Table 3.

Descriptive statistics of model performance: monthly rejection rates.

Dataset	Model performance
	Low (<10%)	Accurate ($\pm 10\%)$	High (>10%)
Fitting period (2012–2015)	9.3%	51.2%	39.5%
Validation period (2016–2017)	9.5%	52.4%	38.1%
Total period (2012–2017)	9.4%	51.6%	39.0%

Deliveries computed from the hourly model within each month classified as Low, Accurate, or High if the model predicted monthly rejection rate is more than 10% below, within 10% of or more than 10% above, respectively, relative to the historical observed monthly rejection rate. The table summarises the proportion of months with each classification for the respective dataset.