How Haas reveals Formula 1's biggest problem

In Mexico last month the strongest Haas driver in qualifying was Esteban Gutierrez, who lined up 17th, 2.697 seconds (3.427%) off the pace. Two weeks later in Brazil Romain Grosjean qualified seventh, 1.201s (1.698%) down.

Those figures encapsulate the biggest problem Formula 1 has as far as fans are concerned. There's simply not enough of this kind of fluctuation in performance to deliver the surprises - like Haas struggling one race weekend, flying the next - to keep those watching entertained.

Unpredictability is the lifeblood of any sport. In F1 of late it's been relatively easy to predict the identity of the drivers who will reach Q3. Over the season so far, of the 200 Q3 slots available only 17.5% have been filled by drivers not in a Mercedes, a Red Bull, a Ferrari, a Force India or a Williams. And the majority of those have been achieved through exceptional performances from Fernando Alonso and Carlos Sainz Jr.

Standard deviation is a handy way to quantify such variability - or lack thereof. To put it simply, it measures the variation between the average and the individual data points in a set of figures.

So the higher the figure, the greater the variability there is in the performance figures - the larger the 'spread' into which a car can slot on a given weekend, if you like.

Taking the fastest individual lap set by each car during every grand prix weekend, expressed as a percentage of the fastest single lap, allows us to look not only at the overall mean performance, but the extent of the variability.

The figures show that Mercedes, unsurprisingly, has the lowest standard deviation, with Manor and Haas at the bottom of the ranking.

STANDARD DEVIATION (PERFORMANCE)
Mercedes 0.05143
Ferrari 0.38616
McLaren 0.42569
Toro Rosso 0.4384
Williams 0.50733
Force India 0.54603
Red Bull 0.58102
Sauber 0.59725
Renault 0.676
Haas 0.84252
Manor 1.05987

To put this into context, around a track like Abu Dhabi, where F1 will race this weekend, the standard deviation figure for Manor indicates a window of around one second into which it might slot.

For Haas it's around 0.850s, while Mercedes is basically going to be fastest whatever happens. But most teams are, roughly speaking, clustered around the 0.4-0.6s region.

The average of the standard deviation figures is 0.556. Now, assuming the distribution remains broadly similar, it stands to reason that the higher the average standard deviation figure, the more varied the results will become.

Each car is subject to the laws of physics, so it's fair to say it has an ultimate potential that can be extracted on a given weekend. That will vary a little from track to track based on circuit characteristics and the impact of car upgrades, but it's the job of the team and the driver to extract as close to 100% of the performance potential as possible.

The standard of the drivers today is so high that, for the most part, the fluctuation in their performance is limited (indeed, this plays a bigger part in the fact that we often see drivers lined up two-by-two and overall results being dictated by the car than the claim that F1 cars are too easy to drive today).

Likewise, preparation and manufacturing standards are such that, except in the case of a problem, the cars' build quality is not as big a limiting factor as it once was.

And this is where we come back to the Haas 'problem'. The other factor dictating how much of the potential is realised on a given weekend is the performance of the team. To be precise, the ability to start with a set-up that is firmly in the ballpark, and troubleshoot problems.

Much of this is achieved through analysis of the data, an area where Haas, by its own admission, has lagged behind many of its rivals. Data is often seen as some kind of magic bullet. The process is seen as a straightforward methodology of 'drive about in session, look at data, see solution, fix it'.

The truth is far from being that simple. In fact it requires a huge amount of resource and understanding to drill down into that data, understand what's significant and to draw the correct conclusions on what, if anything, to change. And when you have a gap of, say, a couple of hours between final practice and qualifying, that's a lot of work to do in a very short time.

The reality is that most teams are far too good at this to deliver the kind of unpredictability fans want. As a start-up team, Haas is behind in the area of data analysis. It is gathering the data, it has excellent people, but it just takes time for the processes and techniques to get to the standard of long-established teams. Doubtless, Haas's performance-figure standard deviation will become lower next year.

It stands to reason that the larger the performance envelope a team generally operates in, the bigger the fluctuation in lap time and end result. And if every car has a larger envelope, there is greater potential for overlap and for a change of order. So what F1 actually needs if people want more unpredictability is for teams to become less good at this process of data analysis.

That doesn't necessarily mean this should happen - and even if you tried to make it happen, such skills are not simply forgotten. But what it does do is highlight that the regulations, often blamed for all F1's ills, are not to blame. At least not in terms of the detail.

The only way to, by regulation, guarantee such variability is to introduce the kind of handicapping system that proliferates in many other championships - usually ones that have to work very hard to get people to pay attention to them.

Currently, F1 has very restrictive regulations. And for good reason, for this limits the benefit that can be had from bigger budgets and makes it possible for teams on lower budgets to be closer to the front.

Most people tend to suggest proscriptive means to try to make the cars, the racing, the show - whatever your phrase of choice is - 'better'. But that is a path that has consistently failed because of the law of unintended consequences.

This is why F1 tends to spiral pointlessly around, arguing that freer regulations are needed but that rules must remain tight in order to control costs and prevent the biggest teams pulling away.

Then, the argument goes, free rules allow smaller teams to be cleverer and punch above their weight. But if the big teams can try more, given how knowledgeable everybody is about what makes racing cars work, generally that means they can achieve more.

But there is potentially one way to have the desired impact. You just need to create the conditions in which it's possible for the kind of F1 people want to flourish. Pat Symonds, chief technical officer at Williams, has long advocated a simple solution.

"Budget caps and freer regulations," he said in Brazil last weekend. "I haven't changed my opinion on that in the last seven or eight years.

"Technically, it's funny to look at the old regulations. I think the oldest set I've got is 1983 and it's just amazing. You look at them now and think, 'I could do whatever I wanted'."

So for the technical regulations, a cut-back rulebook with more freedom, but with a spending cap, is appealing. If, that is, it can be enforced and all those involved are willing for it to be policed properly. And there's the rub.

There's also the major downside that, like so many of F1's cuts, it would strike at the heart of what might be termed the 'ordinary people' whose living is based on grand prix racing, because headcounts would go down.

Again, it brings it back to the end point of all such debates - what does F1 want to be? If it needs to become more unpredictable and allow a more level playing field that is financially stable, this is potentially the path to achieve it.

If that's the path F1 wants to take, whether it can be implemented and policed is another matter. But what's important to recognise is that more tinkering in the detail of the technical regulations will not have that effect.

As usual, the key is to work out how to create the environment that will result in the outcome that's desired, rather than trying to make it happen by fiddling with the detail.

Then again, this is pretty unlikely. After all, just look at the sporting regulations. Following the multiple controversies in Mexico, we've had many key players in F1 demanding - simultaneously - fewer regulations on driving standards and a clearer understanding of what is and isn't allowed. Given the need to make stewarding decisions that would stand up in a court of law, that simply doesn't work.

As is often the case, mutually exclusive positions too often become unified in F1, so nothing gets done.

The bottom line is that, when it comes to wanting greater variation of performance, F1 teams really are so good - and this is to their great credit - that they know too much.