Meet Peter Wibroe – the Data Scientist giving Denmark SailGP Team the edge

Jonathan Turner
22 April 2020

What does it take to make the boat go fast in a competition like SailGP? Well, world-class athletes definitely help, and a lot of practice and teamwork.

Peter Wibroe, SailGP

But every top sailing team knows that the real magic happens on shore – and it’s the team and preparation you put together off the water than can really make the difference when you’re on it.

Meet Peter Wibroe: expert Data Analyst and Denmark SailGP Team’s resident number-cruncher, who spends his time wading through SailGP’s vast amounts of data and reading between the lines, in search of every possible advantage.

Peter has one objective – to help the Danish team perform better. And in this game, with boats racing at up to 100 km/hr, even the finest margins can be the difference between winning and losing.

SailGP has a unique approach to data – and it’s quite simple. Everything is open, transparent, and available to view across the fleet, in real-time. When you’re being tracked by thousands of data channels, there are no secrets.

“As a new team into SailGP, we have a lot of catching up to do – and so we spend a lot of time looking at the data from the other boats,” explains Nicolai Sehested, helmsman of Denmark SailGP Team presented by ROCKWOOL.

“It’s unique that all of the teams have access to each other’s data, but I think that the best way to go about it is to share, as it gives us a level playing field.”

Just as ROCKWOOL harnesses data to produce leading and innovative solutions, the Denmark SailGP Team athletes use the numbers to constantly adapt and refine their approach to racing the boat.

But, when every team has access to the same information, how do you find a competitive advantage? The key, Peter explains, is being able to pick through the data – and strike gold.

Hi Peter! You have one of the most fascinating jobs on the Denmark SailGP Team – how did you become a data analyst, and what does your daily work entail?

Well, I am and have always been a bit of a geek. I love numbers. I have a background in Nano Medicine – a PhD from the University of Copenhagen. In that job, it's all about finding patterns in often some very large and confusing data sets. I love it and that was a motivating factor for going through an otherwise quite boring PhD!

That's more or less what this is all about. To be attacked by huge amounts of data, and trying to find sense in that. In SailGP, the beauty is being able to understand it and translate it into something useful to better the team’s performance.

SailGP is quite unique in that all of the data is open source – as a sailor who is passionate about numbers, it must be a dream job!

Being able to be work with such a complete and accurate data set that is open to all athletes across the sport, is quite unique. It's something that really allows our team to have an edge since we're new and we can learn so much and sneak in to what the other teams did last season.

Something quite fascinating but also frustrating is that we see everything in the SailGP data. There are so many sensors and measurements on the boats, and also on the marks on the race course, so we have everything – all the actions and settings that the sailors are doing, not just on our boat, but on the other teams. The next step is being able to understand it, and read it. It's like a new language – the answer to everything is in the data, but we just need to understand what language it's written.

Somewhere between a sailor’s gut feeling, and the numbers – that’s where the magic happens

How does that process work – do you have specific tools and software that you use to crunch the numbers?

It's a long road to be able to get to something useful, but it starts with access to a database which is kind of geeky, and then from there you need to get it into something more colourful. It's easier said than done. To do that, we use different kinds of programming languages to build our own software.

At the end of the day, the aim is to make it look very simple. If the sailors turn around and say to me, 'is that it?' then my job is complete. I'm kind of like the interpreter for the sailors – the link between the feeling, intuition and experience that the sailors have, and the hardcore numbers that we can extract from all of the boats. In that interface, between the gut feeling and the experience that they have, and the numbers that I can see, that's where magic happens. If we can exploit that, then we have a competitive edge on the others.

Fortunately, the athletes on the Denmark SailGP Team – aside from being really good sailors – also speak the analytical language. They respect and understand the numbers that I'm throwing at them. That's a pre-requisite for all this work and for us to be able to leverage the data available.

You have a competitive sailing background yourself, having raced alongside Nicolai for a long time, including in The Ocean Race in 2014-15. Does that practical experience as a sailor help or hinder your ability to ‘stay true’ to the numbers?

It helps a lot to also know how things go from the sailor’s point of view, and know the strengths and the weaknesses of what to put in the different numbers. You can make a sophisticated model that says that with increased wind speed, you go faster – but as a sailor, that doesn't really help you very much. You need to understand where they're coming from to get to something meaningful. You can make a lot of complex models, but it may not be of any value to a sailor who has been out on the water for the entire day and just wants to know if they should pull more or less on a specific parameter. It's about boiling it down to exactly what you need.

It sounds like trust is an important factor between the athletes and the data analyst. Would you say that having a solid relationship is important?

There can be some very brutal debriefs, where it's quite obvious – because everything is recorded, the numbers, the voice recordings etc – who did what, and therefore who made the error or the mistake. On the other hand, we still have some extremely good sailors that quite often have better feeling for what is going on that even the most complex sensors can detect, so it's a matter of respecting their way of seeing things and their observations. Trust is a very important factor in this, especially because it's important to be so open and honest, so we need to be able to bring everything on the table and recognise that we're in a learning process and that we make mistakes, we all do, and we just need to work together to figure out how to minimise those mistakes.

You have all of the data from SailGP Season 1 to work through – does that mean we’re furthering our understanding of the boat, even without being able to physically be out there practicing on it?

There are so many answers still lying in that historical data that we can dig out and find. One of our problems so far has been that we don't know what questions to ask – and if you don't have the question, it's difficult to get the answer. But during the racing in Sydney, we started discovering what to ask for and what to look for. Definitely there are many treasures lying there in that historical data, and in that way, this course can be an advantage to us and we can come out for the next season nicely prepared, with a good understanding of what has brought the other teams to the stage that they're at now.

Sydney was a bit of a crash course for the Denmark SailGP Team – a lot of work to get ready for the start line. Was there anything in the numbers that surprised you, or a particular ‘eureka’ moment?

We saw quite a few things in terms of how the other boats were setting up in their settings, but also in the dynamics of the different trim options, so adding that layer of how it has evolved over time was a big eye-opener for us. There are so many systems on the boat, and some of them can vary a little bit, so understanding those systems and monitoring those is also a good advantage for us.