US Election Bet of the Day: Expect Trump's vote share to prove predictable once again

• Paul Krishnamurty is picking seven election bets across seven days

• The final bet is on Donald Trump's vote percentage

• Read more in our 2024 US Presidential Election Live Blog here

• View Betfair Exchange politics markets as you have never seen before at Betfair Predicts

• Place £5 on a US Election market and receive a £5 free bet

---

Bet of the Day 7: Back Trump Vote Percentage to be 45.00-47.99% @ 2.77/4

The outright market bounced slightly back towards Kamala Harris last night after another wave of Trump money, settling around 1.594/7 vs 2.727/4 in the latter's favour. I suspect the range for Trump's vote share is also close to its peak.

Trump's vote share, or approval rating, does not look a very hard number to quantify. Perhaps no politician in US history divides voters so clearly and permanently. A very large number strongly approve or disapprove, with a smaller number in the middle who invariably fall on the same side anyway.

In his two elections, Trump earned 46.1% and 46.8% of the popular vote. He is in a better position this time around so that number can be expected to rise but to get beyond 48%, as the current market implies, remains quite a big ask. I cannot envisage any scenario where he gets less than 45% and if that assumption proves correct, today's bet will amount to taking 2.77/4 about under 48%.

If he does get around 48%, that will more than likely be enough to win the electoral college. Hillary Clinton won the popular vote by 2.1%, and Joe Biden by 4.5%, yet still only won the key states by narrow margins.

In 2020, 98.1% of the electorate voted for Trump or Biden. That was a particularly binary election, without any prominent third party option. In 2016, when third parties were a factor, the 'big-two' only won 94.3% combined. In 2012 it was 98.3%, again lacking a strong third party.

This time, third parties are at least more newsworthy, albeit scattered around on some ballots but not others. Therefore, I think Trump and Harris will struggle to get more than 98% combined. If the number is exactly 98%, then Trump getting below 48% would equate to a 2% popular vote win for Harris. My ballpark figure is a popular vote win for Harris between 1.5-3.0%. On that basis, 2.77/4 makes sense and if that combined Trump/Harris share is below 98%, her required advantage becomes less.

You may notice a theme running through this 'bet of the day' series. My plan is to take positions about both sides, in what is an unpredictable election, which could all or mostly win, regardless of who becomes president, assuming the race is this close. So this bet very much correlates with the first - on the election winner losing the popular vote. Exactly where my current prediction lands - Harris popular vote by 1.5-3.0%, Trump wins the electoral college.

Likewise, that outcome could also see Harris win two of the three states for which I have recommended her in this series, and most permutations in that range produce 270-299 Trump Electoral College Votes. From now, I will use the Live Blog to refer to or update these positions.

Bet of the Day 6: Back Kamala Harris to win Michigan @ 1.855/6

Of the seven swing states, Michigan was the one most experts regarded as the best Harris opportunity. Joe Biden won it by 3% and during August and September, various high-quality polls pointed to more of the same.

Then the Trump surge came and nerves began to set in around the Democrats. It was reported that they were worried about Michigan, and with some good reason. Arab Americans are a key part of the Democrat coalition here and there is evidently a backlash against the administration's support for Israel over the Gaza conflict. Biden won this group by 2:1 in 2020, but polls show Trump and Harris almost tied among Arab Americans.

Furthermore, third party candidates are on the ballot and set to eat into that Democrat share. Just as Hillary Clinton lost in 2016 due to third party votes going to Gary Johnson and Jill Stein, so this time Stein and Cornel West present a mortal threat. This particular angle very much cooled my optimism about Harris winning it and compounded doubts that she can win the presidency. Without Michigan, winning the electoral college is going to be an uphill struggle.

However, it seems to be turning back towards her. While one should never assume too much about the early vote, the general signals seem positive for Harris. Notably an 11% turnout lead for women. The concerted Democrat campaign, supported by the Obamas and Eminem in Detroit, appears to be working.

And further positive signs can be found in the polls. YouGov have her 7% up over Trump, on 52%. Susquehanna have the lead at 5%. On the flipside, Emerson have Trump up by 1%. (This respected firm have generally produced better numbers for Trump lately).

As for those third parties, the effect may not be as damaging, especially after this week's final ruling that Robert F Kennedy Jr. must stay on the ballot. He appears to be taking votes from Trump - hence his fighting to the last minute, in vain, to be removed. Local polling firm Glengariff Group (for Detroit News) have RFK taking 4%, and Trump down to 44%.

Harris is solid around 47%, ahead of late undecideds breaking, even in polls naming the third party candidates. So Trump urgently needs any RFK voters to flip, assuming they haven't voted for him already.

