How long is "the long run"?

One often hears it said that poker is a skill game because, over the long run, the luck evens out. That's true, at least in a theoretical sense. But just how long does "the long run" have to be for the luck to even out?

I recently read (though I can't remember where, and I'm feeling too lazy to search for it) an interview with Gus Hansen in which he estimated that perhaps 90% of his success in a given session could be attributed to luck, but over the course of a year, that number would be more like 2%. That sounds nice, and it roughly matches my personal, subjective sense of what's going on, but it's not very scientific.

I'm reading Poker for Dummies by Richard Harroch and Lou Krieger. It is actually a fine introduction to the game. Their chapters on stud and Omaha are the first pieces of strategic advice I've read on those games, and they have already helped me with both starting hand selection and deciding when to jam versus keep the pot small in the low-stakes HORSE tournaments I've been playing every day lately.

But of even more interest, and the part that is of relevance to this post, was this short snippet, from pp. 35-36:

We used a computer to simulate 60,000 hands of $20-$40 Hold'em. That's
about one year of play if you treated poker as a job and went at it eight hours
a day. The objective was to determine how long it would take to get into "the
long run," that elusive zone where luck is filtered out and only skill
determines who wins and loses.

Because identical player profiles were loaded into the computer, the
long-run expectation was zero. With identical profiles, each player should
neither win nor lose. They should have broken even in the long run.

Nevertheless, there were four losers and five winners. Seat 9 lost $3.18
per hour while seat 6 won at the rate of $1.99. That's a difference of more than
$5 per hour--and it was clear they never got into the long run, even after a
year of simulated play.

The authors then discuss extending the simulation to 50 years of play, and still differences remained, though smaller: up $0.60 per hour for the big winner and down $0.35 per hour for the big loser. They conclude from this experiment: "Maybe the best you can expect over a lifetime of poker is that only 1 to 1.5 percent of your results would be attributable to luck."

This reminded me of a startling assertion that Mike Caro made in one of his Bluff magazine columns last year:

Here’s where people get confused. They think that because luck evens out in
the long run, and skill prevails, that over their career of playing poker,
they’ll get almost exactly the same opportunities as everyone else. This just
isn’t so. A lifetime isn’t long enough for the cards to break
even.

Sure, if you play 10 hours a day for 50 years, you’ll get approximately the
same proportion of top pairs, flushes, full houses, straight flushes, and
everything else that others get. Some folks call it the law of averages or the
law of very large numbers. But it doesn’t tell the whole story. You need to
consider which games you’re playing when you hold that royal flush and how many
players are throwing big money against it. Are you competing for big limits at
the right times? Were you on vacation when the billionaire came to town and
dumped millions into your game?

And what about tournament winners? A top pro can go years without winning a
tournament and another with similar skill might win four times in one year. That’s luck, and – trust me – it won’t even out in 25 years of play. Skill matters a lot, but not enough to definitively determine who’s best – which is another reason I seldom play poker tournaments. I already know I’m best and if I simply declare it, some folks will believe me. If I play hundreds of tournaments and don’t win, they’ll begin to wonder. See?

Life isn’t fair. Some people spend a lot of time in hospitals. Some
businesses fail for unforeseen reasons. Your life equates to a single session of
poker. Luck won’t even out for you. But the more you steadfastly make
good decisions, the better you’re likely to do with the cards you’re dealt.

That's from the March, 2007, issue, posted here, with emphasis added.

So if you're playing several times a week, not doing anything different than your usual winning ways, yet experience a couple of months in which you lose, lose, and lose some more, can it really be due entirely or largely to a nasty streak of bad luck lasting that long?

Oh, yeah.