How to configure the optimization algorithms for determined strict initial state

[ertugruldmr]

Situation:
I did a test about 'employee scheduling' task (modified according to our problem) on optapy. When we gave a best plan case (0 Hard and 0 soft) to optapy as initial state (shift_list data), it gave us a very bad output at the end of the time we determined.
Questions :

• Why didn't it give the same output at least ? Why might the current assignment be corrupted?
• additionly how can we manipulate or change the optimization algortihms ?