Andrew,
I don't think your prediction was totally lucky, there is an element of luck regards weather over the rest of the season, as the truncation of a CT Area drop after 26 July indicated. That applies to my early July prediction. I suppose you will still say I called it too late I was tempted to call a lower around the same time as the posts you quote (20 June) once it seemed to me that the CT Area cliff was over, but decided to wait until all of June's data was in.
However if in March you were predicting a 'recovery' in the sense of higher extent/area than last year - that was lucky. Here is why:
At the time PIOMAS volume was about the same as 2012, but the thickness profile was thinner, as I have shown you previously. You had no way of telling what the spring (May-June) volume loss in PIOMAS would have been, and this period's melt plays a strong role in summer volume loss post 2010. You had no way of knowing whether the summer pattern and its attendant 'dipole anomaly' would appear in 2013 - as things have turned out it didn't.
If you think you had a means of determining either of these in advance I would be very interested (not being infantile here - genuinely interested). I also recall us discussing correlation with end of season minimum and how it is very low before June, although I don't recall you coming back to show how I was wrong.
I do appreciate the flack you're drawing and don't think it is warranted. I've drawn similar flack and have decided some time ago not to challenge the bias you are arguing against. I think it is telling that the blog posts I write which contain words like 'crash' draw in far more traffic than the less excitingly entitled, but in my opinion more important posts I have written. All I have to do to 'go viral' is write posts implying an impending disaster - that aren't too complex.
Oh I don't think anything close to accurate could have been predicted in March. I predicted 4.2 (3.3-5.0) SIE, and 2.7 SIA. Both are bumps from the previous year, but not by much. Both would be 2nd worst all time. I will be lucky if we finish in my MOE for SIE. But I do believe a small bounce over 2012 was the best central estimate at the time, and in retrospect that looks a lot better than the near 2012 or new record predictions.
Perhaps most telling would be a comparison of our volume predictions. I predicted 3.7. I don't see a # in your blog post, but you suggest using the post 2010 thinning profiles which would yield 3.88, 3.67, and 3.43. But then you suggest that it might be more melt this year and that you lean lower. I'm trying to parse your thinking here, but it sound like you were thinking very close to 2012 at 3.3 or 3.4 if you were forced to give a #. Possibly lower I'm not sure.
So we were similar, with you being about .3-.4 km3 lower. I think the difference in our SIA, SIE, and SIV predictions can primarily be ascribed to me being more confident that 2010-2012 represents a skewed climatology of our current climate. I believe that those were 3 exceptionally warm, hostile years. The sample size of 3 was not and is not enough to convince me that that was the new normal. While I leaned heavily towards a 2010-2012 climatology (or a 2007-2012 weather-only climatology ignoring ice differences pre-2010), I hedged slightly against the possibility of a cooler year because of the small sample size.
I want to emphasize that this slight hedge is not remotely similar to the belief of WUWT-types that a recovery is imminent, that 2007,2008, 2010-2012 were all flukes, and that every year will be like 2009, 2013 or colder. I am simply trying to make a subjective recognition that 2010-2012 is a small sample size, and the possibility exists that we will still get years like 2009 (or even less hostile).
Next year, barring some massive recovery in PIOMAS, I will be predicting very similar #s for SIA, SIE, SIV as I did this year (2.7,4.2,3.7), which will all be much lower than the result this year. No recovery.
On a side note, you say the 2013 thickness profiles were thinner but I don't see how this is possible. Both 2012 and 2013 had similar volume and area in March, which means thickness must have been the same as well. Personally, I don't think the distribution matters much. If it is centralized it will protect the ice late season, if it is peripheral then it will delay the start of melt season and positive feedbacks. 2013, 2012, and 2011 all had similar volume in March.
Thus, a good starting point for SIA, SIV, and SIE predictions was an average of 2011 and 2012. If you hedge a little against cooler weather, as I did, you could lean a little closer to 2011, as I did.