2014年3月15日星期六

The Gambler's Fallacy

You've probably never heard of the gambler's fallacy. If not, read on because if you're not familiar with it, you've likely fallen into its trap and it's costing you money.
The gambler's fallacy is a condition that besets nearly everyone at various times in their lives. However, as befits the name, it is famously frequent in gamblers and it is, of course, a fallacy.
It was discovered by psychologists and has been a topic of study for decades, and the "ol' perfesser" here is going to give a lecture on it. Get out your notebooks.
Gamblers are a superstitious lot - poker cheat poker players less so than slot junkies, but we've got our nutty beliefs. One of them is that "we're due" to fill a flush, that it's "our turn" to suck out on the river or that it's "about time" for us to rack up a 200BB win.
We think this because we haven't filled a flush the last 20 draws, or hit a miracle river card in the last 40 hours of play or had a multiple rack'em up session in months.
It is this belief that is the gambler's fallacy. And the reason it's a fallacy is that, no matter what you think, you're not due for anything, ever!
There is no increase in the likelihood that you will fill your flush just because you've missed the past 20. Having gone a hundred hands without hitting a two-outer has no bearing on the next river you pay to see.
Layne Flack
Thought I was due there.
I've raised this topic with lots of players, some quite skilled. Almost invariably they tell me that they are due, that they have to fill that flush - eventually. When I push them, they typically give me some justification, a rationale that looks, for all the world, like it makes sense.
"Look," they say, "I'm not stupid. I know that the probability of hitting my flush hasn't really changed but I also know that things have to work out in the long run. It's getting to be a 'long run' because I haven't hit one in, like, forever.
"So it has to come along because things have to even out. In fact, that's what you math guys tell me. All this luck crap evens out in the long run."
If this feels familiar, if you've said (or thought) something like this, you're probably sitting there nodding your head. After all, the belief that the flush card is bloody, freakin' due is compelling, absolutely haunting.
But is it true? Nope. It's false. The awful truth is that things do not have to even out. The cards are under no moral stricture to treat you fairly.
And the long run is really marked cards, really long. Most people, poker players included, have trouble grasping this compelling mathematical truth.
As a result, they are likely to increase bet size or make marginal calls because they are convinced they are due. The real problem, of course, isn't the belief; it's acting on it that puts dents in your bankroll.
Playing cards are made of plastic. They do not have memories. They do not "know" that they just failed to fill your flush, again. They are being shuffled by a dealer or a machine and the order in which they emerged in the past has no bearing on their order in the future.
Each hand is independent of all preceding hands. The probability of filling a flush is not changed by previous outcomes.
Tuan Le
Seriously. Thought I was due.
There are also, obviously, situations in poker where events are dependent on each other. If you saw someone flash a heart when they mucked their hand and you're on a flush draw, you know there are only eight hearts left in the deck.
Hence, the probability of hitting your draw is changed by previous events and the way you want to play the hand just changed subtly. Oddly, virtually everyone understands this second "dependent" case, but many, perhaps most, just can't wrap their brains around the first "independent" one.
And you don't have to be gambling or at a poker table to see this. Here's a study done some years back.
People were asked to guess which of two lights on a computer screen would come on. They were told that the sequence was random, that the light on the right was as likely to come on as the left.
Virtually every person playing this "guessing" game showed the gambler's fallacy. The more often one light came on, the more likely they were to predict the other side.
If there was a really long run, like seven lefts in a row, they picked the right side on the next trial nearly 100% of the time! When asked why, they said that the other one was "due."
Yes, things do tend to even out in the long run. But note the two qualifiers in that sentence, tend and long run.
"Tend" means just that, tend. It doesn't mean "must."
"Long run" implies an infinitely long run for mathematical certainty and, frankly, you don't have that kind of time. You should not expect, in any relatively short run of events, to see the outcomes conform to theoretical expectations.
Daniel Negreanu
Gambler's fallacy is most seductive, but it is a fallacy.
They are under no obligation to do so. In fact, they might not even be close.
The gambler's fallacy is most seductive, but it is a fallacy. Don't get suckered into thinking that you're "due" for anything other than random outcomes. And don't throw money into a pot based on this belief.
The probabilities have not changed and they don't give a fat flying fig how many times you missed your flush or how many times you've had to buy in already.
OK, if your brain doesn't hurt now, tune in again next week because I'm going to introduce another topic that you also need to understand, and it's even more counterintuitive.
It's the principle known in probability theory as "regression to the mean" and every good poker player understands it - although maybe not explicitly.

