WASHINGTON — To avoid having Stephen Harper's debate preparations destabilized by some scandal-of-the-day, his entourage once banned smartphones from the room and devised a coded phrase to use in his presence.
They feared opposition parties would leak their most damaging material to media just before a debate, to pull the prime minister and his top aides out of debate training and into a time-wasting exercise in damage control.
So they used covert language around the boss.
"Somebody would come in and (say), 'The pizza delivery has arrived,'" former aide Dimitri Soudas recalls of the 2011 campaign.
"(So you don't have) the prime minister saying, 'What's happening? What's going on? Is there an issue?' I wanted to avoid at all costs the prime minister at the time getting distracted from doing debate prep."
These are the kinds of psychological games going on now.
Hillary Clinton and Donald Trump will square off at 9 p.m. ET Monday and huge ratings are expected — the race is close, and it's fair to say one of the participants is indisputably the most unorthodox candidate to ever appear in a one-on-one U.S. presidential debate.
The mind games are well underway.
Both sides are working the referees — and each other. For starters, the Clinton camp has invited Trump's billionaire nemesis Mark Cuban to sit in the front row; Trump has responded by floating the idea of inviting Gennifer Flowers, Bill Clinton's former mistress.
The Clinton camp is also pressing the moderator, demanding he call out lies.
Her communications director Jennifer Palmieri tweeted: "Seventy per cent of Trump's claims are untrue. Special measures are required this year... (We) aren't asking them to call balls and strikes, just debunked lies Trump always tells." Republicans, meanwhile, still chafe at what they view as a 2012 moderator's liberal fact-checking — they're urging moderator Lester Holt to be a neutral observer.
An NBC staffer close to Holt was quoted telling CNN: "Lester is not going to be a potted plant."
They're also working the audience, when it comes to viewer expectations.
Trump's team tells reporters he hasn't prepared much, in an effort many view as setting up viewers to be pleasantly surprised by his performance. Clinton's team says that's unfair. They want the candidates judged by a higher standard, of a potential president — and not by the lower standard of someone managing to exude basic professional normalcy for 90 minutes.
"I'm hoping there isn't a different standard in the debate," said Clinton's running mate, Tim Kaine. "Let it be an even standard for both."
He said Trump must be treated like any other candidate. That means, Kaine said, explaining why he hasn't released his tax returns, like all other recent candidates; and offering detailed policy answers, not just soundbites and bluster.
In the end comes the ultimate psychological test: voter response.
Public perception can be swayed by initial reports about who won, suggests a book on the 2012 election, "The Gamble," by a pair of prominent political scientists. It says a majority of viewers immediately agreed Mitt Romney won the first debate — the margins of victory were 24 and 42 per cent in a pair of polls conducted on the night of the debate. Yet the margin grew; it was 48 per cent in another poll a few days later.
At the end of the day, do any of these head-games matter?
In most elections, the polling swing after the first debate is about two or three percentage points. In, "The Timeline of Presidential Elections," Robert Erikson and Christopher Wlezien compared polling data in the week before and after first debates and found a shift within that range in every race since 1980 — with the exceptions of Mitt Romney gaining four points in 2012, and John Kerry gaining six in 2004.
Even that four-point swing toward Romney was a bit of a mirage, says a former aide.
Stuart Stevens says those were mostly people who'd already supported Romney, but they disliked his disparaging comments a couple of weeks earlier about the 47 per cent of Americans not paying taxes, so they'd temporarily called themselves undecideds.
"The debate really just sort of got back voters that drifted away from Romney over the '47 per cent' thing," Stevens said in an interview.
"The race was pretty stable. A bit like this race."
And that's another thing about the first debate: Its effects are ephemeral. Romney lost, as did Kerry. In the nine elections measured by Erikson and Wlezien, the eventual winner was the post-debate loser six times.
In, "The Gamble," John Sides and Lynn Vavreck describe how it happens. In 2012, right after Romney won the first debate a good jobs report came out. That helped Barack Obama limit the damage.
And then Obama did better in the next two.
Another behavioural quirk of debates is the generally futile wait for the proverbial knockout punch. Soudas noted the best-known examples of the devastating put-down came in an earlier television era, from Ronald Reagan and Brian Mulroney.
Now it's a zinger every few minutes.
"It was a novelty (back then)... The format wasn't as well understood," Soudas said.
"The art of the soundbite has now been mastered."