Moreover, the broader signals point to this state leaning blue. Governor Gretchen Whitmer - a co-chair of the Harris campaign - won by 11% following years of confrontation with MAGA. Democrats won the state Senate elections in 2022 (a year in which Republicans had a turnout bias). Elissa Slotkin remains around 3% up in the Senate race.

---

Bet of the Day 5: Back Republicans to win Wisconsin Senate race @ 3.65

The reasoning behind yesterday's bet revolved around the strong correlation between Senate and Presidential vote shares during the ultra-polarised Trump era, and how that may affect the late swing in these states.

That bet on Arizona (see below) needs the Democrat's Senate candidate to drag Kamala Harris up, whereas today's bet in Wisconsin, the exact opposite dynamic is required, and it seems to be happening.

As far as the presidency is concerned, Wisconsin is on a knife-edge. Trump won and lost it by small margins in 2016 and 2020 and the betting remains close, with the former president now ahead at 1.784/5. This is quite a turnaround, given that the Democrats were matched at 1.528/15 earlier in the campaign.

Likewise a few weeks ago, Tammy Baldwin was consistently and clearly ahead in the polls as she bids for a third election in the Badger State. Suddenly, the race is too close to call. The last six polls all showed the candidates within 1% of each other. That list includes high-quality firms Suffolk, Emerson and Quinnipiac.

Baldwin may well hang on, boosted by being the incumbent. It was notable how the opposite Senator, the Republican Ron Johnson, defied a swing towards the Democrats to be re-elected. This is a purple state, where ticket-splitting is not uncommon. Likewise, Harris may well still win a close race.

However, the value has to be with Hovde at these attractive odds. If Trump is favourite to win the state, and I agree he deserves to be so marginally, then one must think the Republican Senate candidate will get extremely close.

---

Bet of the Day 4: Back Democrats to win Arizona @ 3.613/5

The presidential polls remain remarkably consistent, whether at a national or state level. The three bets put up so far would correlate with those polls being broadly accurate but, in truth, I am somewhat suspicious about their consistency - an idea discussed in my piece from last week regarding precisely how the polls could be wrong. Yesterday's list smacks of herding. In which case, the polls could be wrong in either direction.

Arizona's numbers are particularly hard to comprehend, due to their projecting a level of ticket-splitting which is unprecedented during the Trump era. As laid out earlier in the cycle and updated in the Live Blog, there is usually a close correlation between Senate and Presidential candidates. Above 1% difference is notable and 2% extreme. Arizona has seen 2% ticket-splitting, with popular Senator Mark Kelly outperforming Biden by 2% and in 2016, John McCain massively outperforming Trump.

This time, Democrats look near-certain to win the Senate. Ruben Gallego leads arch-Trump ally Kari Lake by 6.2% on the Fivethirtyeight polling average, well above 50%. He is rated 88% likely to win on their forecast and is trading at 1.162/13 in our Arizona Senate market. Whereas the same prediction models have Trump leading by 2.2% and 66% to win.

I find it hard to believe that the difference between these two polling leads will be the same on election day. It would point to a level of crossover and persuadability between America's two political tribes, which seems highly counter-intuitive in an era of ultra-polarisation. If the reason the national polls are so fixed is the lack of crossover, then why is this such an outlier?

I expect either Lake or Harris will fare much better than expected. Part of the explanation for the difference may have been Harris underperforming with Latinos. Gallego may well be able to keep those voters in the Democratic column and push late deciders her way.

In the 2022 midterms, Kari Lake's chance in the Governor race was massively overestimated in polls, as was the other Trumpian Senate candidate, Blake Masters. The Republicans haven't won a statewide race here since 2018. Arizona's independent voters are made more in Kelly or McCain's moderate image, and have rejected extremist MAGA politics in recent elections. If there is going to be a big polling miss, or betting upset, in any of the seven swing states, I believe Arizona is the place.

---

Bet of the Day 3: Back 270-299 Trump Electoral College Votes @ 4.03/1

Betting on the electoral college margin is always one of the highlights of a US Election. A really tricky puzzle which is extremely hard to price accurately, which can offer great rewards if you solve it. At this stage, my focus is on building a position that will look better once the in-play action develops on election night, from which to trade.

So, let's do the maths. First, assume that only the seven close swing states are up for grabs, and everything else stays with the 2020 winner. If Trump wins all seven, he ends up on 312 electoral college votes (ECVs). If Harris does the sweep, she gets 319.

