Some results stay in the standings. Others get stuck in the throat — of the losers, the viewers, the contest itself. Eurovision in Riga on May 24, 2003 produced exactly that kind of result, and even two decades later it still sparks arguments, though formally everything was by the rules.
Skonto Hall in Riga was not the biggest arena in the contest’s history, but that evening it held the ambitions of twenty-six countries. Latvia was hosting Eurovision for the first time, after Marie N’s victory the year before, and the organizers built the show for television scale: a dynamic stage, aggressive editing, and a bet on viewers across Europe personally deciding the outcome by dialing phone numbers. By 2003, the televoting format had been promoted for several years by the EBU as the contest’s main democratic mechanism — a counterweight to the old national juries, long suspected of political deals and backroom arrangements.
Three acts stood out among the participants, and their rivalry defined the evening’s entire drama. Sertab Erener arrived from Turkey with the dance number “Everyway That I Can,” with Eastern motifs and choreography designed for the TV screen. Belgian ensemble Urban Trad offered something unexpected — the folk ballad “Sanomi,” performed in an invented language, beautiful and strange at once. And Russia sent t.A.T.u. — a duo that at that point was perhaps the most commercially successful pop act among all the entrants: their singles were already storming the European charts. Their song “Ne Ver’, Ne Boysia” sounded harsher than the usual Eurovision standard, and the group’s very presence in the contest felt like a statement — for the first time, Russia had sent artists Europe already knew.
When the voting ended and the hosts read out the final lines, the scoreboard showed: Turkey — 167 points, Belgium — 165, Russia — 164. The gap between first and third place: three points. Between first and second: two. Between second and third: one, and it was that single point that separated Belgium from t.A.T.u. Such a tight finish is rare even for Eurovision, a contest where gaps of five to ten points happen regularly, though usually not between all three leaders at once.
To understand what went wrong in Ireland, you have to see the whole mechanism from the inside. In 2003, Eurovision’s televoting system operated under a strict protocol. Each participating country appointed its own telecommunications operator, which opened phone lines for viewers immediately after the final contestant had performed. The viewer voting window was brutally short — five minutes. During that time, viewers called the numbers assigned to the acts they liked (voting for your own representative was prohibited). Once the lines closed, the operator counted the calls, ranked the countries by number of votes, and translated the result into the familiar Eurovision scale: 1, 2, 3, 4, 5, 6, 7, 8, 10 and 12 points for the top ten.
But the procedure did not end there. The results were passed to a notary or independent observer, who was required to verify the data and confirm its accuracy — also within the allotted time. The time pressure was greatest on those due to announce their points early: Ireland was third in the running order that night, and it had roughly eight to ten minutes after the lines closed for the entire conveyor belt — collation, notarial verification, and getting the spokesperson on air. Only then could the country announce its points live. The entire chain — from closing the lines to putting the spokesperson on air — took minutes, and any glitch at any stage could derail the process.
That was why the EBU rules provided a safeguard: every broadcaster was required to form a backup national jury in advance. Usually, it was eight people, selected according to criteria set by the broadcaster within the framework of the regulations. The jury gave its scores after the dress rehearsal, not the live broadcast — in case televoting could not, for whatever reason, be completed on time. The backup jury was not an anomaly but a routine part of the system, just as mandatory as the phone line itself.
In Ireland, televoting was organized by broadcaster RTÉ together with telecommunications operator Eircom. The lines opened, viewers began calling — and they called in large numbers: the total number of registered calls exceeded 106,000. The problem arose not on the viewers’ side, but in processing the data. The results did not arrive and clear notarial verification within the allotted window. The reasons were technical failures, whose exact nature is described only sparingly in public sources, but the essence is simple: the “calls → count → verification → transmission” conveyor belt jammed.
Under EBU rules, if televoting is not completed on time, it is voided and the backup jury’s scores come into force. That is what happened. The eight members of the Irish jury, who had awarded their points during the dress rehearsal, became Ireland’s voice that night. Their scores went into the overall table. For viewers in front of their TV sets, the switch was invisible — the Irish spokesperson announced the points in the usual way, and few in the hall or at home suspected that behind those figures stood not a mass of one hundred thousand calls, but the opinion of eight people.
What exactly did the jury decide? Belgium received 10 points from Ireland — the second-highest score on the scale. Russia got zero. Not a single point from the eight Irish jurors.
