I am not sure it is a good idea to pick your employer like an investor would.
An investor in startups is going for high risk, high reward investments. For every individual investment they make, the chance of any sort of return is quite low; they make up for this because the rare startups that succeed make enough to compensate for this low success rate. Since they are making a lot of these high risk/high reward 'bets', they can be fairly assured they will receive the expected value in return.
As an individual, you only get to place one bet on your employer. For most startup employees, that means they are very unlikely to get any sort of payout because of company success. It doesn't matter that the 'expected value' of working for a more-likely-to-hit-it-big startup is higher than for another company, since the probability is that you will not hit it big.
As an individual, you are more concerned with things like salary, working conditions, and personal enjoyment of the work. Those are the things you are CERTAINLY going to experience; the tiny chance of huge success is most likely not going to affect you.
I think there is a good analogy to poker here; in Texas Hold'em, two aces are the best starting hand. There is almost no situation where it is best to fold them before the opening flop; your expected value is always positive against ANY other combination of opponent hands.
However, there are a few cases where even though you have the highest expected value for your hand, it is in your best interest to fold. One such case is if you are playing in a tournament, you are in the money, and two people with much larger stacks have already gone all-in. In that case, your best bet is likely to fold; even though you have the highest odds of winning the hand out of everyone, if you LOSE, you are out and can't keep playing. Even if you win, you will still have the smallest stack and are unlikely to move up any spots and win more money; if you fold, though, someone else is likely to be knocked out and you can move up.
It seems counter intuitive, but this holds because expected value only works as your primary decision making point when you can make many bets; if you only get one bet, lower risk with lower reward is often better for you even if the expected value is lower.
Think about it this way; lets say you only have $1,000, and you need it to pay your monthly rent (if you don't pay, you will be kicked out and be homeless). Someone comes up and offers you a bet; you bet $1000, and you get a 1% chance of winning $200,000. In terms of expected value, this is a great deal! If you made the bet many times, you will expect to win $1000 for each bet made! However, 99% of the time you are going to be homeless. For most people, the choice is to go against the expected value and play it safe.