Binyamin Netanyahu has been here before. In 2009, Kadima won one more seat than Netanyahu, and he became prime minister anyway. In September 2019, Benny Gantz's Blue and White did the same, and Mr Netanyahu stayed in office for another 21 months. Coming second has not always meant leaving.

Israelis vote on Tuesday, and he is expected to come second again. His likeliest conqueror is Gadi Eisenkot, a former army chief who sat in the war cabinet after October 7th, lost a son in Gaza two months later and founded his Yashar party only last year. Polymarket gives Yashar a 71.5% chance of winning the most seats, up 3.5 points in a week, and four polls published in the past week all put it ahead of Likud, by two to four seats.

But winning the most seats is not the same as winning power. Israelis elect the 120 members of the Knesset, and the prime minister is whoever can command a majority of them. Until somebody does, the old prime minister stays on as caretaker, with almost all the powers of the job. That is Mr Netanyahu's silver lining.

Neither bloc looks close to a majority on its own. In the polls that report blocs, Mr Netanyahu's camp wins 48 to 51 seats and the opposition 51 to 55, both short of the 61 that make a majority. The Arab parties, on about 14, could close the gap, but several of Mr Eisenkot's allies refuse to rely on them. Traders have done the sums: they give Mr Eisenkot only 52% of becoming prime minister.

The most telling price is on Mr Netanyahu himself. Traders give him only 27.5% of being sworn in again, yet just 39.5% of leaving office, or announcing that he will, by New Year. Put plainly, traders expect him to lose power, but not quickly: they see about a three-in-five chance that he is still prime minister on New Year's Eve.

The reason is the calendar. Official results are due on November 4th. The president then asks one member to form a government, with 28 days to do it and a fortnight more if needed. In 2022 the whole process took 58 days, and in 2021 it took 82. At the first pace a new government would be sworn in on Christmas Eve; at the second, in mid-January. All that time, Mr Netanyahu stays in charge. If it fails, Israel votes again with him still there.

After the election of April 2019 nobody could form a government. Israel voted three times in 11 months, and the prime minister sworn in at the end of it, in May 2020, was Mr Netanyahu.

The risk runs both ways. Polls have underrated Likud before: in 2015 the final surveys gave it 20 to 23 seats, and it won 30. A repeat would hand Mr Netanyahu a real victory. Or Likud could slump. Traders give it a 28% chance of finishing below 20 seats, up three points in a week, which would make the opposition's sums much easier. A surge would upset traders' view of the vote; a slump could cut short the wait they expect.

We think traders have it right. Mr Eisenkot will probably come first, and Mr Netanyahu will probably lose his job, but the vote will be the quick part.

Twice before, second place left Mr Netanyahu holding the keys. Traders think the third time will be different, just not by Christmas.

Democrats are about as sure of winning the House as anyone gets nine days before an election: Polymarket gives them 91.5%. Traders have grown keen on their chances in the Senate too, at 66.5%, up two points in a week.

They are different contests. The House has 435 seats, and a national swing towards the Democrats should carry it. The Senate will be settled in a handful of states, most of them Republican ground. Democrats need a net gain of four seats.

In practice that means North Carolina, which traders all but hand them, plus at least three of the next likeliest, Maine, Ohio, Texas and Alaska, while holding every seat they already have. Donald Trump won three of those four in 2024, and Maine has elected Susan Collins five times.

Traders make Democrats favourites in all four. That is a lot of faith in polls that have flattered Democrats in exactly these places. Polling errors also travel together: if the polls are off in one of these races, they are probably off, in the same direction, in all four.

Psychologists have a name for what traders are doing: the halo effect, in which a strong impression in one place colours judgment in another. A near-certain House has lent the Senate a glow it has not earned. We think the Senate is closer to a coin toss than the price suggests.

Some $66.7m has changed hands on Polymarket over whether Jesus Christ returns before 2027. That is more than seven times as much as traders have bet on which party wins the US Senate. The Second Coming trades at 1.4%.

What on earth is going on?

