In Washington DC, I measure out my life in polls and heart palpitations. The polls are relentless, nail-biting, maddeningly contradictory. There are national polls, swing state polls, polls from tiny counties that predict a whole election, partisan polls designed to demoralise the other side.

There are polls on whether a candidate inspires confidence, compassion, leadership. I’ve noticed how, after a bad poll, I start looking for another that tells me numbers I like. I’ve also noticed how, after a good one, I will look for a bad poll to bring me down, as if I’m trying to prick the balloon of self-confidence and remind myself of “reality”.

But the polls never do quite take you to reality. Instead, they shape it. It’s not just what the polls are saying, or even how they were put together, that’s the great problem here – it’s how the obsessive focus on polls is symptomatic of how we view politics.

Polls make politics feel like a race, a game, a sport of feuding personalities. Who’s up? Who’s down? What tactics have they used to get one over on each other? What does it say about their personality? Words are seen as weapons with which politicians show off their ability to subvert or scare the opposition – not as substantive statements about what they intend to do.

And what sort of politician will thrive in this world where political speech is just a game? A candidate such as Donald Trump.

It was the communications professors Kathleen Hall Jamieson and Joseph Cappella who first noticed the connection between describing politics as a series of strategies and a growing cynicism among voters.

This was back in the mid-1990s, when the media was constantly analysing the rivalry between US president Bill Clinton and speaker of the house Newt Gingrich, the early iteration of today’s identity-based partisanship. Jamieson and Cappella found the media was focusing less on the issues the two were debating – often around health reform – and more on how they were competing.

The coverage fixated on who was winning, utilised the language of games and war, emphasised the performance and perception of politicians, put a new weight on polls.

This sort of coverage activated people’s cynicism about politics – the sense that it’s just a game between self-serving schemers – and then made them more cynical about the media.

Decades later, this “spiral of cynicism” is all around us: from the exploding popcorn of polls to the headlines. After Trump’s former chief of staff John Kelly compared him to a fascist last week, the Wall Street Journal wrote: “Harris uses ex-Trump chief of staff’s remarks to paint him as unfit for office”.

The question of whether Trump is a fascist or not was reduced to highlighting a rhetorical tactic. The idea that all politics is just a cynical game, and that the “mainstream media” is not really looking out for the cares of the voter, has become so pervasive it has helped pave the way for politicians who stand on sweeping away the whole edifice of democracy as we know it.

It’s no coincidence that this turn began in the 1990s, when the cold war had finished and the big philosophical debates about policy seemed to be over. Instead, politics became about entertaining performance – the era of Blair, Clinton, Zhirinovsky, Yeltsin. And the media began overgenerating coverage that replaced ideological debate with personality and tactics.

The 1990s were also when the reality show emerged as the dominant entertainment format. It initially grew out of observational documentaries seeking to understand society better by ceaselessly filming ordinary people in their homes in such a way that they would forget about the cameras and be more themselves.

It quickly became the opposite: a circus where all behaviour was for the cameras. Contestants learned to say and do the most vile things just to engineer scandal and generate attention for themselves.

American political TV debates started to imitate the same logic. In a busy primary debate, candidates only get a little sliver of airtime. The way to get more is to attack another candidate in the meanest and most personal way possible, and thus provoke them to attack you back. If you are attacked, then you are allowed more time to respond.

So you quickly got debates where supremely clever candidates sling personal abuse at each other to get more attention. The debate stage was set for reality show host Trump.

Analysis and opinion on the week's news and culture brought to you by the best Observer writers

after newsletter promotion

The design of most social media has followed the same incentives: rewarding taking the most extreme and often nasty statements to generate attention. And Trump has flourished on that as well.

When pollsters gave voters a choice of policies, rather than personalities, the majority, including Trump supporters, preferred Harris’s

The 1990s is when World Wrestling Entertainment (WWE) boomed, with its cabaret wrestlers pulling obviously fake fighting moves, where violence is theatre. Trump was always an aficionado of WWE, even taking part in mock fights, and a member of its hall of fame.

This year the 1990s wrestling star Hulk Hogan spoke at the Republican National Convention; Trump enters his own rallies to the theme tune of the Undertaker, who, at the height of WWE, was the “evil” foil to Hogan’s all-American “goodie”. Many of Trump’s followers apply the cultural logic of WWE to his statements. Sure, the argument goes, Trump might say some very authoritarian-sounding things – but it’s just a game.

So can we ever find a way back to reality? To issues rather than strategies? We can, and we can even use polling to do so. When pollsters recently gave voters a choice of policies, rather than personalities, to choose from in this election, the majority, including Trump supporters, preferred Kamala Harris’s.

Partisan polarisation dissolves when we change how we cover politics. We can also develop different TV political debates, which preserve the excitement of competition but repurpose them to reward collaboration instead of abuse.

Imagine a debate format where candidates had to solve a real policy problem, and show how they would work with each other and with the opposition party to achieve it. We could also scale social media platforms that algorithmically detect the commonalities in political disagreements to generate common policy solutions. Such platforms are already being used in Taiwan.

Of course, there’s appeal in fleeing from reality to the grotesque circus of politics. But if we can’t face facts, others will force us. This month, at the Wilson Center in DC, Jack Watling of the Royal United Services Institute and Sam Cranny-Evans of the Open Source Centre presented a chilling analysis of Russian weapons manufacturing and supply chains.

The slideshow featured satellite photos of munitions factories where freshly cleared tracts of land are being readied to produce more weapons. Vladimir Putin is preparing for a vast war. China’s arms production is on a wartime footing. They are not playing.

Peter Pomerantsev is the author of How to Win an Information War, the Propagandist Who Outwitted Hitler