My favorite reality show Top Chef is back and this time it’s in Vegas. Sin City is not known for any special cuisine so the producers have opted to stir in the town’s most famous ingredient – gambling. In each episode one chef is eliminated and we get to watch them fight and flirt till one is left standing many, many weeks from now. So it’s a pretty big deal if you get a pass to the next round. Because of the Vegas setting, one chef won this great prize by drawing a “gold chip”. In the previous shows, you had to cook your way into this advantage in the “quickfire” round. This round still remains but, at least in the first show, there is now the second random route into the next round.
What effect will this have?
As randomization is not a function of the ability of the contestants but winning the quickfire is, the average quality of the chefs in future rounds will go down compared to the previous series. That’s the statistics angle. The game theory angle is the impact on incentives. The good chefs are going to find it easier than before as only the worst chef is eliminated. If a lower quality chef makes it via randomization, the others can slack off in the next round and conserve their energy and best dishes for the future. The cooking will get worse. The producers should drop the randomization trick if they want to see the best possible cooking.
Cheap Talk 
2 comments
Comments feed for this article
August 21, 2009 at 10:27 am
hern
maybe they dont want to see the best cooking possible.
maybe they want to see if a randomization effect leads to unexpected outcomes, leading to a more emotionally drawn audience, leading to an effect on ratings.
lets say the audience represents a spectrum of cooking interest, from remotely interested to die hard enthusiasts, with the bulk being somewhere in between. the randomization aspect would probably turn away the enthusiasts, and maybe draw in a greater number of the remotely interested. the question is then, will the number of enthusiasts they turn away be greater than the newly interested? id guess the probability ratings drop so much they risk cancellation is extremely low. if ratings increase, then maybe future seasons will introudce a little randomization. if ratings fall, they’ll see it as a lesson learned and probably never do it again.
or maybe im just thinking way too many steps ahead.
August 27, 2009 at 3:12 pm
allan
Does this mean that the best chefs would actually prefer randomization? Yet the worst chefs would also, I assume, prefer randomization. Do any players lose out by switching from a meritocratic to a random model?
How do you model the preferences of the middle (average) players for randomization vs. meritocracy? I imagine this depends on how you model the selection function in the “quickfire” round. We assume that it is a function of talent, but how likely is a middling chef to win? If talent is randomly distributed
Moreover, how generalizable is this? Should high talent people want a less meritocratic system that randomly rewards people at earlier levels of the hierarchy, so that competition for the top jobs is less risky?