Goal Seek doesn't run on iterative code

It might help to have some concrete functions to test on. I tried testing using my Newton-Raphson spreadsheets (https://www.excelforum.com/tips-and-...ind-roots.html ) with mixed results. Most of my testing was to use Solver and/or Goal Seek to get a specific root of those polynomials by changing the parameters of the function.

Goal seek never worked. I observed, as you did, that Goal seek would change the "by changing" cell, but the target cell never changed. Sometimes, when Goal Seek finished, Excel would give me a "circular reference" warning and I would have to force a couple of calculation events (or disable then reenable iterative calculations) to get Excel to correctly calculate the iterative calculation again. I am inclined to believe that Goal Seek is just not sophisticated enough to work with circular references/iterative calculations.

Solver, on the other hand, gave mixed results. Sometimes, Solver gave a "could not find a solution" message. Sometimes, it would correctly find the solution. I also tested on another equation (more of a continuously increasing exponential function, so it would not have the problems with multiple roots), and Solver worked quite well with this equation (Goal Seek, again, would do nothing). Without knowing more about the sequence of calculations between P1 and P2, I can only guess that Solver's algorithms just don't work well or are unstable with your calculation sequence.

I can offer no specific suggestions. You might try different starting values -- perhaps if you give Solver a better initial guess, it will be able to find the solution. Or you might look at your calculation sequence and see if there is a more stable sequence of calculations. Or, if you can formalize the steps you use to manually change P1 to get the desired value for P2, perhaps you can program that sequence of steps into the spreadsheet and completely bypass Solver. If you want us to try anything, you will probably need to provide us enough information about your P2=f(P1) calculation so that we have something directly pertinent to your problem to test.