Using Tableau to Make a Swimming Season Review Dashboard
- Zack Killoran
- Jul 3, 2020
- 8 min read
During my time interning at the Western Australia Institute of Sport (WAIS), I spent most of my time working with the Elite Swimming Squads. These athletes are WA's best, and all en route for challenging for Tokyo 2020 Olympic selection. Obviously, the Olympics have been pushed back a year as a result of the pandemic, giving the athletes another year to prepare themselves. I am excited to tune into the Olympics next summer and hopefully see some of the athletes that I have worked with representing their country and possibly even winning a medal or two! All of the WAIS staff and athletes that I came across during my year were great people, and helped to make my work incredibly fun and exciting. The Podium Department that I was a part of, and the swimming squads were a huge part of this, so thank you to all those practitioners, coaches and athletes involved.
The main bulk of my time was spent creating and maintaining a database for the swimming squads, which included various factors associated with their main training sets and their key performance indicators from their Strength & Conditioning sessions. Main sets were always performed in race suits. I would be on pool deck for most sessions, and on main sets I would record times, stroke rates and take lactate tests. These are all used in combination to monitor the athlete's progression over the course of the season, with the hope that they are becoming fitter, stronger, more efficient athletes.
Although I did not get to see out the season, I have created some fictional data for 'Athlete X', to show how I imagine the database could finally be brought together. Athlete X is competing in the 50 m breaststroke (I know that this is not an Olympic event, this is just to further ensure that data is completely fictional). This is the first opportunity that I have had to apply what I have learnt during the Tableau Analyst course to my own work.
The Dashboard
I have used four worksheets to create this view; key lifts (as determined by the S&C Coach), race performances (with extra details provided by the Biomechanist and Coach), step test performance (as ran by the Physiologist and Coach) and the fastest training times (from main sets prescribed by the Coach).
Title, Logo and Main Filters
The title has been made so that the name part will change as the name of the athlete being assessed changes. The logo is a generic swimming logo that I found from google, which of course would be substituted with the club/institute's logo. Next to that we have a single list filter to switch between the athletes within the squad that we are looking at, and their stroke, if applicable. At WAIS there were three High Performance Squads, so if necessary there could also be a coach/squad single list filter here too for easy movement between athlete reports. Usually, there would also need to be a race distance filter here.
Race Performance
Ultimately, race performance is the most important metric to assess how well a season has gone for a swimmer. Therefore, I have put this in the top left so it is the first thing someone may read, simply including the competition and times.
I used a calculated field to add colour here, suggesting that Athlete X had a target of 29.00 s for their races this season. Rather intuitively, this shows the times in green if he met/beat this target, and in red if he did not. This is explained in the caption also, by stating the target time for the individual. It is important to ensure that this field updates for the individual.
I have included other variables in the tooltip, so when you hover over each time with the mouse, it will show race variables such as reaction time and stroke rate. These do not need to be in the view as the Coach and athlete will have already gone through race reports many times, so likely even remember what was good or bad about each event. It ends up cluttering the report if included and is better as being an added part of further discussion, in my opinion. Of course, these variables can be improved upon as the Biomechanist will produce a much more in depth report, including factors such as skills (dive, free swimming and turn) times also.
We already have the athlete's target time in the caption, this could be further improved by also including times such as their own Personal Best or World Top 8, or World Record if relevant. This would be done to how the Coach/athlete/Sports Psychologist feels is best for that individual.
There could also be a link for each race to take you straight to the race footage also, to be able to immediately go to that and discuss performance factors associated with it.
Key Lifts
This section would be determined by the S&C Coach. For now, I have kept it extremely basic, just showing the maximum weight lifted for three key lifts; pull up, back squat and trapbar deadlift. There could be various other factors included here, for example we took monthly countermovement jump and isometric mid thigh pull readings. The most important variables to measuring performance of a swimmer should be included here, using variables supported by the literature.
In the tooltip, I have included the date that this season's best was recorded, this can be used to assess at which point during the season the swimmer peaked physically. At the start of the season they might focus on generic strength, before moving onto more of a power focus at the backend. Therefore, the variables here should be specific to race performance, and the dates are likely to show whether the plan has been executed effectively, showing good use of periodisation in the training programme.
