The Jerome model published an updated forecast of the election outcome. It currently forecasts a two-party vote share of 44.3% for Clinton, and 55.7% for Trump in Virginia. In comparison, on August 16, Clinton was predicted to collect 55.7% of the vote.
In Virginia, the popular vote is often close. This is why the state is commonly referred to as a swing state, which makes it particularly interesting from a forecasting perspective.
Putting the results in context
Individual models can contain substantial biases, and should be treated with caution. Rather, one should examine how a model's results compare to benchmark forecasts.
The Jerome model compared with PollyVote's forecast
The most recent PollyVote anticipates Trump to gain 47.7% of the two-party vote in Virginia. The results of the Jerome model for Trump are thus 8.0 percentage points above the combined PollyVote in Virginia.