This newspaper yesterday revealed the exciting news that nuclear fusion, long the holy grail for those hoping to create a cheap and limitless supply of energy, might yet become a reality within the not too distant future. If so, that will be one less mystery for mankind to apply its collective mind to.

Irritatingly, seemingly unanswerable questions will not disappear altogether. One in particular is destined to persist. It is: “What do the opinion polls in Britain actually mean?” While we have more polls than ever before, the consequence is yet more confusion. Only last Friday, The Daily Telegraph was reporting an eight-point lead for Labour in a YouGov poll and headlined the findings with the words “Cameron staring defeat in the face” (a turn of events and use of language only marginally more startling than L’Osservatore Romano, the Vatican newspaper, declaring “Pope not really very good” on its front page).

The implication was that an early election was not merely a plausible prospect but virtually inevitable. Yet this morning, a mere four days later, The Times, via Populus, records that the Labour lead has all but evaporated and so, as a consequence, the chances of an election before May next year have disappeared as well. Then again, it was only a week ago that The Guardian, courtesy of ICM, had a Labour margin of five points, which if it were to be repeated next month, may tempt the Prime Minister to dissolve Parliament.

So what does this mean? How can the polls be suggesting a hung parliament, a Labour majority of about 90 seats or a Labour advantage of about 120 seats simultaneously? Could just one ill-timed trip to Rwanda have cost David Cameron several points in the polls only for the shooting of a youngster in Liverpool to have moved the numbers back again?

Partly this is a question of methodological distinctions between the various polling companies (a subject that makes nuclear fusion look like a model of simplicity) and their precise timing. A large caveat is that the month of August, when so many people are away at various times, has a potential for volatility that is so large that in most years few polls are undertaken during it – a rule that has been broken this time principally because of media speculation about the chance of an election this autumn.

The best way to look at how British politics has changed since Tony Blair left (remember him? Go on, you can if you try hard) is to look at the patterns in the three Populus polls before the change at 10 Downing Street and the three Populus surveys that have been published since Mr Brown took over. These show the terrain of British politics has altered much more substantially than a glimpse at the overall voting figures indicates.

The first factor is that there is both a “Brown bounce” and a “Cameron crunch” at work. The Populus survey today has the Labour vote easing back but the approval ratings of the Prime Minister increasing to levels well above that which Mr Blair achieved in virtually the whole of his last four years in office. Meanwhile, Mr Cameron’s personal score is, at best, flatlining while his party has perked up from its fall during the floods of mid and late July.

The “Brown brand” is far stronger than his party’s broader image, while, for the first time since he became Conservative leader in December 2005, the Cameron brand is weaker than the overall impression held of his party.

This leads to an initial paradox. One of Mr Brown’s main objectives on assuming power was to distance himself from the highly personal style and the alleged presidentialism of his predecessor. Yet the optimal electoral strategy for Labour now is to focus on the Prime Minister himself and let him dominate the airwaves and the print press.

Whenever Mr Brown is prominent, voters seem inclined to accept the assertion that there is a new government in Whitehall and hence their natural “time for a change” instinct is largely neutralised. By far the best strategy for Labour is to frame the election choice as “Brown v Cameron” and not as “Government v Opposition”.

There has also been much heated debate in recent days about the extent to which Mr Cameron has reverted to the core vote approach of elections past, abandoning the new politics of sunshine that he preached at his party conference a year ago for the old politics of moonshine (tax, crime, Europe and immigration). This is indeed a legitimate discussion. But what has been missed, though, is the extent to which Labour under Mr Brown has been running its own core vote strategy by stealth. For while the shift from the Blair to Brown era has seen the average Labour vote share increase by 6.5 per cent, that improvement has been strikingly more pronounced among working-class voters (the DEs) than the affluent ABs. To put it crudely, antiBlair old-Labour backers who had defected or abstained have returned to their historic allegiances.

The last factor is the dog that did not bark. In this case, the mutt concerned is the Liberal Democrats. In essence, what the polls of the early Brown era show is not a huge change in the standings of either Labour or the Conservatives since the 2005 general election: both parties are slightly stronger today than they were two years ago. The seismic difference is in the situation of the Liberal Democrats, whose support has slumped dramatically by between five and seven points.

On that form they could lose up to half of their MPs in an early election. If this early contest does not occur after all, the politician entitled to be most relieved – the man who desperately needs more time to improve his and his party’s flagging fortunes – is not Mr Cameron (nor Mr Brown) but Sir Menzies Campbell.

Times Recommends

• Bank charges hit us for being human
• Sorry, Father, thou still shalt not steal
• The shed where Santa’s failures go to die