I recently read an article about a study that some professor did, but unfortunately I don’t remember his name. The basic idea is the following:
About 1500 people are split in 2 groups, group A (we will call it so) has very simple task – listen to a bunch of indie songs and rate them accordingly. The more curious and exciting group we will call B, which has a similarly simple task – listen to a bunch of songs and rate them, the one difference is that B has access to B’s ratings and people see the top list of best rated songs before they choose what to listen and how to rate.
The obvious result is that ratings among A of songs are not much different from each other. That means that rating of the first song is just a little bit higher than the second top rated song, which is just a little bit higher than the third rated song and etc. Group B’s rating however looked different. Song number 1 was far far better rated than song number two and song number three was significantly worse than two etc. From this we can conclude that the exponential results are due to what Alek calls “herd behavior”.
So far, so good.
Now comes the real deal. Group B was divided in 8 sections. All of them listened to the same songs, but each group had their own individual rating lists. The result was very very curious – each had a different top 15 list. In each section, different songs became the most listened. The article reasoned that small differences of luck, or unaccounted factors in the beginning can mean a lot for the success of a song. This way a song that got lucky to be played in the beginning becomes big hit for example, although there are plenty of other songs that technically are not worse in musical terms. If you try to give an analogue of real life, the star Madonna is now, could have been someone completely else due to random small occurrences of luck.
Now that sounds horrible, unfair, and nasty for the people that do believe they have qualities but didn’t make it only because of this luck part of the deal.
I think we can put a different spin on failure however. If there is about 10 songs, and you think you have the skills to make in in top 3, but you were unlucky up to know, I can tell you one thing: doing it over and over again will statistically force the bad luck to ‘run out’ and you will marry success. Same for startups.
I have never thought of it like that, but this is a big part of what people call perseverance.
UPDATE: [ Here is what Marc Andreessen wrote about luck ]
You can apply the same concept to businesses (think of each busienss as a song). If you have two site, X and Y, with similar features, launching at the same time, deciding which one will get a bigger market share is is really a matter of luck, more than quality or anything else.