Even before votes were counted in this year’s Senedd (Welsh parliament) election, speculation among commentators was rife that one campaign narrative had firmly taken hold – that the contest had become a two-horse race between Plaid Cymru and Reform UK. Both parties promoted that framing during the campaign, urging voters to see the election as a straight choice between them. In the aftermath of the result – and Labour’s losses – attention quickly turned to whether the media had amplified that message, including criticism by a Labour Senedd member who refused to talk to ITV News because of its coverage. Clearly the media were not solely to blame for Labour’s decline. However, our analysis of election coverage found that more than one in four TV news items featured an opinion poll, often framing the contest as a battle between two parties. On UK-wide flagship bulletins, that figure rose to more than half. In the final week of the campaign, almost half of all TV news items referenced a poll. A binary choice?From the outset, Plaid Cymru and Reform used campaign slogans that presented the election as a direct battle between the two parties. The implication was that voters should back one of the frontrunners rather than waste their vote on other parties.That framing carried particular significance because this election was held under a new proportional voting system. Unlike Westminster’s first-past-the-post model, proportional systems are designed to produce representation for multiple parties. Seats are allocated according to vote share. Because of this system, the election couldn’t have been further from a two-horse race.Stronger performances by Labour, the Conservatives, Greens or the Liberal Democrats could have translated into representation in the Senedd.But public understanding of the new system remained limited throughout the campaign. Surveys conducted before and during the election suggested widespread confusion about how votes would translate into seats, alongside misinformation about tactical voting.Whether the two-horse race narrative actually changed voting behaviour remains difficult to determine. Post-election research will need to assess whether voters acted tactically, misunderstood the electoral system, or were influenced by campaign messaging and media coverage.Research has long suggested that heavy reporting of opinion polls can contribute to a “bandwagon effect”. This is where voters gravitate towards parties perceived to be gaining momentum.Horse race dominated campaign coverageAs part of our election media analysis project, we tracked all election news coverage across both UK-wide and Welsh broadcasters between April 8 and May 6. This included flagship TV news bulletins on the BBC, ITV and Channel 4. We also looked at online news articles from BBC Wales, ITV Wales, S4C and Sky News, and social media content produced by these outlets. Our previous analysis found that broadcasters produced many explainers during the campaign. These included videos outlining how the D’Hondt proportional voting system works.But in day-to-day reporting, explanations of how votes translated into seats were far less common than stories about which parties were rising or falling in the polls. Instead, coverage increasingly focused on the electoral horse race, particularly in the final stretch of the campaign. TV news items to feature opinion polls in the Senedd election campaign. Cardiff University, AHRC, CC BY Seat projections also became more prominent as polling day approached. Although not always included alongside polls, they shaped reporting in the campaign’s final days across television, online coverage and social media. This included ten online items in the final week, four on social media and two on TV news.Because Plaid Cymru and Reform were leading many of the polls, coverage often centred on those two parties. On May 5, for example, ITV News’ Wales at Six programme opened by reporting that Plaid Cymru was surging ahead of Reform in the broadcaster’s latest poll.As polls increasingly drove coverage, the election itself came to be narrated as a contest between two parties competing for victory.Did coverage squeeze out other parties?The prominence of polling and seat projections inevitably reduced attention on other parties and on the broader dynamics of proportional representation.It’s questionable whether broadcasters should have amplified campaign messaging that framed the election as a binary contest. After all, without understanding the proportional voting system, people may not have appreciated that it is designed to represent a range of parties rather than produce a winner‑takes‑all outcome.At the same time, it is important to recognise the difficulty of isolating media influence. Broadcasters were largely reporting representative opinion polls that, in many cases, accurately reflected the final outcome. But our analysis suggests that, in the final week especially, the campaign was increasingly understood through the language of momentum, winners and losers.That approach undoubtedly added drama and urgency to coverage. But it also risked diverting attention away from policy debates and from the realities of a proportional political system designed to produce a more representative mix of parties in the Senedd.Stephen Cushion has received funding from the BBC Trust, Ofcom, AHRC, BA, ESRC and Welsh Government.Keighley Perkins receives funding from AHRC for research into broadcasters' impartiality.Maxwell Modell receives funding from the AHRC for research into broadcasters' impartiality.