On last week’s episode of Survivor, it was pointed out that ever since they introduced a brand-new car as a reward, no one who won had won Survivor. This was referred to the “curse” of the car.
But does such a curse exist? Is it merely a superstition, or is there something to it? Since I’m writing about it in this column, you’ve probably guessed that I’m going attempt to take a statistical look at it.
Firstly, here are the ten car winners so far:
| Season |
Car Winner |
Number of participants
in car challenge |
| Australia |
Colby |
4 |
| Africa |
Lex |
5 |
| Marquesas |
Sean |
5 |
| Thailand |
Ted |
5 |
| Amazon |
Matthew |
5 |
| Pearl Islands |
Burton |
5 |
| All-Stars |
Rob M |
5 |
| Vanuatu |
Eliza |
6 |
| Palau |
Ian |
5 |
| Guatemala |
Cindy |
5 |
As can be seen from the above table, the car challenge usually occurs when there are five survivors left. This means that the car winner has approximately a one-in-five chance of actually winning Survivor (it’s approximately because of the two seasons where the challenge had a different number of survivors).
For ease of analysis, I’m going to assume that the car winner has exactly a one-in-five chance of winning. Of course, it’s not true that everybody has an equal chance of winning, as there are many additional factors which influence a survivor’s chances of winning. However, these can’t be quantified, and for the purposes of this article I’m going to assume that the chances of winning only depend on the number of people left.
If the car winner has a one-in-five chance of winning, then that means out of ten car winners, you’d expect two to have won the million dollars. Clearly this hasn’t happened, but does the fact that none of them have won mean that something sinister is going on?
Well, no it doesn’t. Consider this example: if you were to toss a coin ten times and it came up tails seven times, does it mean that the coin isn’t evenly balanced?
I think most people would agree that based on those results (seven out of ten), you can’t say for certain. In order to determine whether there is any bias, you would have to toss the coin many more times.
It’s a similar situation with the car winners. You may expect two to have won Survivor, but the fact that none of them have won is not an indication that your expectation isn’t wrong. It’s another case of not having enough data to say for certain.
In fact, in a situation where you’d expect two out of ten to win, the result that none of the ten would win occurs approximately 10% of the time. So while none of the car winners going on to be the overall winner is not the most common outcome, it’s far from being a statistical anomaly.
But despite this, there may be some validity in the idea of a “car curse”. If you look at the car winners, you see they many of them were the most physically dominant of the survivors left at that point.
This means that they posed a physical threat, which resulted in them being voted out or, in the case of Colby and Rob, played a role in the weaker of the final two winning. So ultimately, it’s not a curse on the car winner but rather a curse on someone who poses a threat by being too dominant at challenges.