Kelly Criterion makes no sense in this context as it assumes every bet is completely independent. As states are highly correlated, kelly betting will lead to vast overbetting and negative returns (even if your predictions are much more accurate.) I made a >20% return betting on 538 and kelly correctly.
We propose a market-based scoring method for evaluating the performance of probabilistic forecasts of the 2020 U.S. elections for President, Senate and House of Representatives. We demonstrate this method by evaluating the forecasts posted on the FiveThirtyEight website based on their performance against the prediction markets at PredictIt.
In addition to the analysis presented here, a running tally of results from the above analysis was updated and reported throughout the 2020 campaign at https://pivs538.herokuapp.com/.
➤ Version 1 (2020-10-26)
Harry Crane and Darrion Vinson (2020). Models vs. Markets: Forecasting the 2020 U.S. election. Researchers.One, https://researchers.one/articles/Models-vs-Markets-Forecasting-the-2020-US-election/97212967bc124e2bf3af31a5/v1.