3 ms·
"...compensation package has a higher expected value..." Expected value is a good measure when you're summing over lots of instances, e.g. if you're a VC fund
by martincmartin 9y ago
"...compensation package has a higher expected value..."
Expected value is a good measure when you're summing over lots of instances, e.g. if you're a VC fund investing in lots of startups.
As an employee, where you're working for a single startup at a time, robust statistics[1] suggests that the median is a better measure of what you'll expect to make: you have a 50/50 chance of making more/less than the median.
More than half of startups either fail, or don't succeed wildly enough for options to be worth more than the equivalent salary.
(If you work for 5 startups in your career, the best measure might be "sample 5 startups and sum the options payout to produce a value; repeat that many times and take the median of the result." But that's a lot harder to intuit, and is no doubt closer to the median than the expected value.)
[1] https://en.wikipedia.org/wiki/Robust_statistics https://en.wikipedia.org/wiki/Robust_statistics
- jdavis703 9y agoWhy take the median? For me personally all I need is one year where I make a couple million bucks. What I really care about for my personal financial position is either the sum or mean, because that's what hits my bank account.
- martincmartin 9y agoBecause you'll only work at a handful of startups in your lifetime. You will not make the sum / mean over all startups, only the ones you work at.
- thedufer 9y agoSome lotteries that accumulate when there are no winners have weeks with better-than-even odds. By this logic, you should wait until the odds are sufficiently in your favor and then dump your entire savings account into lotto tickets, for an immediate gain. Of course, even if you can pick up a 20% gain in EV, that small chance you end up a billionaire won't save you from the 99.99..% of cases when you're eating cat food in retirement.