[BUG] check if we are calculating cost of retry without counting the suggested no. of retries

[Aditya26189]

is there the logic for detector retry loops that when calculation cost saved, it is not calculation the 3 retries, like it is accepting that since we have taken the 3 retries as ideal then the all other should not be there, but the cost of 3 would be there, that would not be in potential saving