What is written below correctly describes the Planner functionality but there are a couple of new updates since this was written, namely:

  1. Forecast method. For athletes for whom Crickles has uploaded at least a year of data there are two new Forecast methods for projecting your exercise load. The method described in the post below is now called Average based on Date range. The first new method is called Trend. This takes your average CSS per day from exactly one year before the forecast period (which is from now to your chosen Target date) and scales this up or down according to your current v. last year prior exercise load. The advantage of this over the Average method is that it accounts for both seasonality and year-to-year trend, which are typically very significant factors. Like the Average method, this generates ‘per day of week’ exercise load that you can adjust to suit your training programme. The second new method is called Repeat block from the past. When you choose this you’re prompted to enter a Start date to repeat from. Given this, the actual exercise loads that you incurred from that start date are projected forward from now to your Target date. For this method the adjustable per day of week levels do not appear. This enables you to see exactly how your Fitness, Fatigue and Form will evolve between now and race day if you repeat a training block from the past.
  2. Taper periods. The default taper periods described below were taken from analysis of the tapers of successful marathon runners prior to a marathon. These allow for a substantial amount of recovery and lead to very high Form levels on race day at a cost to Fitness (which is the point of tapering). These defaults have been changed to levels that are recommended for (and typical of cyclists) – a two week instead of a three week taper and of more modest extent. Of course, you can freely set the Taper levels as you see fit and this only affects the defaults.


The Crickles Fit-Fat tab shows you how your Fitness, Fatigue and Form have evolved based upon the Crickles model for Cardiac Stress. Now, you can also use the new Crickles Fitness Planner to anticipate how Fitness, Fatigue and Form will respond to a training programme geared towards an event. This is accessed on the Planner tab.

Entering the data

When you first visit the Planner you’ll see that the sidebar changes to the following:


Target date defaults to today + 3 months. You should change it according to your fitness planning horizon – say the date of your next significant athletic event. (It won’t accept a date less than one month into the future.)

The Sundays to Saturdays fields in the CSS Regime section are populated with your average CSS (actually your ‘extended CSS’ or XSS) for each day of the week sampled over the Date range. In this screenshot, I’ve extended my Date range back to 01/01/2015 to get a longer-term sampling period. You should adjust these numbers to reflect your planned training programme over the period up to your event. It’s best to do this using typical observed CSS values from the Activities tab. For example, in my case my usual park loops on a Wednesday or a Saturday park run will each score me about 40 CSS points; an hour on the turbo will typically get me 80; and a weekend club ride or equivalent might score about 320. Of course, if I want to increase my Fitness I’ll have to do more than I’ve done in the past!

The three Taper period fields enable you to specify how you plan to step down your exercise in the three weeks leading up to the target date. The fields are pre-populated with typical percentages – e.g. a reduction to 70% of a normal training week in the third week before the target date – and you can change these if you wish.

As you change all of these values the charts will change accordingly.

What you should aim for

The general aim of a Fit-Fat-Form training programme can be summarised in three principles:

  1. Add training volume to increase Fitness
  2. Taper before your event to flush out Fatigue and hence increase Form
  3. Don’t take on so much exercise load that your Fatigue gets to unhealthy levels.

Of course, there are many other important factors too such as working on strength and flexibility, managing any medications you may be taking, and balancing non-exercise factors such as diet, workload, travel and sleep; these lie outside the scope of this tool.

There is no established science on correct or optimum levels for Fitness or Fatigue. However, you can see if your own Fitness is increasing, staying flat or falling. Regarding Fatigue, the charts show you two useful levels as horizontal dashed red lines:

  • What is a high Fatigue level for you. This is calculated from your 90th percentile of Fatigue over all of your data on Crickles.
  • What constitutes a high Fatigue level for all of the athletes on Crickles. This is calculated from the 90th percentile of the 90th percentiles of each Crickles athlete.

Each of these two levels is marked on the Planner chart. As a fact of statistics, most of us will find our own high Fatigue level lies at a smaller absolute value than the high Crickles Fatigue level. When you train hard and do tough sessions you will occasionally push yourself beyond your own high Fatigue level. However, when planning a training programme if you see that your Fatigue levels are sustained for several days at a greater negative level than your high level, and especially if they are also sustained beyond the Crickles high level, it would be prudent to build in more recovery days.

So with reference to the three aims of the Fit-Fat-Form programme given above, you want to see that:

  1. Your Fitness at the target date is higher than it is today
  2. Your Form at the target date is comfortably positive – ideally at levels of 25 or more for the days leading  up to your event
  3. At no point during the course of the programme is your Fatigue hanging for days on end below the high lines, and especially not below both lines.