That is not a statement of faith. Almost all the money is on the other side: traders buying 'No' at about 98.6 cents and collecting a dollar after New Year. That earns 1.4% in ten weeks, or more than 7% a year, at a time when American Treasury bills pay a little over 4%. For a big trader it is a high-interest savings account that happens to mention the Messiah.

The $66.7m is real, but it is not $66.7m of conviction. Only about $1m is staked at the moment; the rest is the same money changing hands many times over. The price behaves like a yield too. It peaked at 4.65% in February and has drifted down as the calendar has run out.

Our analysis puts it at 0%.

That is a judgment about the contract, not about theology. The 1.4% is made of three things. About half is what the money tied up in 'No' could earn in Treasury bills instead. Some is a premium for leaving cash on a crypto platform until January. The rest is the small chance that the contract settles wrongly: its verdict rests on 'a consensus of credible sources', decided in a dispute by a vote of token holders.

For our readers who are not au fait with the french national motto of Liberty, Equality and Fraternity, the nation is about as far away from that as possible.

Regional Mayors declared curfews, restricting movement after dark. Riots erupted due to societal inequality. Citizens burned their historic monuments to the ground.

Polymarket gives France a 6.5% chance of declaring a state of emergency by the end of October. That looks about right. The prime minister, Sébastien Lecornu, says curfews suit the situation better, and they are already in force in parts of several cities, Paris and Lyon among them. Here is the other side of the argument.

Start with what has already happened. On October 17th more than half a million people marched over the cost of living and the state of schools, and the night turned into rioting and looting. More than 2,400 people have been arrested, almost 1,000 police and gendarmes hurt and at least 3,000 businesses damaged or looted.

The interior minister, Laurent Nuñez, offered to resign. Emmanuel Macron refused.

Then listen to the people in charge. Asked about a state of emergency on October 19th, Mr Lecornu did not rule one out: 'If it is necessary, of course we could do it.' Mr Macron called for firmness.

The calendar is the next problem. The yellow-vest coordination behind the first protests called fresh action for October 24th and again for October 31st, the last day of the contract. Student organisations have vowed to stay mobilised, and unions have called a national day of action for November 5th. The government is sending soldiers from Sentinelle, its anti-terror patrol, to guard sensitive sites so that more police can work the streets.

And France has done this before. In 2005 Jacques Chirac declared a state of emergency on the twelfth day of rioting, under a law of 1955 that lets local authorities impose curfews, search homes and ban gatherings. It stayed in force until January.

The case against is strong.

So the contract comes down to the coming week, and one last Saturday. If this weekend passes quietly, 6.5% will look generous. If not, a State of Emergency is the next step.

Keep an eye on France next weekend.

We’re up

A real $10,000 book on Polymarket. See all of our current positions - @thebasementview.

Yes on Scott Gillingham winning Winnipeg's mayoral election: 371 shares at 89%. Cost: $330.16, with no fee.

Rationale: Winnipeg votes on October 28th, and the candidate with the most votes wins, with no runoff. When we bought, the only poll, taken in September for the Winnipeg Free Press, gave Mr Gillingham a 32-point lead among decided voters. The final poll by the same firm, Abacus Data, taken between October 16th and 22nd, cut that to 17: Mr Gillingham 36%, Kevin Klein 19%, Mike Vogiatzakis and Noah Redden 14% each. The vote against him is split three ways, and every Winnipeg mayor who has stood for re-election in the past 60 years has won. Polymarket priced him at 89%. We think it is nearer 94%.

Expected return: If Mr Gillingham wins we collect $370.97, a profit of $40.81. Otherwise we lose $330.16. At a 94% chance that is worth about $20, roughly 4.5% in a week. It only loses money on average if the true chance is below 89%.

Risks: Nearly two thirds of decided voters back someone else, and Mr Gillingham's share fell 19 points in five weeks. If the anti-incumbent vote gathers behind one challenger in the final days, a 17-point lead can go. CBC News also cautions that the latest poll used an online panel and two questions that could be read as leading. Advance turnout hit a record, and nobody yet knows whom it favours.

Nothing this week.

Know one person who is fed up with traditional news? Forward them this.

Been forwarded this? The button gets you your own copy every Sunday.