I am not saying that the below is exact science or validated. It’s not. The fact that it uses math (really just arithmetic) does not mean that it comes to an exactly correct conclusion.
That said, I think it’s useful anyway:
When dropping weight to achieve more reps, if the result of this is that the product of the weight times the reps is now less than it was before, then you are going to too high a rep value for that exercise for you at this time.
The form and tempo must be essentially identical in the comparisons, by the way. The comparisons are best done in successive workouts with very similar previous work done, and most preferably at or near the beginning of the workout. But it can be acceptable to do it as first and second work sets with a good rest inbetween if experience has shown that you have little drop-off in performance between first and second sets under the conditions in question.
For example, suppose (for the sake of round numbers) you can use 100 lb in an exercise and get 5 reps. The product is 500.
Let’s say when you use 90 lb, you can get 6 reps. The product is 540. So 6 reps is not too high as examined by this method.
Let’s say all this remains true – the product keeps going up or doesn’t drop – down to say 75, where you can get 9 reps and the product is 675.
It turns out that to get 10 reps, you have to drop all the way to 65 lb. The product is 650. This being under 675 suggests that maybe 10 reps is getting too high for you for this exercise at this time, but on the other hand the value is pretty close and so certainly you might see if you can do better next week and either get 11 with the 65 lb (in which case even 11 reps is good) or get 70 for the 10 reps, in which case at least the 10 reps is good. Can’t make a conclusive decision with the numbers being as close as 675 vs 650 a single time.
Let’s say however that even 60 lb cannot yield 11 reps. You have to drop to 55 lb to get 11 reps, for a product of 605.
The conclusion at this point is not that you necessarily never go as high as 11 or more reps in this exercise, but that you are getting past your optimal training range – in my opinion – by this point. I would recommend doing the great bulk of training of this exercise at values less than or up to 10 reps, in this instance.
Just as a rule of thumb, not an exact science.
A different exercise for the same person could yield quite different results, as could the same exercise performed in a different manner, or a different person performing the same exercise.
A briefer way of communicating the principle involved is that when you have to drop quite a lot of percentage of weight to gain another rep, well, the reps are getting too high then.
A differing principle in the same subject area is to figure it based on 1RM. As personal opinion, less than 50% 1RM is generally getting outside the most effective training range, and for that matter the above-50-but-under-60 percent range seems kind of marginal to me.
If having a 1RM and where the 1RM is well-trained then I prefer the second principle to the first. But where for example the 1RM is not substantially higher than say the 3RM and quite a lot better than say the 5RM, or of course when it’s not known, then the first principle is the preferable of the two.