SVISS Supernova candidates, third screening

The total number of candidates in the different quadrants are:
B1: 7 (2)
A2: 8 (2)
A3: 4 (2)
B4: 1

New stuff in this version:
Cutouts and lightcurves only shown for uncertain candidates (the number of uncertain candidates are shown in paranthesis above). For each candidate the following information is available:

Identification number and quadrant (e.g. a2-12)
Best fitting template
Best fitting parameters with the above template, M_B(peak) is the fitted absolute magnitude at peak, z_fit is the fitted redshift of the SNe, t_e is the teomporal offset in days between the explosion date and the first search epoch and extmod is the extinction model (1-7 Cardelli with R_V=3.1, 8-14 Cardelli with R_V=2.1 and 15-21 Calzetti and E(B-V) from 0 to 0.6 for each).
Fitting quality (P/L/chisq) for the above template
Host galaxy magnitude, m_I(host)
Host photometric redshift, z_p (1 sigma uncertainty)
Comments (my comments on the data and fit quality)
A panel with the observed light curve plot (R, I, R-I) and the best fit light curve for each template.
A panel with the image stamps, containing the reference image, the seven detection images and the seven subtracted images.

Ratings:

A, primary SN candidate
B, secondary SN candidate, typing questionable/difficult
X, not a supernova, likely AGN
0, not a supernova, likely nothing

Please give me your comments on each of the problematic six. The rest (14) have been designated A or B by all of us, those I will include. The final 11 ones have been designated 0 or X by almost all of us (in some cases there is one differing opinion, but if there's more than one it's in the problematic list).

B1 quadrant

b1- 50
Type: IIn with M_B(peak)=-18.64, z_fit= 0.76, te= 39, extmod= 1
Quality of fit: P=0.710431, L=0.23373785E+10, Chi²= 0.11936187E+03.
Host properties: m_I(host)=22.9329, z_p= 0.75 (0.22)
Comment: This source is tricky. The bestfitting type (IIn) has a fairly good likelihood, but an atrocious chisq (which is evident by looking at the curve). The source is also negative in the last 2(3) epochs in both R and I (although quite faint). It shows nothing in the 2 year baseline subtraction. I think the problematic fitting, together with the negative sources indicates that this is not a SNe, but rather an AGN. Final verdict: X.

b1- 51
Type: IIP with M_B(peak)=-17.65, z_fit= 0.54, te= 31, extmod=14
Quality of fit: P=0.985121, L=0.49792648E+09, Chi²= 0.27840198E+01.
Host properties: m_I(host)=21.2523, z_p= 0.67 (0.14)
Comment: This one actually has a very weak negative feature in the 2 year baseline subtraction, and an even weaker negative feature in the epoch 7 I band image. However, it is fitted fairly well with a SN light curve, the caveat being epoch 5 in which the source got a bit brighter again. This point is the reason why a IIP is preferred (which means that the assigned redshift is quite low, and the peak mag quite a lot brighter than the mean IIP one). This is really at the limit of what can be accepted, but I think its more likely that this is a SN than an AGN. Final verdict: B.

a2- 56
Type: Ibc_bright with M_B(peak)= -19.23, z_fit=0.60, te= -73, extmod= 7
Quality of fit: P= 0.684955, L= 0.74698975E+00, Chi²= 0.87300124E+01.
Host properties: m_I(host)= 20.3099, z_p= 0.47 (0.18)
Comment: I think this is a SN. There are some problems with residuals in the first epochs (in primarily I but also R) which makes the first three data points suspect. However, in epoch four the true source suddenly becomes brighter (or actually appears) simultaneously in both R and I. This indicates that this is a real thing. There is nothing to indicate that this is an AGN. The first 3 epochs might be affecting the fit in for this source, I should probably rerun the fitting with them treated as upper limits. Final verdict: B.

a2-167
Type: IIP with M_B(peak)= -14.42, z_fit=0.08, te= -49, extmod= 21
Quality of fit: P= 0.999999, L= 0.10465594E-04, Chi²= 0.29524527E+01.
Host properties: m_I(host)= 22.0182, z_p= 1.72 (1.46)
Comment: Clear detection in the 2-year baseline subtraction. I think this indicates that the source is an AGN. Final verdict: X.

a3- 23
Type: 91t with M_B(peak)= -19.84, z_fit= 0.64, te= 141, extmod= 1
Quality of fit: P= 0.645455, L= 0.31771995E-02, Chi²= 0.29791376E+01
Host properties: m_I(host)= 21.5613, z_p=0.58 (0.50).
Comment: Very faint source, so extremely hard to say anything definite about (let alone type, although my simulations include SNe as faint as this so the mistyping should in principle be understood even for sources like this). I think this is a SN. The photometry is shaky but it just doesn't look liek a residual to me (and based on the 2 year subtraction its not an AGN). Final verdict: B.

a3- 26
Type: IIP with M_B(peak)= -14.42, z_fit= 0.08, te= -17, extmod= 17
Quality of fit: P= 1.000000, L= 0.51966120E-52, Chi²= 0.17545867E+02
Host properties: m_I(host)= 21.8209, z_p=0.73 (1.20).
Comment: I don't like this one. It would be a very nice IIP candidate but for the first I band epoch. There's a clear detection there (~5 sigma) and that ruins the fitting completely (forcing the SN to have low redshift and to be very faint, in order to make the observed data point as close as possible to the model 0 flux). Performing a fit with this point as an upper limit gives a much better fit for a IIP at higher z (but this does not feel very ethical). All in all I don't think that point can be disregarded, the host also has a very high CLASS_STAR parameter making it a possibility that the source is actually a variable star. This also fits with the colour essentially staying constant. Final verdict: 0.