The following data was collected from the latest Decision Desk HQ 2020 Election Forecast updates on September 29th. PredictIt’s market prices correspond to October 1st. Numbers, screenshots and analysis are subject to change over time.
We can all agree this has not been an easy week for Donald Trump, Republicans and their options to win in the upcoming 2020 Election. The New York Times report on Sunday about the president's tax returns over the past decade has now been joined by the hangover of a rough and chaotic first presidential debate which has not exactly favored the image of the GOP candidate. However, these last events have not managed to put forecasting models and betting markets in full agreement yet.
PredictIt, the most notable betting exchange in the US, can give us a clear idea of the latest political betting sentiment in the country. Joe Biden and the Democratic Party are currently trading at 63¢ to win the presidency next November. A Donald Trump re-election is at 42¢ according to traders.
These numbers differ a lot from most of the pundit data models released over the past few weeks. Democrats are given more than 80% chances to win in most cases, leaving Republicans with a likelihood of 15-20% to win. This extends to a greater or lesser extent to practically all the races up for grabs on Election day: popular vote by state, Senate and House control. Democrats dominate in the models.
The following comparison will use the numbers provided by the Decision Desk HQ 2020 Elections Forecast as a pundit model, and compare them with the prices on PredictIt's betting exchange. Remember you can follow all the action on these markets and those from other international exchanges through guesser.com. Keep reading to find all the details about the data below!
This PredictIt market, which accumulates the highest trading volume within the platform, saw how Biden and Trump odds tighten almost 10 points during summer, but the Democratic candidate is now taking advantage again with, at the time of writing, 63% chance of being elected president, a 5% more than the last week.
Bettors are still uneasy about a comfortable victory for the former Vice President, contrary to the Decision Desk HQ model, that currently predicts for Biden more than a 82% chance of winning, 20 points above the traders consensus on the exchange! The model also sets right now a 311 votes victory at the Electoral College for Democrats, while Republicans would only get 227.
For his part, President Trump keeps coming out highly favored by PredictIt prices, which boost him to a 42% likelihood of being re-elected. This, compared to the 16% that DDHQ sets on its model, may be hopeful for the Republican Party expectations, which is being highly handicapped by all the recent pundit models.
PRESIDENTIAL POPULAR VOTE BY STATE
When it comes to analyzing the 2020 Election popular vote, most of the swing states that boosted Donald Trump to victory in 2016 get all the attention within forecasting models, as well as the largest betting volumes in PredictIt.
In this case we are going to compare the states that the Decision Desk HQ model sets as ‘toss-up’, that is, with the probabilities of both parties more evenly matched, against their respective PredictIt prices. We will also look into those that differ the most in probabilities from one site to another.
Top 3 Toss-up states according to the DDHQ Model:
Ohio is the current most contested battleground in the DDHQ forecast, which expects a Republican victory with only 51.4% chances.
This highly disputed status doesn’t change too much for the state within PredictIt, where political bettors currently give the victory to Republicans by only 10 points, setting their odds at 56¢ vs 46¢ Democrats. Important to attend the latest GOP’s price trend: 10¢ drop in a single week! They may not have much left as favorites in the betting market.
Reminder: Donald Trump won this state in 2016 with a 8.13% margin of victory.
In the state of North Carolina the DDHQ forecast expects this time a lean Democratic victory with almost 54% chances.
Options for the Democratic Party are slightly reduced in PredictIt, where they are currently priced at 52¢, setting this way one of the most contested markets along the whole betting platform. Republicans have recently lost their lead in this state, where their price has dropped 5 points during the last week.
Reminder: Donald Trump won this state in 2016 with a 3.66% margin of victory.
Same as Ohio, Georgia leans slightly for Republicans in the DDHQ model, who get a 54.6% chance of victory.
