WKO5: Cycling and triathlon analytics with Tim Cusick | EP#199


Tim Cusick is the TrainingPeaks WKO5 product leader and the owner of Velocious Endurance Coaching. Tim coaches world champion Amber Neben, road pro Emma Grant, and numerous other pros and age group racers. His expertise in the WKO software and its application in real-world coaching makes him the ideal person to discuss the ins and outs of the new and updated WKO software: WKO5.
Discuss this episode!
- Let's discuss this episode and the topic in general. Post any comments or questions in the comments at the bottom of the shownotes. Join the discussion here!
In this Episode you'll learn about:
- What's new in WKO5 compared to WKO4?
- How to use Aerobic and Anaerobic Training Impact.
- How to use Dynamic FRC (Functional Reserve Capacity).
- The Power Duration Curve and physiological modelling for running.
- Tim's coaching workflow in WKO5.
- Tim's top tips for athletes and coaches to get the most out of WKO5.
Sponsored by:
Precision Hydration
One size doesn't fit all when it comes to hydration. Take PH's free Triathlon Sweat Test to get personalised hydration advice tailored to what you're training for. Use the promo code THATTRIATHLONSHOW to get your first box for free!
ROKA
The finest triathlon wetsuits, apparel, equipment, and eyewear on the planet. Trusted by Lucy Charles, Javier Gómez-Noya, Flora Duffy, Mario Mola, and others. Visit roka.com/tts for 20% off your order.
Shownotes
Tim's previous episode
Why the upgrade to WKO5
5:53 -
- The goal with WKO4 was to bring a truly analytic engine to endurance sports.
- People have better access to data and information than ever before, we wanted to give people a tool to use this data in a custom and unique way.
- We've found that because the development of the custom analytics had a learning curve, it was people who saw themselves as a high end user that were leveraging the advantage of the software.
- With 5, we wanted to take the learning gathered from using WKO4, and make it a more user friendly environment so more people could access it.
- With analytics in sport, it's the everyday user that would benefit the most.
Someone who is already deeply educated and does a lot of analytics already may get a couple percentage points, but the everyday user may stand to gain much more. - We wanted to make sure it was easy to navigate, and had the great insights right at the surface.
- We're still adding more and more data in the endurance sports world - I recently worked with someone adding glucose insulin, which we can do in this software.
- We have packed in some great new features - some are really obvious from the start, but others are more important downstream.
New features in WKO5
12:49 -
- We've introduced new physiological metrics - we've modelled VO2max and some muscle fibre types and make better use of them in the system.
- There's a new Training Impact Score (TIS) which is a new way of looking at the impact of training on energy systems.
- We have introduced dynamic functional reserve capacity - a system to look at workouts, track anaerobic energy and how you expend it during a workout.
- One of the coolest functions is a dynamic multithreshold management.
In the multisport world you can't count on threshold being the same number across disciplines, so we have developed a system for that. - We also now have Smart Segments, which is a really fun feature.
- We've got some new training analytics including indoor outdoor modelling so we can better tell what's going on with your data.
- We're also now syncing subjective metrics.
If you're loading HRV and daily feelings, workout feelings, sleep etc into Training Peaks, we now sync it download these metrics.
You can then use them in any analytic, which is really powerful and has shown some really interesting correlations.
WKO5 changing coaching techniques
15:55 -
- I work with some pretty high level athletes, and life at that coaching level is always the razor edge.
You're pushing to extreme capability and you really need to do a great job and understand the dose response relationship. - For me in my coaching process I require athletes to write subjective notes post-workout.
It's been really helpful having this automated! - An athlete that spends time getting in tune with their feelings, the communication of their subjective data will align and predict the objective response.
Potentially more so than the actual data performance. - Once you match this with HRV and sleep, that whole picture really gives you a much better understanding of the timing of training.
- I often preach about how coaches focus way too much on one workout, and not the overall training rhythm.
- By adding more information to the big picture and look through the athletes subjective eyes, they predict when they should be training and when they should rest.
- It helps you hone the athletes skills to give you better feedback. You're using their performance data to improve their performance.
- I'm getting close to the point where subjective data is the data and the objective data will be the colour commentary.
- Data is a number based on whatever system you're using.
Analytics is when you can put a pool of data together, and that pool creates knowledge and a decision basis.
Sometimes the analytic is a one number thing, sometimes it's a group of data, sometimes it's a chart going up or down. - When I say subjective data, we're collecting it all: RPE, stress, sleep, HRV.
I'm getting all of the variables you couldn't get before, and we're doing it through analytics not just a data strain. - What I'm taking is perceived exertion, life stress, morning HRV, sleep, etc.
