Thanks Ned, appreciate your perspective. Interesting analysis, and I agree with your math that we need to see 0.69 C figures for the rest of the year. Though I would suggest that this isn't an unlikely outcome, and here's why.
My tendency has been to not look at the year-to-date averages when projecting to the rest of the year, since the monthly anomalies (at least taking GISS as an example) have apparently somewhat seasonal differences. The summer months June and July have typically ran weaker anomalies relative to, say, September through November. I'm just picking a random but fairly recent starting year, but let's say from 1981-2013, what are the averages in anomalies we would find? In units of C, here is what I've derived:
January, 0.428
February, 0.443
March, 0.464
April, 0.398
May, 0.398
June, 0.391
July, 0.390
August, 0.392
September, 0.394
October, 0.397
November, 0.398
December, 0.392
More recent years are even more dramatic. From 2001 to 2013 I get:
January, 0.570
February, 0.553
March, 0.638
April, 0.578
May, 0.565
June, 0.544
July, 0.545
August, 0.565
September, 0.609
October, 0.625
November, 0.646
December, 0.542
The 1981-2013 set shows a more accentuated Jan-Mar period, though the trajectory is ultimately similar to the 2001-2013 set, with a concave shaped curve, and the trough occurring during the summer. I'm not positive how that is the case (possibly ENSO which has relatively little effect on summer climate, but given La Niña dominance in 2001-2013, it cannot explain all of it); but in both examples, June-August (and December, as well sometimes) show the lowest average anomalies, whereas spring and autumn show the most dramatic ones. So whereas the global warming trend between these months has likely been similar, the 1951-1980 averages, as they were, suggest lower anomalies (but only for those specific months.) I held suspicions that June and July numbers on GISS would print on the lower side this year, based on that. The 0.62 and 0.52 for June and July would seem to confirm that. By my logic, we should expect that September through November will have a greater chance to print a higher anomaly, and that something consistently in the 0.65-0.75 range is reasonable. My contention is that the year-on-year average (which is to say from August 2013 to July 2014) is 0.650 C, such that assuming an August 2013 through December 2013 repeat would guarantee 2014 in 3rd, just a nick below 2005 and then 2010 (which finished at 0.653 and 0.662, respectively.) I don't know what percentage odds I'd give it that 2014 may be the record hottest, but factors I am considering to make up the gap include possible El Niño and then the background global warming trend. Tying with 2005 is very possible, and the range of uncertainty and the frequency of revisions are such that this may bear out over time. If it races ahead of 2005, with revisions, seeing a back-and-forth between 2010 looks more likely as well.
As I've stated elsewhere on this thread, I think that if a record happens, it won't be by any grand margin. More likely, 2014 ties with 2010. Either way, I'll concede otherwise that it's just a WAG beyond basing my projections on the near past and the uncertain future. I do think seeing this year be a record warm one would help shift the message away from "hiatuses" and "pauses", and back into the reality that there's a freshly minted record year to tick into the geologic history, consistent with the expectations of climate change.