Table 1.

Comparison of forecast performance between Dante and DBM.

Metric	Model	States	HHS Regions	National
Skill	Dante	0.372 (0.060)	0.413 (0.073)	0.439 (0.078)
	DBM	0.337 (0.053)	0.383 (0.066)	0.426 (0.074)
MSE	Dante	3.164	2.441	1.921
	DBM	3.390	2.674	2.244

Dante and DBM average forecast skill and mean squared error (MSE) comparisons across geographic scales. For skill results both the FluSight challenge’s forecast skill and the exponentiated, proper log score are reported, with the proper skill in parentheses. Forecast skill and MSE each improve for both models as the geographic scale coarsens. Dante outperformed DBM at all geographic scales under both performance metrics.