I have lots of subjective data focused on how well the athlete is going through a workout and recovering.
I then have two analytics based off the data that I look at in an acute relationship (2-7 days) and a chronic relationship (14-30 days). - You've got the give the athlete some guidance with this, and you can use the objective data to help quantify feelings at the start.
- Once you've established your analytic, the athlete predicts it - it aligns with sleep, or HRV patterns - not always both.
There's a correlating group of data that goes from green to yellow to red, and it predicts training workout and short term training rhythm success.
Training Impact Score
24:21 -
- The Training Impact Score (TIS) is a great example of using insights, where we're bringing data to the surface.
- We make energy in two ways: aerobically and anaerobically.
WKO4 and 5 have a robust power duration model, and it's a human performance model.
When you have this type of model as a basis, and the individual athlete goes out a does a workout, the power duration curve is two curves: aerobically and anaerobically. - Since we understand how to make energy, when you go out and workout we can understand which of those two systems you were impacting for that workout.
- You can begin to measure how the workouts that you do strain those systems.
- Stress is external, and a training stress score is just how much stress you apply to a system.
When you put stress on a system it goes under strain. This is the internal response - what happens to your body under that load.
It is strain that leads to adaptation. - You can apply the same stress to five different athletes but it doesn't mean they go through the same strain.
They may have different muscular metabolic system and make energy differently.
They therefore won't have the same TIS because it's specifically measuring strain. - We've taken that system and we've simplified the scoring and it's now 1-10.
- 1-3 means you haven't done enough work to have any impact.
4-7 means some varying form of maintenance.
8-10 you're putting load to put quality stress on that individual system. - It gives feedback on the system that was impacted, and insights into future effective workouts.
For example if you do a 3 x 20 min at sweet spot workout and you do it 10 watts too hard, you might be generating a lot more anaerobic energy than you thought.
You go out the next day and do it 10 watts slower, and you can see aerobically it was really impactful.
It helps quantify the output.
Dynamic Functional Reserve Capacity
29:41 -
- This isn't the same as maximum power available but it's similar to this metric.
- This is your anaerobic work capacity generally - Functional Reserve Capacity (FRC) is different because it takes into account the small amount of aerobic contribution when you're over threshold.
- The simplest way to think of FRC and Dynamic FRC is the measurement of how much continuous power you can do over threshold.
It's measured typically in kilojoules and joules.
To simplify: it's your anaerobic battery. - Go really far over threshold, you're draining the battery fast, go a little over and you're draining at a controllable rate.
- Dynamic FRC is an optional piece of data - you can see over the course of a workout, or a series of workouts, how quickly you're draining that anaerobic battery.
You can also see the rate at which it restores. - Dynamic FRC is starting out with the full battery (e.g. 20 kilojoules).
You go out and do a 5 minute near maximal effort, which might drain your battery down to 5.
You then rest for 3 minutes in between which might restore it back to 12.
Then the next effort might drain it down to 1 because you didn't filly restore the battery. - If you're using Dynamic FRC as an observation of intervals or short intense work, or even longer harder races, you can see how you were burning down and the effect over time.
- Dynamic FRC is based on the idea that we have an anaerobic battery and it's more limited than our aerobic system.
We can get the burn rate down pretty easy, but it's hard to quantify the recovery rate.
Ours is now good observational data but I don't think it ever would be or could be perfect. - If you went out today and did 4 x 5 minutes near max effort with 4 minutes rest in between.
We can score the burn rate pretty well.
If you do that well rested having eaten well, the burn rate is the same but your recovery rate is likely good because you entered it well rested.
For the next four days you don't ride your bike, you don't sleep or eat well.
Repeat the same workout, the math would say your recovery should be the same but it would actually be pretty rare for you to do as well in that workout because your personal training environment hasn't been good. - We think we have enough data to begin to get these things down, but the reality is to understand recovery rate you need to know exactly what the athlete is doing at all times outside that workout.
- Subjective data begins to connect these dots, and increase the precision of metrics like this.
- By understanding and using these metrics you're putting together data to make an analytic, and then using this to prescribe and predict a performance.
- If an athlete goes out and says they found it really hard but the session output doesn't agree, I go back and check other factors.
Was the life stress too high maybe.
However, if I don't see those, maybe they're just not good at going hard or they're not mentally ready for that.
You then might need to do some work on that side - using the data for example can be really helpful.
Multiple thresholds
39:04 -
- In the last 6 months development of WKO4 running power began being measured, so we knew we wanted to have this in WKO5.
We needed to spend time vetting and testing, and making sure it was applicable to the analytics. - Running, cycling, swimming and rowing are the core ones we see activity in.
FTP or threshold is not the same across these disciplines.
