Confused about TSS value for similar workouts


Hey folks - I was looking through a workout for today, and was surprised by the TSS values for the two and was hoping someone could help me understand. Here are the workout details:

  • Clouds Rest + 1: 60 minutes, 81 TSS. Description: Clouds Rest +1 is 3x12-minute sets of 30-second VO2max repeats at 130% FTP with 20 seconds of active recovery between repeat and 5 minutes of recovery between each set of repeats.
  • San Joaquin: 60 minutes, 90 TSS. Description: San Joaquin is 3x12-minute sets of 50 seconds at 130% FTP with 50 seconds of active recovery between each interval.

Intervals for both are at 130% FTP, and the active recovery in each won’t add much to the overall TSS.

For Clouds Rest, there are 42 30-second intervals at 130% FTP, for a total of 21 minutes at 130% FTP. For San Joaquin there are 24 50-second intervals at 130% FTP, for a total of 20 minutes at 130% FTP.

Given that - how does San Joaquin (less overall time at interval intensity) have 10% higher TSS than Clouds Rest +1? Thanks in advance for any thoughts.


The warm up looks like it is a minute longer. I doubt that would account for 10% but it would add a little.

This discrepancy probably arises due to how normalized power is calculated. IF and then TSS are calculated from NP. NP uses a 30 second running average and a fourth power which means it does a little filtering and interval length will matter. Its not just a ratio of time at intensity. I’m not an expert on this and I’m sure someone else can do a better job explaining or link to a page.

I think the effect here is that maintaining those 130% efforts for slightly longer is considered to be harder even though the total time at that level is shorter.


Really good question, @KatuskaMTB. I’ve actually struggled with this in the past when designing workouts and deciding which form of cruelty to inflict on myself that day. Pretty simply, the TSS calculation incorporates Normalized Power, which is basically weighted power. So the longer intervals–50s vs 30s–up at 130% account for higher intensity and that higher intensity translates to higher TSS even though the work performed is strikingly similar between both workouts. Crazy, right?


It’s all driven by the math used in the NP formula. In the following examples, let’s assume the NP formula uses a 2 second rolling average instead of 30 second, and the workouts are very short :grin:(the concept is the same when you extend to 30 second rolling average and longer workouts, just makes it easier to see in an example).

Workout #1 (second: FTP%):
1: 0%
2: 100%
3: 0%
4: 100%
5: 0%

NP = ((0.5^4 * 4)/4)^1/4 = 0.5 = 50%
Average power = 40%

Workout #2 (second: FTP%):
1: 0%
2: 100%
3: 100%
4: 0%
5: 0%

NP = ((0.5^4 + 1^4 + 0.5^4)/4)^1/4 = 0.73 = 73%
Average power = 40%

So both workouts have the same average power, but workout #2 has higher NP due to the 2 second block at 100% FTP.

Intuitively this makes sense - it’s harder to do a single block at 2x duration vs 2 blocks at 1x duration with a rest interval in between.


That makes a lot of sense @chad , I had thought TSS was more “mechanical” than that but glad to see it incorporates the clearly more taxing 50-second intervals well.


Thanks for breaking it down this way - definitely makes more sense now. If I had extended the analogy further and gone from a bunch of 30 or 50 second intervals to one 20 minute 130% FTP interval (obviously impossible) then I would expect a massive TSS for that!