Step Test Performance
A step test can be done at several intervals throughout the course of a season. By completing the same set (e.g. monthly), the Physiologist may track changes in fitness levels (i.e. by the athlete producing less lactate at the same swimming velocities, showing a greater use of the aerobic energy systems). The Coach and Biomechanist could also look at some performance variables during this such as stroke rates. As with all scientifically valid tests, the test must use repeated measures, meaning the same warm up and step test are used, if possible other performance related factors should also be controlled e.g. diet, fatigue, etc.
In the dashboard, the caption explains what the test was and when the measurements were taken. This is just an example test, it should be designed specifically for the athlete's event. We see many data points being used for each date that the test was completed on. The fastest times for the 50 m breaststroke is shown through the solid green circles, and the corresponding maximum lactate by the end of that set is shown by the orange bars. I chose this format as a line graph for each doesn't clearly show when the test was completed, and the minimal changes in the time are hard to see with a line. The bar graph allows the two to be easily discerned from one another and also clearly shows large differences in lactate production. I think using different types of marks shows that they are different metrics, but placing them on the same graph shows that they are related. The axis are not synchronised as the scales are vastly different from each other, this allows the whole space to be better used.
The tooltip again is used to give further context to each datapoint. The lactate and time mark share the same details on the tooltip for matching times. This is likely to be the most reliable source of changing training performance throughout the season due to the more scientific nature of the test.
Fastest Training Times
In this section, I would only include main sets where the swimmers are wearing race suits. It is important that the data collection is standardised, so the same data is collected week-on-week. Although coaches will prescribe different main sets for most sessions, there will be certain similar themes or qualities to sessions. It is important to be able to see these, or discuss with the coach to help you identify these. Once this is done, similar sessions can be grouped together. For this dashboard, lets assume that all recorded data comes from Race Starts (i.e. diving off the block for freestyle), and the instruction is to go at '50 m race pace'.
We can see a clear trend with decreasing times over the course of the training season, which is obviously what you would hope for, peaking later in the season for the main races. As our athlete competes in 50 m breaststroke, the only training times the coach would feel relevant to assessing their performance is across 25, 35 and 50 m distances. This can be filtered according to what is relevant, or to get greater insight on just one distance to understand where their strengths and weaknesses may lie.
Each distance is coloured differently, with the corresponding legend shown next to the graph, and the selection of distances shown underneath. Arguably, we don't need to include the legend as those involved with the sport would immediately understand what times correspond to which distances. I used two different mark types here, to be able to show each session with the solid circle, and then connect each using the line, to clearly show when each session was and the decrease in times. It is beneficial to have both as we are working with tenths of a second, so it is hard to see when the different sessions were and when improvements occur. Again the tooltip can offer greater insight into the session, the most beneficial part here being the set description. Often main set descriptions are long and difficult to understand for those who aren't on pooldeck every day, so it is important that abbreviations are kept constant throughout the database and that key stakeholders know what is meant for each set description. The benefit of including the set description and total main set volume is that if the fastest repeat distance was to be much slower than usual, for example, we could look here and understand why, realising that perhaps this session was much harder than we originally thought when the coach planned it. The slower speed doesn't necessarily mean that the swimmer is injured, or not trying, or lacking fitness. As always - context is king.
If there is enough data collected, with clear definitions of the different intensities that the athlete is being asked to swim at, we could even go into further detail to understand their training performance for different types of main sets, such as those with a lower or higher amount of rest. Buzzwords such as anaerobic power, lactate production or aerobic power might be used for separation here.
Overall, coming into the sport, I knew little to nothing about swimming, but feel grateful to have been given the opportunity to learn from so many great people within the sport. To now be able to talk about it in this post and feel confident that I am speaking a language that those involved within the sport would understand, with it being to a technically sound level, is possibly the greatest achievement that I have made during the year. Ultimately, I would not be able to bring anything to a team if I couldn't communicate in a way that coaches, athletes and other practitioners understand.
I think that this dashboard can be adapted to any level, and indeed various other sports also. As part of that adaptation, it is important that it fits the needs of the organisation, this may also include various other details, one being overall training load. How this would be presented is up to people such as coaches and Performance Directors. I think weekly swim load could be included in the tooltip of the training sets, but also could be within its own graph to show further details, such as how many metres completed at each different intensity.
Again, ultimately it is up to the discretion of key stakeholders, again, context is king.
Comments