# How to calculate percent difference in a balanced system with theoretical value=0?

This is for a physics lab. We're supposed to calculate the "percent difference" (it could mean something else (I have a feeling it means percent error) it's a lab typed out by my school so it could have errors). What we were measuring is the net torque in a balanced system. SO of course, the... show more This is for a physics lab.
We're supposed to calculate the "percent difference" (it could mean something else (I have a feeling it means percent error) it's a lab typed out by my school so it could have errors).
What we were measuring is the net torque in a balanced system. SO of course, the [sum of the torques] theoretically equals 0, but the experimental torques I got don't add up to 0 (there is about a 1 to 2 N*m difference).

So usually, when asked for percent difference, I always use
| Theoretical - Experimental |
-------------------------------------- x 100
Theoretical

But here, of course, the theoretical value is 0...so I can't divide by it.

How should I go about calculating the percent difference/error of my net torque?
Should I find the difference, and divide by the bigger torque? Or the smaller torque? Or should I divide it by something I haven't even thought of?

i.e. Torque1=19.1; Torque2=20.4
The difference between the net torque is theoretically 0, but in my case, is 0.7...