The longer you give yourself to train, the easier it will be to meet all three of these training goals. Over three months it is certainly possible, and over a longer period it is easier. To take my numbers given on the input panel above as an example, if I project forwards based on no change except for a taper period I’ll find that my Fitness at the target date is lower than my Fitness now, while my Form will be much higher/better: both effects are simple consequences of the taper period. However, if I increase my average exercise load on Sundays from 131 to 250 while leaving all other values the same I’ll get this:


My Fitness increases by about 10%, my Form is nicely high leading up to the target date and although my Fatigue sometimes dips below my own “high level”, it never stays there for long; moreover, my Fatigue never falls below the “Crickles high fatigue” level set by our hardiest athletes.

A final thought

There are limitations to any model and, as I’ve noted, there are important health and fitness factors that are not addressed at all by this kind of model. However, the Crickles methodology for assessing cardiac stress is robust and can be applied consistently to any form of exercise (although the estimate quality is reduced when heart rate is not captured). Furthermore, if you set a CSS-based exercise programme using this approach and stick to it then – as a mathematical fact and not just with the vagueness of an estimate – your Fitness, Form and Fatigue on the target date will be exactly as they were projected to be at the start of the programme!



  1. Ian,
    I’m not sure what I’m doing wrong but whatever I do I seem to end up less fit in July than I am now! I’m using the last years CSS to populate the tool. Even if I pick a very short period of the year where I did quite intense riding the improvement in fitness is minimal with a big drop off in the tapering. It’s hardly worth me trying if I look at this. Yet if I use a PMC chart in Golden Cheetah or the fitness and freshness in Strava I can see that from where I am now I can make significant gains. Does your tool use my Power numbers or HR?

    1. Hi Nick, Thanks for bringing this up. What I’d suggest is (1) start off with no taper – set the three weeks of taper all to 100 to begin with; (2) pick a Date range where you were most active and that most closely matches your likely Spring/Summer training load for this year – if no historical period matches it, pick the closest (maybe May/June/July last year) and increase the CSS regime numbers by whatever you think reasonable (15%?); (3) then pick a date in Summer, say mid/late July, as you have. This should give you an improved Fitness level by the target date. If it doesn’t email me offline and I’ll look at the numbers in depth. If you do get an improved fitness this way then you can choose how to balance the Form/Fitness payoff by adding a taper back in. By preference, Crickles will use HR data whenever it’s available and power when that’s available but HR is not. I’m happy to go through this with you in detail offline with your numbers if what you see isn’t making sense. Ian

      1. Hi Ian,
        Thanks for your prompt response. I set a date range of 1/6/17 – 12/7/17 as I seemed to gain a lot of fitness then (50 to 76 on Strava Fitness and freshness and similar improvements during this period in Golden Cheetah – both of these use power data). I also choose these dates because I’d been away during May and done virtually no training. I had the E’tape du Tour to ride at the end of July so this was the period I focused on bringing my fitness up with time to taper before the E’tape on the 16/7/17 (138 miles/12,800ft) and ride London on the 30/7/17 (127 miles/4,700 ft) .With zero taper there is an improvement in fitness – from 126 to 149. However using your planner (and training like that from now) with a 2 week taper of 75% and 30% puts me back where I started – which I’m sure would not be the case.

        I guess this is because HR data is used. My HR range is not that great. It seldom goes over 153 ish even when working really hard and at a low endurance pace it’s about 120 – not a lot of range to discriminate. Also If I had a couple of hard days riding it probably wont go above 140. Even when rested if I do high intensity VO2 intervals and my HR wont go above 140 until about the 5th interval – so HR tracking probably wont reflect the effort I’ve put in.

        Interesting tool – it would be nice to select if Power or HR tis used – I know the site is all about HR but I use Power to guide my training and HR to help calibrate my fatigue (put simplistically). I always record my HR and power.

  2. Interesting. I’d be surprised if there are any significant CSS~TSS discrepancies. I’ll have a look at your data over the weekend, including your power data. Are you okay for me to email you regarding your detailed info using at email address from Strava/Crickles and to update these comments with conclusions once we’ve looked into it? Ian

    1. Ian,
      That would be fine – it would be interesting to have an independent view of the data. I just checked the Strava fitness and freshness for the period in question selecting HR only and it gave me a big boost in fitness – similar to the power data (Fitness went from 52 to 80 Power, 50-76 HR during the period).

  3. For anyone following the comments, you’ll see two new methods on the Planner for generating the estimates of future efforts based on history. The fit-fat calculations themselves are correct and Nick’s data showed CSS (derived from HR) to be a remarkably good proxy for TSS (derived from power) – rather the improvements concern how to infer an athlete’s future training load from his/her past and Nick’s suggestions have been most helpful.

Leave a Reply