Three years ago, the Lib Dems were still in government. Danny Alexander and Nick Clegg comprised half of the quad, the inner circle that fixed the government’s direction. It feels like a lifetime ago now. The Lib Dems were reduced to 8 MPs in 2015 and recovered only to 12 MPs last year (with a slight decline in vote share nationally), despite being the only party to advocate remaining in the EU. Replacing Tim Farron with Vince Cable has not given them any more airtime or sense of relevance.

The Lib Dems are urgently in need of a strategy. That strategy needs to be founded on a basis that can produce an electoral recovery for them. In turn, that means understanding what they need to defend and where they are best placed to gain ground.

Defence first. Here’s the list of Lib Dem seats. It’s a short list, obviously. There are two points of interest. First, the Lib Dems are not immediately challenged by Labour in any seat that they hold. They have now lost every last Lib Dem / Labour marginal. Secondly, they aren’t particularly safe anywhere – a uniform 5% swing against them would see them lose all but four seats and a uniform 10% swing across these seats would see every last Lib Dem seat fall.

Oh, so you think a 10% swing is big? Let me introduce you to my next list, the list of Lib Dem prospects, organised by swing required. As you can see, a uniform 5% swing to the Lib Dems would increase their tally by just 9 seats. A 10% swing gathers a further 10 new seats for them. A 15% swing brings in just 18 more. And by this stage, the table is actively unhelpful in understanding what’s going on: many of the seats feature simply as a function of a distributed vote – the Lib Dems finished fourth and lost their deposit in Edinburgh North & Leith, for example. Anyone fancy the Lib Dems’ chances in Kensington? I have asterisked where the Lib Dems finished third (and added additional asterisks for each additional drop in the voting order). As you can see, even on this shortish list, asterisks proliferate.

For targeting, I suggest the Lib Dems would do better to concentrate on seats where they already have a substantial vote share, which to my mind gives a truer reflection of where the Lib Dems have potential strength and chances of real progress. Here are the 52 seats where the Lib Dems have more than 20% of the vote.

This to me is a particularly interesting table. First, there are only 52 seats on the list. The Lib Dems held more seats than this up to 2015. If you wanted an indication of just how badly the Lib Dems have been smashed, there’s as good a measure as any.

Thirdly, it shows that the vote has polarised in each constituency – the same effect that has harmed the Lib Dems nationally is making life harder for them in their target seats too. Outside Scotland the Lib Dems hold no seat with less than 40% of the vote. There are three seats in England where they beat that mark and failed to take the seat. This is in part a function of the Lib Dems losing the battle of the left. The English seats they once held on lower vote shares (like Norwich South and Bradford East) were usually Lib Dem / Labour marginals where a fair chunk of the Conservative support refused to vote tactically. Now those have gone, the Lib Dems’ chances of coming through the middle have correspondingly declined.

Fourthly, the seats cluster strongly geographically. The Lib Dems have husbanded their strength well in Scotland. The Lib Dems have considerable residual strength in south west England and in quite a few seats south and west of London. Conversely, they are almost non-existent in the Midlands (a longstanding weak area for them) – there’s a hole bounded by Cambridge, Witney, Montgomeryshire, Hazel Grove and Sheffield Hallam where not a single seat can be found where 1 in 5 voters plumped for the yellow team last time round.

The Lib Dems will need to fight the next election exclusively in target seats – anything else will be a waste. I strongly suggest that if the seat isn’t on this list of 52, it should essentially be ignored next time around. Candidly, 52 right now looks far too high a number to be aiming at.

Next, the Lib Dems need to build policies to help themselves in their target seats. That means taking on the Conservatives. There is no point in taking on Labour because there are next to no Labour seats that they can take. Fortunately, there’s a nice big space between a Labour party that has taken the next left turning after radical socialism and a Conservative party that has decided to major on implementing a nationalistic Brexit, where the Lib Dems could hunker down and turn their guns on the Conservatives. So the opportunity is potentially there. They just need to take it.

Alastair Meeks