I've forgotten to say, the width of the curve is more or less 2L.

And what I'm trying to do is: I'm trying to solve some tunneling corrections for un unsymmetrical Eckart barrier for a chemical reaction. So on the left side of the curve (the left tail) is the educt, on the right side the product and the maximum is a transition state. So I've modified the data, like they have do in the publication, where I have the model function (Eckartfunction) from. I don't know if it is saying anything to you, but I hope it will help.

Thanks ones more

You are right, there was some trouble with my data's. And some of the value's are negativ. New data:

data := [[-1.15926589, 0], [-1.08606248, 0.121741500004191e-2], [-1.00976677, 0.265672099999392e-2], [-.93094726, 0.431098900003235e-2], [-.8503015, 0.61591659999749e-2], [-.76834172, 0.816635000001042e-2], [-.68540474, 0.102850050000143e-1], [-.60214647, 0.124569199999769e-1], [-.51778321, 0.146089130000746e-1], [-.43325872, 0.166652540000314e-1], [-.34843929, 0.185401810000485e-1], [-.26338805, 0.201447860000599e-1], [-.17813459, 0.213917739999943e-1], [-0.9271974e-1, 0.222022390000802e-1], [0, 0.225149170000805e-1], [0.9283385e-1, 0.221928170000183e-1], [.17851801, 0.213258170000472e-1], [.26435016, 0.199344170000586e-1], [.35030094, 0.180637170000182e-1], [.43638693, 0.157797169999867e-1], [.52260379, 0.13163817000077e-1], [.60894663, 0.1030291699999e-1], [.69542035, 0.728131699997903e-2], [.7820355, 0.417761699998209e-2], [.8687898, 0.106451700003163e-2], [.95568486, -0.199408299999959e-2], [1.0427333, -0.494428299998617e-2], [1.12995557, -0.774228299997048e-2], [1.21739502, -0.103532829999722e-1], [1.30510885, -0.127513830000225e-1], [1.39318442, -0.14918683000019e-1], [1.48173484, -0.16845582999963e-1], [1.57100811, -0.185307829999601e-1], [1.66083359, -0.199799829999847e-1], [1.75141836, -0.212065829999801e-1], [1.84268948, -0.222312829999964e-1]]

But the fit is worse...

I thank you all for trying to help me.

But the Eckartfunction [y=((A*exp((2*Pi*x)/L))/(1+exp((2*Pi*x)/L)))+((B*exp((2*Pi*x)/L))/(1+exp((2*Pi*x)/L))^2)], I want to fit to my data's can't be modified, it's fixed.