It’s setting the bar pretty low to say that Saturday’s byelection for the Victorian state seat of Broadmeadows was more interesting than expected. Everyone thought it would be a walkover for the ALP’s Frank McGuire, and so it was, but nonetheless there were a couple of features of interest.

In any byelection such as this, there are two things we might hope to get an indication of. First, how are the new government and opposition travelling — what does it say about the overall Liberal versus ALP position?

Secondly, how do Liberal voters behave in the absence of any official direction — with no Liberal candidate or how-to-vote card, where do those votes go?

Unfortunately, it’s the nature of the case that the result can’t tell us both of those things at once. If we think we know the answer to one, we can use it as an assumption to get information about the other. But if the assumption is wrong, we’re back at square one.

The first interesting thing about Broadmeadows is that we have what looks like a reasonable proxy for a Liberal vote: although McGuire won comfortably with 53.8% of the vote, independent Celal Sahin came a creditable second with 19.7%, compared to the Liberal vote of 25.3% last November.

Sources on the ground on Saturday said that Sahin’s campaign appeared to have Liberal backing, and his how-to-vote card prominently told people to put Labor last. Moreover, the pattern of his vote by polling place is not a bad fit for the Liberal primary vote in November, although Sahin’s vote in the Glenroy booths was much lower. Not enough to label him as a stand-in Liberal, but enough perhaps to adopt that as a working hypothesis. (Don’t bother checking his website for more information.)

If we count Sahin as a Liberal, then the Labor/Liberal balance has barely changed, although with a much larger field of candidates the primary vote is down for both majors. Labor won 73.2% of the raw two-party vote, as compared to 71.1% in November. In November that translated to 71.0% two-party-preferred; the VEC hasn’t done a McGuire/Sahin preferential count (and it’s not clear whether it will), but Antony Green’s estimate puts Labor on 68.9%.

But all those numbers should be treated with suspicion, because they’re percentages of a lower total: more than 22% of the electorate, unsurprisingly, failed to vote, double the number from November. That could be Liberal voters staying home due to the absence of a candidate, but it doesn’t look that way: the decline was less, rather than more, in the stronger Liberal booths.

The Liberal voters who turned up and didn’t vote for Celal Sahin — of whom there must be some, since presumably Sahin had some support from former Labor voters in the Turkish community — seem to have gone all over the place. The DLP vote more than doubled to 5.6%, the S-x Party debuted at 5%, and an independent who had the donkey vote beat both of them with 6.2% — an obvious option for otherwise homeless voters.

We can, however, be pretty confident about one thing they didn’t do: they didn’t vote Green. That’s the second interesting result from Saturday. The Greens completely failed to benefit from the absence of a Liberal candidate; their vote dropped off in much the same way as the Labor vote, from 7.5% to 6%.

The problem for the Greens wasn’t just primary votes; they did badly on preferences as well. We know this because the VEC did a Labor/Greens two-party count on the night, even though in reality the Greens finished a distant fourth. If voters for the other candidates had all followed the how-to-vote cards, the Greens would have had 28% two-party-preferred; instead they only got 24.6%.

Broadmeadows is poor Greens territory by any standard, so the party probably won’t worry too much about what happens there. But it looks as if a significant number of Liberal voters have now got into the habit of preferencing against them, even when the how-to-vote card (as Sahin’s did) puts them ahead of Labor.