2014年3月12日星期三

£10,000 No-Limit Hold'em Main Event - Live Updates

10/01/2009 (4 years ago)

Shulman is Champion!

Barry Shulman
Don'e spend it all at once!
The 2009 WSOPE presented by Betfair is over and Barry Shulman is champion! It was a heartbreaker for Daniel Negreanu but the silver lining is that with this cash he becomes poker's all time leading tournament money winner.
It was a marathon heads-up match that ended in a crippling blow followed by the coup de grace.
The Crippler
On a J 8 5 flop Daniel Negreanu checked, Shulman bet 300k and Negreanu check-raised to 900,000.
Shulman shoved and after a few minutes in the tank, Kid Poker emerged with the call. It was aces for Shulman and Negreanu needing help holding Q J.
As the dealer burned and turned the crowd at the Empire roared in appreciation as the miracle J hit the felt and it appeared the title was Negreanu’s.
Appearances can be decieving, however, marked cards and the river was Shulman’s turn for a miracle, as the A came down.
Shocking stuff!
That hand left Negreanu with a massive deficit - down to 1.2 million versus Shulman's imposing 8 million stack - he was in big trouble.
The Death Blow
When Negreanu found pocket fours with just ten big blinds, there was no escape.
Heads-up
Mano a mano.
Bad luck for Negreanu - Shulman had him trussed up with pocket tens.
Needing to spike to win - instead it was Shulman who made a set, a ten hitting the flop to leave Negreanu virtually dead.
Running fours or running diamonds were his only salvation and when the turn was neither of these - it was all over!
Massive applause for both these gladiators who have slugged it out over a marathon heads-up game. Negreanu refused to give in - but ultimately Shulman made it to the winner's circle alone.
Second place for Daniel Negreanu, he picks up £495,589 for his performance - which happens to propel him above Jamie Gold at the top of the tournament prizes leaderboard.
Meanwhile congratulations to Barry Shulman who collects a huge £801,603 for his performance as well as his second gold bracelet!
Level
29
Blinds
50000/100000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

Kid Poker Makes His Move

Daniel Negreanu
I'm comin' for ya Barry!
With the blinds a whopping 60k/120k, winning a few pots in a row really changes the status quo.
Negreanu just did that, taking three pots in a row, meaning he was close to Shulman's chip lead for the first time in a while.
The final hand in that sequence saw Negreanu re-raise an opening min-raise by Shulman to 650k, Shulman making the call.
The flop fell K T 6 and Negreanu led out for 850k. Shulman passed and Negreanu showed aces for the 1000th time since they started.
Ok, a slight exaggeration but Negreanu has been seeing those bullets more often than you'd think.
And then, suddenly, Negreanu leapt into the chip lead for the first time since he lost that huge pot with aces -again!
The hand saw Negreanu push Shulman off a huge pot.
On a ten-high flop with two hearts Negreanu bet 240k, Shulman raised to 940k and Negreanu shoved!
Shulman tanked for a moment before electing to lay it down, meaning Negreanu now holds a slim lead by just 100k - or less than one big blind.
It couldn't be closer!
Level
29
Blinds
50000/100000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

I Want That Bracelet So Bad!

