Britain’s election is off to a flying start. On the first night of the official campaign, the opening televised “debate” saw prime minister David Cameron and opposition leader Ed Miliband facing a monstering from veteran political interviewer Jeremy Paxman, and rather less aggressive questioning from an invited studio audience.

The broadcast was watched by around 2.6m viewers – lower than the 9m who watched Britain’s first televised leaders’ debate in 2010. Inevitably, the focus during and after the broadcast, was on “who won?” But establishing that is surprisingly hard.

For a start, it depends who you ask. Immediately after the debate, BBC Newsnight journalist Allegra Stratton, speaking from the “spin room”, called the contest for Miliband, based largely on her reading of the candidates’ performances and the body language of their debate teams as they returned from the studio.

While she was on air, results were released from a Guardian/ICM poll of 1,123 viewers, which claimed they saw Cameron as the winner, by a margin of 54% to 46%. Although good news for the PM, it was hardy evidence of a knockout – and could even by spun as surprisingly good news for Miliband: given how low pre-debate opinion of him was, getting so close on the night could be presented as a considerable achievement.

But that wasn’t the end of the matter. Another poll conducted in the immediate aftermath of the debate by YouGov for the Times produced an effective dead heat (51% calling a Cameron win to 49% for Miliband).

Meanwhile, analyses of trending comments on new social media during the debate suggested that, among those commenting on Twitter and Facebook, the advantage lay with Miliband. One study, conducted by Ipsos MORI, looked at the relative balance of positive and negative tweets about each leader during the debate. Of 29,598 tweets mentioning Cameron, 35% were positive but 65% were negative. For Miliband the picture was reversed: 10,755 tweets during the debate mentioned him, and his positive/negative ratio was 62:38.

Confused? Don’t be. Two things are worth bearing in mind: advance preparation makes a clear-cut result less likely; and to a considerable extent, viewers call the debate based on their pre-existing prejudices. While the debates are not inconsequential, they are not necessarily the game-changers which news organisations might suggest.

Great expectations

In some ways, we should not be too surprised that perceptions of the debate outcome placed the two leaders so close. Both Cameron and Miliband had been preparing extensively for the ordeal ahead. Though car-crash performances can never be ruled out, the parties and their leaders are leaving as little as possible to chance. And in the opening debate, both protagonists kept their respective cars on the road without major mishap.

But party leaders enter these events with a weight of expectation already on their shoulders. Many viewers have already made up their minds about them and are more likely to pick up on features of their debate performances which confirm those pre-existing views rather than on features which challenge them.

In 2010, the first-ever round of televised leaders’ debates during a British election saw just this phenomenon. The British Election Study (the leading academic survey of British voters) asked respondents to its internet campaign panel whether they had watched that year’s debates and (if they had) who they thought had “won”.

Two-thirds claimed to have seen at least one debate. Of this group, only 17% thought Labour prime minister Gordon Brown had come across best overall in the debates, and 61% thought he had performed worst; 31% thought David Cameron (then the leader of the opposition) had won, while 34% thought him the loser. But the stand-out winner was Nick Clegg of the Liberal Democrats, rated the winner by 52% of those BES respondents who watched a debate (only 4% rated him the overall loser).

So far, so conventional. The clear narrative of the 2010 campaign was that Clegg won the debates. But a more interesting question is who thought well or badly of each leader’s performance. The BES is useful here because it was a panel survey, interviewing the same people before, during, and after the campaign. Before the campaign (and therefore before any of the leader broadcasts had taken place), respondents were asked how they intended to vote (and, for those with a view, whether they were definite or just leaning more to one party than to any other).

As the diagram below shows, each leader’s debate performance was rated much more favourably by people who were already leaning towards his party than by people who were not. For instance, while Gordon Brown’s ratings were uniformly dire among those who were already thinking of voting Conservative or Liberal Democrat, he was much more likely to be rated the debate winner by those who were already thinking of voting Labour (48% of those who said before the campaign that they were definite Labour voters and who had seen a debate thought Brown was the winner). The picture is similar for David Cameron: those who were already thinking of voting for the Conservatives were very likely to rate him the winner, while very few indeed of those intending voting for the other parties did so.

Author provided

There’s an important lesson here. How viewers see the debates is heavily coloured by their prior partisanship. We tend to see what we want to see.

But another key lesson emerges from the 2010 experience. Nick Clegg really did do well in the debates. Unlike his two rivals, he picked up substantial plaudits not only from those intending voting for his own party, the Liberal Democrats, but also from those intending voting for the other parties. Among those who initially thought they would definitely vote Labour, for instance, he ran neck-and-neck with Brown in perceptions of who won the debate.

Why? In large part, this was a function of visibility. Going into the 2010 campaign, Brown and Cameron had already been subject to extensive media exposure and public comment. But Clegg was much less of a known quantity for most voters. While they had largely already made up their minds about the first two, the debates were the moment when they first encountered Clegg – and they liked what they saw.

Novelty factor

There is a subtler lesson from 2010 too. Three debates took place on consecutive weeks during that campaign. As the following graph of opinion poll trends during the 2010 campaign shows, so-called “Cleggmania” peaked immediately after the first debate, when he did most to introduce himself to most voters: his poll ratings, and his party’s, surged, while both the Conservatives and Labour fell.

But in later debates, the novelty of a new and congenial figure on the stage wore off, and the other leaders did more to counter his charms. Support for the Liberal Democrats fell back somewhat (though remaining above pre-debate levels), while Conservative support rallied and Labour’s position stabilised.

By the election itself, the Liberal Democrats were only slightly ahead of where they might have been expected to be, given trends before the debates – while Labour was only slightly behind (the Conservatives came out of the process in more or less the same position as they went in). While the debates galvanised interest in the campaign, therefore, whether they had any real effect on the outcome is moot.

Author provided

Roll the picture forward to 2015. This year, too, voters’ already strong views of the major parties mean that (barring major gaffes) their leaders will have to work hard in the debates to change their parties’ fortunes substantially. The big change since 2010 is that this now includes Clegg.

All three leaders are now debating, not just against each other, but against the public’s heavily entrenched views of them. While that does not rule out dramatic reversals of fortune, it does make them much less likely.

But, in the coming debates, attention will turn not just to the major party leaders, but also to the leaders of the Scottish and Welsh nationalist parties (major figures in their own regions, of course) and of UKIP and the Greens. If the experience of the 2010 debates is anything to go by, that’s where to look for debate-induced drama. If the televised debates are to make or break any careers in 2015, watch the dark horses.

Charles Pattie does not work for, consult to, own shares in or receive funding from any company or organisation that would benefit from this article, and has no relevant affiliations.