Even what is recorded to generate the power is not the same, so we had to re-invent power analytics for each sport. - In cycling we had honed it already and it's a very controlled sport, but we had to deal with all the variables in the other sports and we feel this now works well.
- We have the universal - a sport type selection with different sport types and you can individual pick which ones you want to see.
- Or you can hard code it as you build your own custom analytics if you choose.
We've given really good sets of these so you shouldn't feel you need to, but if you choose to you can do it there.
With sports specific you can use the sport picker. - When you do different disciplines across different days you're not using the same threshold and you're not generating the same threshold score.
You're doing it off the individualised threshold of each and they'll simply interact and work.
Kevin Williams - the driver of WKO - calls it 'automagic'. One setup will manage all the data individualised to each sport. - The power duration model that we use, designed by Andy Cogin, is highly accurate but it's a highly accurate estimate and needs a certain amount of data.
It needs one hard short effort, one hard medium effort and one hard longer effort. - Most runners involved in endurance sports don't do much sprinting, but if you do add the sprint the model works well, but it's not perfect.
It doesn't know what you can do, it takes what you've done and estimates what you can do. - Some of the best coaches in running mix the power duration curve and the CP curve, which might be a better use for runners.
You can run that in WKO! - I personally would run both the power duration and the CP model and use both as the basis for my analytics.
The CP model is probably a more appropriate tool than the power duration if the athlete doesn't want to do a 100m sprint. - If you have an athlete running who will give you a 100m sprint every once in a while - say 12 seconds - it will give you good numbers.
- You can now, in WKO5, make recommendations for what durations to test to get an even more accurate power duration model for estimating.
This is in the insight section, and it even gives you a power to exceed rather than just a time. - The model does a good job of estimating what you can do it just can't predict.
Smart segments
50:35 -
- Kevin Williams and I were doing a bike ride and we were doing climbs, we had a friendly competition and we thought after it would be cool to have all the analytics of that.
We came up with the idea of doing smart segments! - A smart segment is where you do a workout and you have your local climb - for example Alpe D'Huez.
At the bottom you hit your lap button and the same as the top, so that climb becomes a lap.
When you load that into WKO you can save it as a smart segment. - Then, every time you do that smart segment WKO will search and find it and it'll show you all the times you've been up there.
In the simplest version you can then rank by speed or by power.
You can then see where you do better and worse. - You can then use the WKO compare mode and build custom analytics to look for trends of when you perform great - e.g. cadence, power trending up, going harder at the bottom etc.
- You can do the same on any geographical track - most of us train in a few specific locations.
You can then create a dashboard to particularly analyse that segment if you wanted too. - I was over in Harrogate in England and we were pre-riding the course. It's highlight unique because it has three distinctive areas, and in the middle one that also has three distinctive areas.
We rode it day 1 just to experience it, I marked my laps and saved them as smart segments.
The next day we rode it closer to race pace, and then looked at those numbers - we learned something between slow and fast.
We then ran the course and focused on each segment, and by the time we were done we had 10-12 of each segment which meant a whole lot of learning!
Tim's coaching workflow in WKO5
57:58 -
- I have a three step process whenever I'm interacting with the athlete's data - whether it's writing a training plan, analytics or preparing for a meeting.
- I review, I make decisions, I create solutions.
- I don't coach a lot of athletes because I often go quite deep and each athlete can consume quite a lot of time.
- I start with a review of key performance indicators (KPI's). I want to surface what I believe are the key things I'm looking at.
For me, I want to know compliance - not in a penal way, just whether it's what I expected or not.
Then I look at performance in the objective format.
Then I move onto subjective data - I read the notes and look at the subjective scoring. - If all the KPI's look good I don't go any deeper unless I'm specifically looking for something.
But if there's something that alerts me either good or bad in the KPI's, I begin the drill down process. - That leads me to step 2: making decisions.
If the KPI was bad, e.g. compliance was down, notes where 'wheels came off' I now need to gather information to make a decision.
It's not data science, it's decision science. There is no perfect solution or ideal model, we can't measure all the necessary variables for that. Data improves the odds of correct decision making. - If they had a bad workout I start looking at data - maybe it was really hot out, or the athlete was having a bad day.
My decision might be 'this is a fail but let's press on'.
Or maybe the environmental was right but the subjective data is bad, or the HRV is all over the place.
Then the decision might be to pull the plug for a day, rest is better. - Or maybe they went above what was expected because they're feeling great, and then the decision may be to add more load.
- If we follow through with these decisions but in a few days we're still not fixing it, then we need to go deeper into the data to try and understand the solution.
For example it may mean going to the doctors to check they're not sick, or find out how they're sleeping and what their recovery looks like. - If you have an athlete that you're just starting to work with and you get them to do 2 x 20 minutes at threshold.
