The full result from Selkirkshire hasn't been published yet, but I thought you might enjoy the headline while we're waiting. Before you get too excited, the Tories were defeated by an independent candidate rather than the SNP, but even so.
UPDATE: Here's the result, and a finer advert you could never see for why it's so important for SNP supporters to use their lower preferences in local council elections. The Tories were leading on first preferences by a seemingly comfortable six percentage points (albeit their share of the vote was sharply down on their result in the ward last May). But they still lost, because enough SNP, Labour, Lib Dem and Green voters gave their lower preferences to the leading independent candidate. The beauty of preferential voting is that even if your favourite candidate or party is well off the pace, you can still help prevent your least favourite party from claiming the seat, and that's exactly what happened here. Well done to the progressive people of Selkirkshire for using the voting system correctly.
Selkirkshire by-election result (first preference votes):
Conservatives 35.7% (-6.4)
Independent - Penman 29.7% (n/a)
SNP 19.8% (-2.4)
Independent - Gunn 6.3% (-3.5)
Labour 3.8% (+0.1)
Liberal Democrats 2.7% (-0.6)
Greens 2.0% (-1.0)
Even after five counts (ie. after four candidates were eliminated) the Tories were still narrowly ahead by 1307 votes to 1231. But the decisive moment came when the sizeable pile of SNP votes were transferred on the sixth count - 291 went to the independent, and only 35 went to the Tories, leaving the independent victorious by a margin of 1522 votes to 1342. If no SNP voters at all had used their lower preferences, Selkirkshire would now be stuck with another Tory councillor.
There's always the fascination of seeing how Labour votes transfer when both the SNP and Tories are still in contention. In this case the winning independent was the biggest beneficiary of Labour transfers, but nevertheless more than twice as many Labour voters transferred to the SNP as transferred to the Tories. Clearly the much-vaunted unionist voting bloc has its limitations, even in territory such as the Borders where you'd think the anti-Nat pitch would have some purchase.
The slight fall in the Liberal Democrat vote from 3.3% to 2.7% may not look terribly significant - but when you realise this was a ward in which the Lib Dems took 19.7% of the vote as recently as 2012, it puts things in a completely different perspective. In Westminster terms, this is part of a constituency that was a Lib Dem heartland until very recently. So much for the "fightback".
Although the SNP vote fell a little, the fact that the Tory vote fell significantly more means there was technically a swing from Tory to SNP of approximately 2%.