Question regrading the results

[maxanes]

Hi,
Recently I had a few samples where a single model couldn't effectively fit all of the data, as in the example below. Do you have any suggestions for handling such cases?
Thank you."
Model 1

Model 2

Model 3

there are 9 models more but worse than 1 and 2