Wisdom of the Crowd

We have posted this picture last week and we let people weigh-in on the number of Mentos candy inside the Jar, you're probably wondering if this is a sponsored ad by the Mentos people. Well it's not. The idea behind this experiment is to test the wisdom of the crowd and if it's reliable in terms of accuracy.

The correct number of candy in the Jar was 77. We had a total of 92 guesses across our social media channels [Facebook, LinkedIn & Instagram].

Below is a distribution of all the guesses we've received

The wisdom of the crowd entails that the collective opinion of a group of individuals will far more accurate than that of a single individual. Experiments that have been carried out over the years such as the famous Ox experiment case by Sir Francis Galton and BBC's prof. Marcus du Sautoy, have reaffirmed this model.

If we calculate the average of the data we've received referring, we'll find that the crowd guessed an average of 62 (~19% error) with around 81.5% of the guessing below the correct answer. The average shows a pretty good indication of the amount of candy in the jar, seeing that participants had nothing but a single perspective of the object to use as reference.

Experiments have shown that the larger the participants the more it converges towards the correct answer as long as the error of the participant is statically independent of the error of other individuals - which was not guaranteed in this context. Under the right conditions, the noise generated with each individual judgement is reduced upon taking the average of a large enough group.

So, is the wisdom of the crowd a reliable method? Well as we have mentioned above, it's been proven to be relatively accurate when trying to find a discrete solution (under the correct conditions) as for more complex problems additional measures must be taken to assure satisfactory results.

Thank you for participating in this little experiment of ours! We have many more to come!