One of my friends is a member of a card counting team that plays blackjack in the western US: Las Vegas, Reno, Tahoe, Indian casinos, etc. He says team play is advantageous because in blackjack the edge is so small and the variance is so high, that a single individual playing alone would need more than a year's worth of play just to find out whether or not he had an edge. Anything less than 50,000 hands and you just don't have enough data to really know; the noise is bigger than the signal. (50 weeks/year * 5 days/week * 4 hours/day * 50 hands/hour = 50k hands). On the other hand a team with N members finds out whether collectively they have an edge, in (1/N) years. Summary: blackjack card counters evaluate themselves by looking at their profits after 50,000 hands of play. Maybe there is a parallel in trading. Maybe traders can evaluate themselves by looking at their profit and loss after 10,000 trades. That's enough trades to give a high confidence level that you are measuring signal (trading ability) rather than noise (random good- or bad-luck).