Hacker News Reader

Comment by liftbigweights | Hacker News Reader

Yes, the median would be a better data point. Even better would be a qualification of "successful".

Are they equating successful startups that are worth a million to trillion dollar companies like apple and amazon?

> I always struggle when thinking about average vs median.

Median takes away the bias of the extremes that averages can include.

For example, if 9 people make $1 and 1 person makes $1000000, the average is about $100,000 while the media is $1. The $1 is more representative of the sample.

But in this "research", there are far more data to unpack. Like the quality and quantity of "successful".

View on HN

> Even better would be a qualification of "successful".

Did you... read the article?

Fourth paragraph:

"Among the top 0.1% of startups based on growth in their first five years, we find that the founders started their companies, on average, when they were 45 years old. These highest-performing firms were identified based on employment growth. The age finding is similar using firms with the fastest sales growth instead, and founder age is similarly high for those startups that successfully exit through an IPO or acquisition. In other words, when you look at most successful firms, the average founder age goes up, not down. Overall, the empirical evidence shows that successful entrepreneurs tend to be middle-aged, not young."

You can _disagree_ with their definition of success, but they certainly have one.

>Median takes away the bias of the extremes that averages can include.

What sort of extremes do you expect to see in age data for starting a business?

There will be no 10 years olds, and likely no 90 years olds.

>But in this "research", there are far more data to unpack.

Why the quotations? Do you not believe this is research? This is a 1000 word summary of their work; do you think these researchers, from some of the best schools on the planet, didn't consider the sorts of questions you pose?

... and the mode is 1.

Measures of central tendency worth keeping in mind when talking "averages":

mean - what people usually mean by "average"

median - central point on a number line (ie, the same number of items are greater, and lower, than the median)

mode - the number which is most common in the set

Reply on news.ycombinator.com