Later, once the scandal was already gathering momentum, the Irish televoting data was released after all. And this was where I came across a detail that breaks the simple conspiracy narrative: Irish viewers, too, gave Russia zero points. Belgium would have received 2 points in the televote — fewer than from the jury. At the same time, of the 106,000 calls, nearly 57,000 were ruled invalid. More than 15,000 of them because Irish viewers voted for their own entrant, Mickey Joe Harte, which the rules prohibited. The remaining invalid calls came after the five-minute voting window had closed.
This is where we enter territory in which facts and emotions intertwine until they cannot be told apart. The Eurovision fan community is one of the most meticulous in the world: people maintain databases for years, recalculate points, model alternative scenarios. When it became known that Ireland’s televote had been replaced by a jury, the first question was obvious: if televoting had worked, would the result have changed?
The arithmetic looks like this. The jury gave Belgium 10 points; the televote would have given it 2. That is an 8-point difference in Belgium’s favor from the jury. Russia got zero in both cases. So, with working televoting, Belgium would have scored 8 points fewer overall: 157 instead of 165. Turkey and Russia would have remained where they were. The final top three: Turkey — 167, Russia — 164, Belgium — 157. Russia would have climbed to second place, but Turkey would still have won.
You might think that settles it. But fan logic works differently. The argument went like this: we do not know how Irish viewers would have voted if all calls had been valid, if the window had not been so short, if there had been no technical problems. Maybe t.A.T.u. would have received 7, 8, 12 points from Ireland? Maybe that would have been enough to overtake Turkey?
Counterfactual history is fascinating, but it is a bottomless pit. The truth is that we know this: even based on the calls that actually came in — and were partly rejected — Russia did not make Ireland’s top 10. Zero points. The hypothesis that, under “proper” televoting, t.A.T.u. would have received enough from Ireland to win rests on no data whatsoever — it rests on a feeling of injustice, and on the fact that a three-point gap looks painfully small.
The EBU took the situation seriously. The union initiated an investigation that lasted almost a month. On June 18, 2003, it published its official conclusion: RTÉ had acted in full accordance with the regulations. Televoting could not be completed on time for technical reasons; the backup jury was activated under the procedure laid down by the rules; the jury’s points were legitimate. The contest results remained in force.
For the EBU, this was the only possible conclusion — not because the union wanted to bury the story, but because the mechanism had worked exactly as it was designed to work. The backup jury existed for situations like this. Using it was not a violation, but the application of the rules. Blaming RTÉ because the jury voted differently from how viewers might have voted is like blaming an insurance company for paying out under an insurance policy. That is what the policy is for.
But the EBU investigation had another function: it recorded the scale of the problem with the Irish calls. More than half of the 106,000 calls turned out to be invalid. Some were votes for Ireland’s own contestant; others came in after the deadline. That says something about the system itself: a five-minute window was too narrow for a country where viewers had actively joined in. Ireland, for all its Eurovision tradition — seven wins by that point, a record — had run into an infrastructure problem, not a conspiracy.
I’ll admit it: this is the part of the story I find most fascinating — and most uncomfortable for everyone involved.
The backup jury is not proof of vote-rigging. The televoting data released later confirms it: Russia would not have received a single point from Irish viewers. Belgium would have received fewer than it did from the jury. Turkey, in all likelihood, would have won anyway. The EBU conducted an investigation and cleared the broadcaster. Formally, there is no scandal here — there is a technical incident, handled according to the rules.
But by 2003, Eurovision had spent several years building its identity around the idea of viewer voting. Promotional clips, on-air hosts, press releases — everything stressed: this is your contest, you decide. Televoting was not merely a technical instrument, but part of the brand. And when it emerged that in one of the countries the viewers’ voice had been replaced by the opinion of eight people — however legitimately, however much by the rules — it struck at the very rhetoric holding the contest together.
The paradox is that the jury as an institution never went away. It still makes up half the final Eurovision score today (since 2009, the contest has returned to a combined system). But in 2003, the jury was something like a reserve parachute — something people preferred not to remember existed until it was needed. And when it opened, viewers saw that their voice was neither the only one nor a guaranteed one.
For Russian fans watching t.A.T.u. finish three points away from victory, the Irish episode became the perfect object on which to dwell over a lost chance. For Belgian fans, it was confirmation that their Urban Trad deserved more. For Turkish fans, it was irritating noise around Sertab Erener’s fair victory. Each side found in the story what it was looking for, and none received a final answer to the question: what would have happened if?
The contest moved on. Turkey celebrated its first victory. RTÉ revised its technical procedures. The EBU later tightened requirements for televoting infrastructure. And the 2003 Riga final remained in Eurovision history as the evening when the emergency mechanism worked flawlessly — and precisely because of that, people still do not believe it.