As you can imagine, the GB Rowing Team schedule is very much like a full-time job. Training lasts for 49 weeks a year, from the middle of September to the end of August. During that time we generate vast and varied data sets. All of these data are discipline specific, but to make things simple we can break it down into three categories: Performance data, training data and health/wellbeing data.
Performance data is international regatta results and internal trials (ergometer and water). There are lots of elements such as intermediate splits, weather conditions, crew combinations to consider and track.
Training data is generally timed work, including any kind of varied efforts or interval sessions. It’s all recorded for time so that we can review, analyse and amend training schedules accordingly. It’s then stored on one database that contains data on every session we’ve run. We use this to calculate the time that we think it will take to win a gold medal – and train to that target.
Fitness, health & wellbeing involves monitoring the impact of training on the athletes. We collect blood lactate data, which helps assess intensity. In professional sport it’s not just about training hard, it’s about training smart.
So, obviously, we do collect a lot of data, but historically it’s all been stored in different databases; it’s very siloed. Therefore, our first stage is to move all of this into one place…that way you can see all you need to for any athlete, any time. You can also then take one athlete, one boat, one squad and look at it any time. This will be particularly important in the run-up and during competitions. You need to know if you’re pushing someone too hard or not hard enough. When it comes to a regatta, you need that information right there and then.
However, our biggest challenge is trying to ensure the current information is comparable with historical data, so we can look for trends.
We’re not just trying to achieve the best performance or find the perfect rower, though. Data can be used in a multitude of ways. For example, when assessing injuries, football and rugby teams have done a good job of trying to pick up and predict these before they happen. So, we want to do the same with our rowers. Generally speaking, professional rowers get quite a lot of back and rib issues that can have a lasting impact. What we need to know is: what are the indicators that make people more susceptible to injury? What puts them at higher risk? If we spot something that’s out of the ordinary in terms what we’d expect a rower’s stroke to look like from the data, it could indicate the onset of an injury requiring the training regime to be altered to allow recovery.
At the same time, we are also looking to predict future champions. Relatively, we’re not a big participation sport, so when people take it up, we need to be able to guide them and see that they’re on the right track. If we have the capacity to know if someone will flourish in a couple of years, even if their performance right now isn’t up to scratch, that could be invaluable in terms of our future success.
Also, we want to peak for the Olympics, but we can always do better at learning how to get people at their peak at the right time. The answer to this lies in combining the different types of data to determine what’s important and when.
With SAS, we feel that we can achieve all of this. Once we have all of our data in order and in one place, we will be able to mine and analyse data on individuals and groups. We can’t do that immediately at the moment. Over time we can start to build up training schedules that are more suited to each individual rather than everyone following a similar regime.
We’re also looking for the things that we don’t yet know. Are there some patterns that we haven’t thought about that could have a big impact? I’m sure that there are, but that’s the excitement of the unknown.
