Today’s Ashcroft poll of the Liberal Democrat battleground provides a mixed bag of emotions for the embattled coalition party just weeks out from May 7. There will be reasons to celebrate in LDHQ – Ashcroft implies Lib Dem holds for 4 of their 7 polled seats and shows support to have pulled ahead since September 2014.

But there is cause for concern from today’s numbers. In addition to projecting the loss of North Devon and St Austell & Newquay to the Conservatives, Ashcroft projects a two-point lead – and win – for Labour in Nick Clegg’s own Sheffield Hallam, again.

Clegg has responded robustly, predicting victory and highlighting the consistent Liberal Democrat-voting behaviour of Sheffield residents since 2010. While Cicero Elections can’t confirm this to be an explicit reference to the Real Votes Tracker, there is logic to the Deputy Prime Minister’s defence: Ashcroft’s poll, and local election results, do not reflect national polling.

The parties that led in the polls in late 2014 have consolidated their positions in all eight constituencies included in Ashcroft’s poll. Just as Conservative gains have grown to a 7 point lead in North Devon and a 6 point lead in St Austell & Newquay, the Lib Dem lead in St Ives has grown from 1 to 3 points while currently-held North Cornwall and Torbay have both moved from ties to 2 and 1 point leads, respectively.

Sheffield Hallam and Cambridge – Labour’s targets within this Ashcroft lot – appear to be bucking this entrenchment trend though. Both constituencies posted Labour gains in previous Ashcroft polls, but are now coming back yellow.

The speed of this shift is drastically different between the two constituencies: Julian Huppert, darling of the technologically-savvy liberal academia, has secured a 10 point shift to give himself a 9 point lead. Sheffield has shifted a 3 point lead for Labour to a 2 point lead. It is a small movement, but worth considering.

Ultimately, what this poll points to as a whole is the sheer lack of credibility of national polling ahead of May 7 – especially when attributed to Liberal Democrat chances. Internal Lib Dem polls have long indicated a bump for the party when its individual candidate is named, rather than simply providing respondents with party options. Ashcroft critiques this approach, utilising a different yet similar technique by asking respondents their voting intentions for their specific constituency, rather than a generic national voting intention.

However, the outcome is the same – albeit with less of an effect – Lib Dem support jumps when the question of voting intention is localised. Rather than Clegg’s head on Labour’s Sheffield spike, it is Ashcroft’s subtle underlining of the durability of local Lib Dem MP reputations that will dictate the mood of LDHQ today.

This is not necessarily misplaced optimism. Ashcroft’s numbers show the party exactly where efforts must be made in Sheffield. While Liberal Democrat numbers rise 8 points and Tory numbers drop 6 when voting intention is queried locally, the Labour number stays firmly even at 36.

Clegg must somehow appeal on a personal and constituent level to these Labour voters to have any chance of survival. Canvassing is evidently needed urgently and without delay. There’s a 19:03 train from Kings Cross to Sheffield. Clegg would do well to be on it – if he’s not already there.