How F1 teams kill racing

Before Max Verstappen won in Austria, Mercedes had won the first eight races of the 2019 season.

Verstappen had pushed Hamilton hard in Monaco, and Ferrari lost almost certain wins in Bahrain and Canada, but Mercedes still won. We have to face facts - this is a sport predicated on predictable outcomes.

When major technical changes occur, such as in 2014, one or two teams usually make best use of the step-change in technology or performance, followed by a tight midfield, then a tail-end dealing in battles of the non-racing variety: people, money, or both.

As regulations mature, the midfield will close on the leading teams, but by then everyone is bored with one or two-team domination, so the rules are changed again, which, ironically, guarantees we repeat the cycle.

Qualifying neatly sorts the field into pace order and, drivers being good at exploring a car's limits, we often end up with a Noah's Ark grid - two by two in team order.

From lights out we all know that the fastest cars are out front, the second fastest cars behind etc. Williams is not going to come through the field; Mercedes is not going to flounder around having just locked out the front row.

Strategies are data-driven, optimised and, in this environment of marginal gains on steroids, mistakes are minimised. Most cars finish most of the races most of the time. Reliability has never been better.

Neil Martin of Random Logic Ltd is a data scientist who enjoyed long spells in race strategy roles at McLaren, Ferrari and Red Bull, and confirms what many fear.

"Teams at the front are risk averse," he says. "They develop sensible 'boring' strategies aimed at de-risking as much as possible and guaranteeing an outcome.

"A dominant team like Mercedes can develop a safe lead, pit one lap after their main competitors - thus mirroring their pitstops - meaning there is no possibility of performing the undercut."

Martin confirms that dominant teams can even afford the odd error without affecting the outcome. If the top two or three teams have 0.5s per lap over the midfield, that means a half-minute advantage by race end.

Such is the performance of the leading teams - fixed in the pecking order of relative speed - they don't even bother modeling race strategies against their midfield 'rivals' or backmarkers, apart from considering traffic after a pitstop or when lapping traffic. This confirms another point - the lesser teams are not even in the same race.

Teams also don't like their drivers to race each other, so they often race as a one-team unit. We don't have races of 20 independent drivers, rather 10 unified strategies, further managing outcomes.

In this de-risked, data managed sport, the only variables that can upset the outcome are - ahem - predictable.

Martin confirms these as: pitstop blunders, strategy errors, grid penalties, driver errors in qualifying, safety cars, a wet race, or incorrect tyre choices.

Consequently, the top teams learn how to respond to safety cars and avoid penalties, and employ drivers who minimize mistakes, are good in the wet and obsess about tyres.

As F1 has moved from analogue to digital, the outcomes are almost as predictable as when you click-to-buy on your Amazon account and a box arrives a few hours later. F1 teams know precisely when their package will arrive at the chequered flag.

This is the central problem for a sport whose fans crave racing yet whose players strive every day to produce the opposite.