Of course more extreme wins are possible, but they involve very big upsets on a scale almost unheard of in presidential elections - eg New Hampshire for Trump is a 7.06/1 chance, Minnesota 9.08/1. while Texas for Harris is a 15.5 chance.

Based on the outright odds, Trump is 67% likely to win. In the Trump Electoral College Votes market, 270-299 is a 4.03/1 or 25% chance while 300-329 is 3.65 (27%). Anything higher involves the aforementioned massive outlier.

I reckon the 270-299 band is considerably overpriced. If Trump loses one any of Pennsylvania (19 ECVs), North Carolina (16), Georgia (16) or Michigan (15), he falls short of 300. Alternatively, Trump could land on 270 whilst losing four of the seven key states - Nevada, Wisconsin, Arizona and Michigan. This band covers many highly plausible permutations.

All the seven are very close and likely to remain so during the live counting. If Trump looks very likely to win at that stage, I'd still be very hopeful of being able to cash out this 4.03/1 bet on the upper side. Perhaps with just one bet on the single state which looks most problematic.

Likewise, if Harris is ahead or looking likelier to win, it won't be too hard or expensive to exit this position, as it would equate to any Trump win (ie, once a Trump sweep is off the table). Indeed from a trading perspective, the risk appears very low, compared to substantial rewards.

---

Bet of the Day 2: Back Kamala Harris to win Pennsylvania @ 2.265/4

For logical reasons, Pennsylvania has become seen as the single key state. Its 19 electoral college votes appear essential to both candidates. That is certainly the case for Harris, whose pathway to electoral college victory looks extremely thin without it.

Historically, this is seen as the toughest task for a Democrat among their firewall in the Great Lakes states. I am less convinced about that this time, and think it may prove easier than either Michigan and Wisconsin. In other words, she could very well win Pennsylvania yet still fall short in the electoral college.

Michigan is problematic due to the Arab-American vote, and anger about Democrat support for Israel's actions in Gaza. Wisconsin's layout is predominantly rural counties and therefore ripe for Trump to juice turnout. But Pennsylvania has been trending blue for some time and those factors are slightly different.

Whereas foreign policy may hurt Harris in Michigan, Trump may be damaged here by his support for Vladimir Putin. As explained in the Live Blog last month, there are over 100,000 people of Ukrainian descent in Pennsylvania and 800,000 Polish-Americans.

More widely, Democrats have high hopes of flipping the state legislature in November, overcoming major disadvantages due to gerrymandering. Governor Josh Shapiro hammered the Trump-endorsed, election-denying Doug Mastriano in the 2022 Gubernatorial race by a 15% margin.

Trump has assumed favouritism here due to a decent run of polls, but they have rarely been from firms with pedigree. This may be an example of Republican-tied firms swamping and skewing the averages. Whereas A-grade Siena College and George Mason University had Harris between 2-4% up, and on 49/50%.

Note too that there appears to be meaningful resistance to Trump in the state. In the GOP primary, Nikki Haley secured 157,000 votes long after withdrawing from the race. These are the critical voters being targeted relentlessly by Harris and the senior Republicans endorsing her such as Liz Cheney.

If you think Harris has a good chance of winning overall, my advice is this state is a better bet at 2.265/4 than 2.68/5 for the presidency. I make her favourite to win here.

---

Bet of the Day 1: Back 'Yes' in Will Election Winner Lose Popular Vote? @ 2.747/4

If Trump is the favourite, this bet must represent good value.

The odds to win the US Election have swung sharply towards Donald Trump in recent weeks. Whether that proves to be vindicated or not, those odds around 1.664/6 do not appear to correlate with this bet at 2.747/4.

No question, Trump can win the electoral college because the margin in each of the seven swing states is tiny. He did so in 2016, and failed narrowly in 2020.

But winning the popular vote is quite another level of achievement. Last time, he lost to Joe Biden by just over 7 million votes and even against Hillary Clinton, he lost by just under 3 million.

In fact, the only Republican to win the popular vote since the 1980s was George W Bush in 2004. At the time, Bush was the incumbent, with the country relatively united in the aftermath of the 9/11 terror attacks and at war on two fronts. The context with this ultra-polarised Trump era is incomparable.

Note that this bet would land if today's forecasts are correct. The polling models all make Trump a slight favourite to win the electoral college, and Harris a stronger favourite for the popular vote.

Betfair odds are more confident about the former, with Trump's odds to win implying around a 58% chance, while Harris is around 68% likely to win the popular vote. Odds of 2.767/4 about this bet imply a 36% chance. Too low.

---

Now head to the Politics live blog for today's latest election news!

---