Table 1.

Performance comparison.

Track	Our model	Stanford model (Ho et al., 2017)	Human
Farm	98.33	97.46	94.07
Raceway	166	129.09, 1 DNF*	125.30
Mountain	213	138.37, 2 DNF*	129.50

For each track, we run 10 races in real time and calculate the mean race times as the final result. The Stanford results were borrowed from the Stanford report. Human results are obtained by real human participants playing each track twice: The first time is to get used to the track, and the second time is to record the time they finish the game. DNF

^*signifies that the autopilot got stuck and was unable to finish some number of races.