They have a local climb and do their 2 x 20 at 300 watts.
The next week everything looks good and you review and continue, so they do the same session but on the flat and only generate 280 watts.
A decision might be they're not feeling great and they're fatigued, a power drop off is normal.
A solution would be because they have a different pedal stroke when they climb compared to flats means they can't produce that power on a flat, so in future if I want that power I need to put them on a climb.
Or the solution might be, most of their performances are on the flats so I need to try and adapt this athlete. - The solution is most related to their goals, and it's something you're putting in play longer term.
- Whereas a decision might be a short term thing such as rest for a day.
- When you have a training strategy, a solution means you have to make changes to that strategy, but a decision means you make changes to my tactics.
Top 3 things on WKO5 for time efficient coaching
1:07:17 -
- We put a set of good basic views that should give you 99% of everything you need.
- What you should do when you first bring data into WKO5, look through the different views and find the one that matches your athlete.
Look through it, click through everything, and see if you've got everything you need. If not, try another.
Once you've found one, start using it and stick with one at a time. - Then read the insights dashboard.
It's always changing, and it's things we can feed to you as the user. Any new science and features can come through this way.
We're trying to give the best new science in actual insights that can help your coaching. - Join and ask questions in the Facebook Group! It's a polite and interactive group.
Most questions have been answered so always search first, and if not ask away and you'll get a plethora of views.
Facebook Group
Learning more about WKO5
1:11:18 -
- On WKO5 there's a question mark in the top right, and this will launch the quick start guide.
This is a series of walk throughs with screen shots to help you get rolling. It also has short video tutorials. - It also has a link top right called 'education centre' which has five key categories, and it's true learning for the software.
Each has a recorded webinar that lasts an hour, and also has follow on videos.
1. Getting start with WKO - how to maximise a setup early on.
2. Individualise your training - benefits of a power duration model, and other things.
3. Training and coaching with WKO - that's literally the recorded process that I use. It has an overview, initial testing, strength and weaknesses etc - all in individual videos.
4. Track and monitor performance - insights on how to use ongoing data.
5. Creating analytics - Kevin does a one hour webinar of how to create your own, and we roll out individual videos on the different steps. - We've got a much better support centre now in WKO5 because we've learnt from WKO4.
WKO as a tool for coaches and self-coached athletes
1:16:12 -
- One of our main goals was to give the self-coached and everyday athletes a tool - that's why there is a lot of focus on the user experience.
- The demand to use it for the individual athlete is high, which is why we've tried to reduce that learning curve.
- An individual athlete should probably use WKO5, and will find it easier than WKO4.
If you're totally new to training with data you might want to start with TrainingPeaks premium and then WKO, but WKO5 should be totally fine for an every day athlete too. - It's the same experience for coaches and athletes, but because coaches often invested the training in WKO4 they're still the masters.
- I think the best relationship is when both the coach and athlete are using WKO.
Accountability of both the coach and the athlete is the pathway to success, and this synergistic cycle improves relationships.
- Example of an amateur athlete with two peaks in the year, so they have two 16 week macro cycles - 12 week for capacity development.
Key takeaway
- Moving from data science to decision science is important - data is only valuable if it helps you make better decisions and find solutions to problems.
- Set up a workflow that works for you, that allows you to make decisions.
Tim's workflow was review, decision, solution. Yours doesn't need to be the same, but it's important to have one in mind so you know why you're doing what you're doing. - Subjective and wellbeing metrics are so important e.g. HRV, athletes view of the workout, sleep etc.
Links, resources and contact
Links and resources mentioned
Connect with Tim Cusick
Connect with host Mikael Eriksson

Hi! I'm your host Mikael,
I am a full-time triathlon coach and an ambitious age-group triathlete. My goal is podium at the Finnish national championships within the next few years.
I first started the website Scientific Triathlon in autumn 2015 as a passion project to share my learnings with a larger triathlon audience. Later on, in early 2017 I started the podcast That Triathlon Show.
I sincerely want you to contact me to
- Send me feedback
- Give constructive criticism
- Request topics and guests for the podcast
- Send me your triathlon-related questions
- Tell me that you've rated and reviewed That Triathlon Show so I can give you a shout-out on the show and tell you how much it means to me!
MORE ON THAT TRIATHLON SHOW
- Go to the That Triathlon Show main page
- Go to the full Episode Archives
- Check out these popular episodes:
Hi Mikael! You and Tim mentioned that you can make recommendations for intervals/unstructured tests to help update the power duration curve. Is there a specific chart that explains that?
Hi Jake,
Yes, it is the “Normalised Residuals” chart.