversion (fear of losing track position), overconfidence, anchoring on pre-race expectations and reacting to rivals rather than data.
The best strategists are decision-makers who balance computed outputs with real-world uncertainty and driver feedback.
A single decision can be worth millions. A well-timed undercut can gain track position worth several millions in season standings while a pit stop error or bad strategy may cost a team seconds. In a sport like Formula 1 where tens or hundreds of millions are invested into car development, engines, and operations, every thousandth of a second is worth its weight in gold.
This is what makes F1 a uniquely valuable economic case study. It is a real-world model of high-speed, high-pressure optimization where technology and decisions directly affect financial results.
The views expressed in this article are the author’s own and may not reflect the opinions of The St Andrews Economist.
Image Credits: Formula 1 and CrowdStrike