After the first race of the Formula One season in Bahrain there was a multitude of articles in the press and on blogs about how F1 was broken. This was quickly followed by an amazing race weekend in Melbourne where the complete opposite was seen.

Is F1 broken? Can races be made more exciting? These are the questions that seem to be continually asked. Part of the answer lies in the following statement: increase the likelihood of Random Events.

Lets start at the beginning. If you read the previous blog on First Order Performance variables you will see that engine power, aerodynamic downforce/drag and tyres drive the performance of the car.

Engine power is not really effected by any events that happen around them. As long as the engine receives fuel, air and electrical power, there will not be any random events. The only issues that can then effect the engine are failures, perhaps caused by errors such as cooling and ancillaries.

Aerodynamics can be effected by other events. Again, other than failure of a part, the next biggest effect comes from other cars disturbing the air that the car travels through. The only other real effect comes from weather. High winds can also disturb the aerodynamics, in a similar way following another car.

Tyres are probably one of the biggest changing factors in a race. They are highly non-linear. Their performance changes with a large number of variables making them very difficult to predict: track temperature/surface (which work to define grip)/condition (water!), heat input (driver style, car downforce level, engine power, etc).

Then there is the “software” element: the driver and the team! This is one the the largest areas for randomness. However, Formula One is the pinnacle of motorsports, so the level of skill is the best available, they don’t really make many mistakes (I realise this is a little controversial!): not compared to many of the series that lead to F1.

Teams in Formula One work very hard to reduce the risks in any of these events causing them problems. I think this is summed up in an interview with Ron Dennis in “Director” magazine (2009), ““When I came into motor racing so many things were a black art,” “But black art was a cloak for ‘we really don’t know’. It was intuitive engineering. I decided to make it a science. We will develop science to take away uncertainty to make winning a certainty”. This comment was reiterated when Martin Whitmarsh responded to criticism in Malaysia (2010), “The weather radar has been pretty reliable here. It predicted that it would pass through and we thought it would be dry by the end of the session. If you send a car out to do a banker, there is a risk”. McLaren focus’s on reducing uncertainty, sometimes it doesn’t pay off but you don’t notice the other 1,000 times that it delivers results.

An interesting part of this focus on reducing uncertainty is that smaller teams can focus on these events to gain! When I ran Super Aguri F1 Team, this is exactly what we did. Probably the best example was in Canada in 2007 when Takuma Sato famously overtook Fernando Alonso. This occurred because we were able to switch onto a different, non-standard strategy in order to get onto the better tyre. McLaren had to follow their predetermined best strategy. Why? Because in general, changing strategy in the heat of battle does not deliver results. In general reducing risk, sticking to a plan and focusing delivers results. But it certainly makes for interesting racing! So at SAF1, we planned specifically for these events.

Another interesting example is weather. At SAF1 we also knew that at a track like Bahrain, the weather would not vary much from year to year. In fact we would plot a chart showing the temperature over the previous 10 years, every hour, for the race weekend and 5 days past and previous: there were a number of tracks where nothing really happens year on year. With this level of statistical analysis applied to any variable that could provide a difference, it is easy to see how much effort is applied to reducing randomness!

Bahrain (2010) – predictable weather, temperature within predicted range (hence engine cooling, tyre choice etc all within the expected “window”), predictable buildup to qualifying (all previous preparation on simulations and simulators within the expected “window”) resulted in the grid order fastest to slowest: as expected.

Melbourne (2010) – unpredictable weather and temperature, unforgiving “walls” that punish small errors. In these circumstances it is easy to see that all of the preparation in the name of reducing uncertainty are thrown into chaos. With the race thrown into chaos, mistakes are made as the permutations and combinations increase the randomness of events.

Is it possible to reduce the preparedness of the teams? Probably not. Only incredible reductions in funding could achieve this. With more people looking at more problems in more detail (specialisation), the understanding that Ron Dennis targets is achieved. It would be hard to go back to the “old days” when the team owner also drove the truck and the race engineer was also the chief mechanic! However, this is where some of the racing series such as Formula Fords, F3 and others are. This is some of the reasons that racing is often more exciting in these series.

So how do we increase the chaos? Increasing the permutations and combinations of the event.

Can we predict the results of a race? In Bahrain, following qualifying, yes, we can be fairly accurate. The permutations and combinations that feed into the teams race simulations are not too far wrong. Baring mistakes by the driver or the team and failures, the variables discussed above are dominant: tyres. How well the tyres performance is predicted is related to how well preparations are made on Friday testing and previous testing. Now that testing has been banned between races the focus has moved to simulators: but it does increase the chances of errors.

This made me think of sports and games that have a large number of permutations and combinations. They are hard to predict. One thought I had was that games where the teams oppose each other, i.e. they have to go head to head. Like football or chess. The fact that the players clash when they pass through each others area on the way to the goal (or other winning area) results in a large number of possible interactions and hence permutations and combinations. Obviously in football, teams practice and practice formations and plays to reduce errors. In a game like chess, the number of possible different moves is constrained slightly by the squares on the board, but the theoretical number of outcomes is in the order of 10^27002 (http://answers.google.com/answers/threadview?id=389840) even though there is a maximum of 12,600 moves. First hypothesis: can we somehow introduce more ways in which the drivers interact more?

It is easy to see that tyres have a large effect and so many people say that the tyre war should be brought back into the sport. But the problem is this increases spending as the budgets will increase in order to minimise the variables associated with the tyres. Others want Bridgestone to make the tyres more on the limit, but this hurts their brand image and sometimes the tyres could be made too close to the limit and make it impossible to race! But effecting the tyres predictability will produce more variables. Second hypothesis: have races in places and at times when weather is more variable! (Like Melbourne and Britain!)

What about artificial random events like pace cars, reverse grids, push to pass buttons and different racing lines? Pace cars have become acceptable in Formula One, but reverse grids are seen as too artificial. Bernie Ecclestone recently suggested having 5 chances to take different lines around the track to pass people they are stuck behind. Go karts often have a number of races where drivers line up in a number of separate races where they get different grid positions each race and receive points based on the number of cars they pass and positions in each race. These points add towards the final race where these points decide the final grid. Many chances for the cars to clash in order to provide different possible permutations and combinations for the final race.

No real answers, but a look at many of the questions associated with randomness and how it effects the excitement of racing.

More thoughts to follow