The Liberal Democrats have had an awful election so far. The pain is not because their vote share is going down since the last election – the polls all agree that their vote will be significantly up since 2017 – but because they are falling so far short of what seemed reasonable expectations only a couple of months ago. While the idea of winning 80-100 seats was always quite a reach, it was not long ago that 40 looked a reasonable target. After all, they achieved a good second place to the Brexit Party in the European elections and had enjoyed a very successful set of local elections earlier this year.

There are several feasible explanations for the slow puncture that has seen the Lib Dem share of the vote deflate from a little over 20 per cent in polls taken in mid-September to the region of 14 per cent where they seem to be at the end of November. The onset of the formal election campaign has not been a boost for the party as it did in many previous elections; with a relatively little-known leader in Jo Swinson there seemed to be the potential for a new version of the ‘Cleggmania’ that upended the 2010 campaign. But Swinson fever did not attain epidemic or even endemic proportions.

There have been some political missteps as well; the policy of straightforward revocation of Article 50 was bold and simple but it invited questions and doubts about what the party would do if it failed to gain an overall majority for revocation, and whether the policy was really feasible. It created political space for Labour to shuffle a bit further into ‘exit from Brexit’ territory with its renegotiate plus referendum policy. By attacking Corbyn so harshly, an entirely reasonable strategic step to appeal to its target vote of Conservative Remainers, they also made left of centre Remainers, who now had an adequate policy from Labour, suspicious of the party’s intentions and wary of a repeat of the coalition of 2010.

Perhaps the withdrawal of the Brexit Party ended up hobbling the Lib Dems more than anything else

I wonder if the withdrawal of the Brexit Party as an effective national competitor actually ended up hobbling the Liberal Democrats more than anything else. With the BXP in contention, the national debate would be more on the Brexit/ Remain axis – if there was polarisation going on here, the Lib Dems could position themselves as the clearest antagonists to the BXP and Swinson would have been in her element taking on Nigel Farage directly. A genuine four-way contest would also lower the threshold for winning seats, allowing the Lib Dems to escape the ‘wasted vote’ trap of First Past the Post. With the consolidation of the Leave side around the Tories, the dynamic changed. For wavering centre-left voters, the prospect of a Tory landslide encourages them to shore up Labour; for wavering centre-right voters, the prospect of Corbyn as Prime Minister encourages them to hold their noses and vote Tory. The two big parties, like two leaning walls of a decaying building, support each other through their mutual decrepitude.

So what does await the Lib Dems when the 2019 votes are counted? We can look at YouGov’s MRP model for a glimpse of what is possible. I should caution at this point that Lib Dem support might be significantly harder to model than the relative votes for Labour and the Conservatives, as it depends so much on candidates and constituency context. The generally accurate YouGov MRP model in 2017 was off-beam for the Lib Dems – missing 5 of the 12 constituencies they won and predicting victory in one seat (Ceredigion) that they lost. It might be even more off this time given the higher Lib Dem national share.

Health warning duly noted, what does the model say? The total of 13 seats, only one more than in 2017 despite the higher vote and the wave of defecting MPs, would be terribly disappointing for the Lib Dems. Even if the model is pessimistic, it is still hard to see how they get much more than 20 seats: there are 15 more seats where the Lib Dems are within 10 points of the leading party, and another 21 where they are between 10 and 20 points behind.

Looking at where these 49 constituencies where the Lib Dems should win, or become competitive, in this election, a pattern emerges. Of the 13 where the Lib Dems lead, three did not vote for the party in their last good election in 2010 (Oxford West & Abingdon, Richmond Park, St Albans). Imagining for a moment that the Lib Dems win all 49 prospects, there are some interesting changes with respect to 2010. The Lib Dems are thinned out in Scotland and the South West, and lose nearly all their scattered urban seats, both university quarters (Cardiff Central, Cambridge etc) and working-class footholds (Burnley, Birmingham Yardley) alike. But they would make striking gains in the Home Counties and the wealthier areas of London, many of which were not even remotely on the radar before this year – Chelsea & Fulham and Esher & Walton for example. Their representation would gain greater demographic coherence, concentrating in the South East. This would be a major reversal for the Liberal strand of politics, which has drawn its strength from the periphery – the South West, the ‘Celtic fringe’ and so on – ever since the dawn of party electoral politics in the Victorian era.

Map: Lib Dem gold – Lib Dem in 2010 and in the 13 2019 model; Yellow – Lib Dem in 2010 and among the other 36 plausible seats; Pink – seats that were not Lib Dem in 2010 but are now projected to be in the top 49 Lib Dem prospects in 2019; Necrotic grey – seats that were Lib Dem held in 2010 but are not remotely feasible prospects in 2019.