PredictIt bettors have here a clearer idea of who’s the state favorite, pricing the GOP at 62¢ over the Democrats, which trade at 42¢ only. But the Republican’s trend has not been especially favorable during these last days, it’s price having dropped up to 4 points.
Reminder: Donald Trump won this state in 2016 with a 5.13% margin of victory.
Top 3 states which differ the most between predictions:
Despite common consensus around Republican dominance in Texas, this key state is the one that currently has the largest differences between the forecast and the market.
While the DDHQ model sets a 63.4% probability for a GOP victory, PredictIt’s market raises the odds up to 74%, a spread of more than 10 points that practically makes the state go from a lean to a likely Republican scenario. Donald Trump’s party has seen its odds rise more than 10¢ over the summer and it seems they’re managing to maintain this wide advantage for the moment. Traders are pricing Democrats at 29¢ as opposed to the 36.6% set by the model.
Reminder: Donald Trump won this state in 2016 with a 8.99% margin of victory.
The way bettors and the model differ in New Hampshire is also remarkable.
Here we can find almost a 8% spread in the Democratic win, which is predicted with a 71.3% probability in the DDHQ model but priced at 79¢ on PredictIt.
Consequently the Republican victory odds go up to 22% in the exchange compared to the 28.6% set in the model. The price trend for the Republicans has been really unfavorable in the betting market during the last week, having dropped up to 7 points.
Reminder: Hillary Clinton won this state in 2016 with a 0.35% margin of victory.
Florida is our last example of battleground state that notably differs between the model and PredictIt. Here DDHQ is forecasting a 59.3% chance to win for Democrats, President Trump’s options falling almost to 40% in a key state he’s been targeting throughout all his 2020 campaign.
The scenario in the Sunshine State is much more optimistic for Republicans in the betting markets, where their victory is currently priced at 50¢ as traders set a 10 point spread in relation to the model. The Democratic Party remains favorite, but trading at 52¢ only.
Reminder: Donald Trump won this state in 2016 with a 1.20% margin of victory.
The 2020 Senate race is undoubtedly being the most difficult one to predict for pundit models. Based on the fact that the current Senate is controlled by Republicans with 53 seats and that this year’s election 23 of the 35 seats up for grabs are held by them, forecasts tend to set a low margin victory for the Democratic Party. In any case, Democrats are not given, in a ‘safe’ way, more than 50 seats within none of the main models.
The DDHQ forecast is currently predicting a 52 seat Democrat control after 2020, leaving the GOP with 48 senators in the upper chamber. This would suppose a +5 net change for the Democrats, which at the same time is the second most probable outcome according to PredictIt, which prices it at 16¢ on it’s market about Senate change in seats.
But the most interesting prediction we can confront to the betting markets within the DDHQ forecast is the chance of a Senate majority, which at the time of writing is predicted 78.4% in favor of the Democratic Party in the model. The chance of a tied (50-50) chamber is set at 11.7%, leaving below Republican odds, that don’t even reach 10%.
These numbers change a lot if we move to PredictIt. Within the ‘Who will control the Senate after 2020?’ market available on the betting exchange, Democrats are currently priced at only 61¢, more than 17% change with respect to the DDHQ model!
Despite the fact that this is probably, and in general terms, the least contested race in all the 2020 Election, the expected Democratic victory has also different probabilities according to the pundit models and the betting markets.
Forecasters predict Democrat control of the lower house in more than 220 seats of the 435 up for grabs in all cases, with some models eyeing even more than 240 seats. The DDHQ model particularly predicts 237 seats won by the Democratic Party on Election night, giving them almost a 99% chance of a House majority.
The result of this race seems to be clearly advantageous to the DNC in the betting markets too, but Democrats’ prices here are reduced by 13%, with odds at 86¢. Therefore Republicans are trading slightly higher, at 16¢ at the time of writing.
This summary comparison table is the end of the analysis! Let us know what you think on Twitter and feel free to share with your friends. Subscribe below for weekly updates like this one!