I am trying to identify the characteristics of an FIR filter in an ideal case as a sanity check. I observe that there is no strange behaviour, but there is an unexplainable issue in the phase plot of the original and identified filters. To make it simple, I've provided the transfer function coefficients below:

a1 = 1; b1 = [-0.0228015616127155 0.445214656159413 0.776226769351428 0.445214656159413 -0.0228015616127155]; b2 = [-0.0228015616127150 0.445214656159413 0.776226769351428 0.445214656159413 -0.0228015616127150]; bode(filt(b1,a1),'-b',filt(b2,a1),':g') 

The difference between b1 and b2 is only at the level of machine precision, yet I observe 360$^\circ$ phase shift in the phase plot. I understand that a 360$^\circ$ phase shift has no physical significance, but I would like to understand why this happens with the bode function in MATLAB. Also, it would be great if you could tell me how to correct this when plotting?