What constitutes the long run?

[dameblackman]

I've seen, depending on STDev's and distributions, people say anything from 200 to 600 plays, to 5 years to 20 years. Can anyone give me a comprehensive analysis and arguments for thinking about this a given way?

Obv. depends greatly on what "avg." stdev is across the board for ppl with inarguably long (maybe not statistically foolproof but simply large relative to the majority of other bettors) sample sizes of plays, and how your stdev. relates...Nar (forgot the rest of your handle, sorry, your avatar is some silly model if that helps), I understand your argument that obv. 73 games is a worthless sample size (and agree in chief principle with your coin-flipping analogy). However, you stated "the long run is a lot longer than we think"...meaning what, I should never try to extrapolate long-term forecasts from my samples?

Thanks in advance for thoughts.

[Jaguar101]

i say anything over 5 miles is a long run

[Tmoney120]

Take it from a marathoner. Long runs every Sunday - 12-20 miles is standard.

[rjp]

At least 2,000, but this isn't a number set in stone. The more the better, obviously.

[fun160]

[ QUOTE ]
At least 2,000, but this isn't a number set in stone. The more the better, obviously.

[/ QUOTE ]

2,000 miles or 2,000 years?

[rjp]

[img]/images/graemlins/smile.gif[/img]

[Tmoney120]

I would guess 2000 plays. It takes me a year to log 2000 miles.