Heaven only knows what the Guardian are doing releasing a poll on Boxing Day, but that appears to be what's just happened...
Scottish voting intentions for the 2015 UK general election (ICM) :
Liberal Democrats 6%
There doesn't seem to be any word on fieldwork dates yet, which of course are all-important. I'll update the Poll of Polls at some point, although let's not forget it's Boxing Day...
UPDATE : The fieldwork took place between the 16th and 18th of December, which means the poll is more than seven days out of date, and consequently I can't add it to the Poll of Polls. On a positive note, it was entirely conducted after Jim Murphy became Scottish Labour "leader", so this adds extra weight to the evidence from Survation that there was no immediate "Murphy bounce". Unlike Survation, though, there have been no other ICM polls since September to provide us with pre-Murphy baseline numbers.
I haven't been able to access the datasets properly yet (possibly because I'm on my mobile), but Anthony Wells has noted that without ICM's "spiral of silence" adjustment the SNP would have had a lead of 19%, almost identical to the lead they enjoyed in the recent YouGov poll. The adjustment basically takes a portion of Don't Knows and allocates them to the party they voted for in 2010. This is based on past evidence that undecideds are disproportionately likely to go back to the devil they know, but whether that general rule will hold true in these highly unusual circumstances is anyone's guess. Some of the undecideds who ICM have allocated to Labour will be people who since 2010 have voted for the SNP in the Holyrood landslide, and voted Yes in the independence referendum. Can Labour really be said to be the default choice for such people after all that?