That’s not to say that machine learning can’t play a role.
In this season, minor prizes are awarded after each challenge to the team members that were not vulnerable to elimination in the following Gauntlet; i.e.
on challenges before male Gauntlets, prizes are awarded to female members of the winning team, and vice versa.
The grand prize is $300,000, which is split among the remaining members of the team that wins the final challenge.
In the Gauntlet, the game played is determined by spinning a wheel.
Each team has to unscrew an oversize wing nut until it has been removed off a track, which contains the key to the next checkpoint, which is based on the "Sliders" elimination game.
Note: Prior to this checkpoint, Eric collapsed and needed medical assistance, and the Veterans team became disqualified when he was removed from the final challenge.
This, of course, does not mean that it can’t vary, but we do mostly understand what can cause these variations.
There are internal/natural cycles that can produce variations, but there are limits as to how large these internally-driven cycles can be and how long they can last.
The residual is then used to estimate the ECS, which they suggest is around 0.6C. It’s essentially just a complicated curve-fitting exercise.