RPE

[Elephant]

How do you make sure your RPE estimates are correct?

One way would be to go to failure occasionally. For example, if you think something is RPE 8, see if you can do two more reps and can't do any more than that. However, it seems no one does this.

Is RPE supposed to be purely feel (in which case how can you ever be wrong) or is there supposed to be some relation to objective reality (ten minus the number of additional reps you could actually do) or something about bar speed or something else?

[perman]

Maybe it's main purpose is to train submaximally with shorter breaks, better overall rep quality, higher volume without getting systemically fatigued as much and that these normal objections about RPE precision which are common here ignore that precision is not necessary to get the bulk of the benefits of RPE training?

[Elephant]

False precision is certainly an issue. However, many RPE charts include both integers and x.5 (e.g., 7, 7.5, 8), which suggests a fair degree of precision.

Another scale would be easy, medium, hard, couldn't do more, which might correspond to 6, 7-8, 8-9, 10.

[DeanT]

How do you make sure your RPE estimates are correct?

One way would be to go to failure occasionally. For example, if you think something is RPE 8, see if you can do two more reps and can't do any more than that. However, it seems no one does this.

I test my RPE gauge on a regular basis. I started using RPE around a year ago. I have discovered that I was terrible at determining RPE for anything less than a 9.5. To get better, I test my RPE gauge once or twice each month. I plan for it when I am setting up my workout and I don't do it on the fly. I have been able to better determine RPE 8 and up especially for singles. Honestly, I still suck at RPE for anything less than 8. I am especially bad at RPE on sets with 4 or more reps.

As with anything else, I think RPE is a useful tool. My biggest frustration with it is that it isn't a good workout planning tool for me. The RPE charts for multiple reps are much more optimistic for e1RM than I am able to produce. This makes me overestimate my work weights when I'm planning the workout and I generally have to adjust weights down (which I am often not disciplined enough to do). It took me a long time to figure this out.

End of the day, I'm better suited to use percentages for planning purposes and RPE to ensure I'm not getting greedy.

[OCG]

End of the day, I'm better suited to use percentages for planning purposes and RPE to ensure I'm not getting greedy.

Yup. I personally much prefer pre planned workouts using percentages to pure RPE style programming. Works much better for me. But, RPE is still a very useful tool to make sure I'm not killing myself and when to stop.

[Dalton Clark]

I have seen many RPE estimate charts that use percentages/reps. I would argue that if this conversion is used, it isn't really using RPE as the training metric. The reason I argue this is because it is difficult to know prior to training on that day what the RPE for a particular lift at a particular intensity and volume would be. RPE is a difficult thing to relate to any particular metric. Bar speed could be an indicator but there are many lifters who lift explosively and get a few reps but fail extremely quickly. You wouldn't see the failure coming based off of their bar speed. There are also others who are good at grinding out reps and can get many more than I would think they should based on bar speed alone. Similarly, I don't believe additional reps follows a linear scale. This goes back to what I said earlier. You can only know how you perceived the exertion of that set AFTER the set was completed not before. Basically RPE is an anecdotal "I know how challenging this typically is so I will ballpark it" approach. Which is also roughly equivalent to using percentages.

[Michael Wolf]

For those who don't use bar speed trackers, watching a video of the set afterwards can be very helpful to help separate the reals from your feels. However whether you have a video to watch afterwards or not, you have to have a bunch of experience lifting yourself and preferably watching others lift as well. If you train long enough on a purely additive or percentage based program, you will miss reps. You start to accumulate experience of what it's like to miss reps, what the rep before that feels like, what the rep before that feels like, and so on. You also see what that looks like in others. IF you pay attention. Given enough time, and if you have the ability not to bullshit yourself to get out of harder work (90% of people) or to do more even when you shouldn't (the other 10%), then you can gauge RPE with a high enough degree of accuracy to make it useful for whatever it is you're going to use it for.

While you'll find a range of opinions among SSCs and other experienced lifters here on what to do with RPE and how long before it can be used as a relevant factor, it takes time and experience under the bar to be able to gauge it with enough accuracy and precision to use. This process can be sped up if you have a coach watching your sets and telling you what the RPE is or otherwise guiding you about it, but it still takes time to calibrate this.

[AndrewLewis]

How do you make sure your RPE estimates are correct?

RPE is, by definition, not something that correlates perfectly to an objective measurement. You're not testing hook's law with a spring and force meter. The problem with RPE is that it varies highly among the different phenotypical expressions of athletes.

For example, one of my clients (a 37 year-old male) expressed that the last rep was an RPE of 10/10 and his bar speed was actually really good. Conversely, another client (35 year-old female) did a set of 5 at her heaviest ever weight, had two reps that were probably 5 seconds long each, and thought it was like an RPE of 7/10.

This difference is most notable with women versus men; women can be pushed more even if you see a super slow/hard rep. However, it does extend to differences in phenotypical expression even within one of the genders. A neuromuscularly inefficient male will lift and precieve more similarly to a woman than a highly neuromuscularly efficient male.

This is primarily why it's the job of the coach to be able to say "you're done, rack it" or "do another, you can do it" even if the athlete believes the contrary.

Unless you're well into your intermediate progression or advanced, the only time to use RPE is to say "There is zero chance I will add weight to the bar next workout and do 3x5" in which case you need to change your programming in some fashion. In this case, it is binary. You either believe you will make the next workout or you won't. It's not a continuous scale.

[John Hanley]

I have seen many RPE estimate charts that use percentages/reps. I would argue that if this conversion is used, it isn't really using RPE as the training metric. The reason I argue this is because it is difficult to know prior to training on that day what the RPE for a particular lift at a particular intensity and volume would be. RPE is a difficult thing to relate to any particular metric. Bar speed could be an indicator but there are many lifters who lift explosively and get a few reps but fail extremely quickly. You wouldn't see the failure coming based off of their bar speed. There are also others who are good at grinding out reps and can get many more than I would think they should based on bar speed alone. Similarly, I don't believe additional reps follows a linear scale. This goes back to what I said earlier. You can only know how you perceived the exertion of that set AFTER the set was completed not before. Basically RPE is an anecdotal "I know how challenging this typically is so I will ballpark it" approach. Which is also roughly equivalent to using percentages.

When you program set and reps for an advanced lifter at a given percentage, how do you dose the stress? Say, we're doing 75% established-1rm on the squat...how many reps per set? How many sets? What's your logic for these decisions?

[John Hanley]

Similarly, I don't believe additional reps follows a linear scale.

Also: bar speed decay is pretty damned linear. Different sloped for different folks, but linear.