Daniel Negreanu
Cut me some slack poker gods.
Such a frustrating game for Negreanu - every time he looks to be gaining parity through a series of small pots - he loses a big one and Shulman regains chip monster status.
That's exactly how the previous passage of play has panned out, Shulman being forced to fold or being shown the best hand by Negreanu.
Shulman loves his check-raises and as the pair went to a queen-high flop, Negreanu led and again Shulman popped it to 600k, Negreanu making the call.
The pair cautiously checked down the remaining streets and Shulman showed down a queen to take the pot down, sending a super healthy 1.5 million pot the older combatants' way.
Negreanu back to a 3-1 disadvantage. He refuses to give up though...
Level
29
Blinds
50000/100000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

Angry Negreanu!

Daniel Negreanu
Don't make me angry!
Kid Poker is not a happy man right now!
He's become increasingly agitated and annoyed at the way the cards have been falling.
One of the more recent hands saw Shulman make it 200k from the button - a min raise - only for Negreanu to re-raise to 600k.
Shulman made the call.
The players checked down a board of Q 5 4 2 6 and Shulman turned over 9 6 for a rivered pair, causing Kid Poker marked cards contactlenses to leap in the air and slam down his cards, shaking his head at how unfair the poker gods are sometimes.
"Come on Negreanu!" yelled someone from the crowd in a bid to re-invigorate the Canadian who is so desperate to get his hands on that bracelet.
An unmoved Barry Shulman simply added the chips to his stack - still sitting on a formidable lead.
Level
29
Blinds
50000/100000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

Clash of the Titans

Heads-Up
The battle for the bracelet continues.
It's a titanic struggle between the two players right now. Negreanu had started to make some progress against Shulman, chipping back up to 3.5 million, before a series of check-raises when Negreanu had raised the button, saw an increasingly frustrated looking Negreanu begin to concede ground again.

He shipped a pot right at the death when his rivered pair of sixes took down Shulman's pocket deuces so right now here's how they stand as they take a twenty-minute break.
Shulman - 6.5 million
Negreanu - 3.5 million
The kid won't go down without a fight, but Shulman shows no signs of being intimidated by Negreanu and he won't be able to simply bulldoze his way back into contention.
They'll be resuming any moment now, so we're off to re-join the action.
Level
27
Blinds
30000/60000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

Shulman Shows Heart!

Barry Shulman
Sorry Daniel, I'm good at hitting those draws!
Suddenly the two players tangled in an enormous pot!
Negreanu raised up the button and Shulman re-raised - Kid Poker making the call.
The board peeled off K 8 6 - Shulman led out and Negreanu moved all-in. Call from Shulman!
Shulman showed down a heart flush draw with A 5, whilst Daniel had him on the backfoot with a pair of black aces. Just needing to fade a heart to pick up the bracelet, Negreanu studied the flop intently.
Shulman has been all-in for his tournament life with the worst hand before though and he's made a habit of hitting the card to stay in and chip up.
And guess what? He did it again!
The 2 turned to a mighty roar from the watching rail as Shulman made his flush, leaving a disonsolate Negreanu drawing dead.
Looking very upset, the Canadian removed some of his towers of 25k chips and handed them over to Shulman, who now enjoys a big 3-1 chip lead for the first time since the contest began - his 7.5 million dwarfing Negreanu's 2.5 million.
Can Kid Poker recover from this cruel set back?
Level
27
Blinds
30000/60000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

Shulman Fights Back!

Barry Shulman
Game on!
The posturing is over and now the two prize-fighters are starting to throw some big shots.
A poised Negreanu had put a stuttering start behind him, soaring into the ascendency with oily precision as he began to systematically slip through the many gears his years of poker experience have crafted.
Negreanu had begun to pick up chunky pot after chunky pot as he made great value bets, got Shulman to fold in some key situations and chipped away till Shulman eventually found himself with under 3 million for the first time.
Then a big pot saw the pendulum swing.
The two combatants saw an 8 6 3 board peel off and Negreanu called a bet from Shulman, who went on to check the 6 turn card.
Negreanu bet out 500k, and now suddenly Shulman sprung what appeared to be a trap, counting out his stack and announcing all-in.
Kid Poker wasn't loving it, and tipped the rim of his baseball cap with a scowl as he considered whether to call.
Eventually he laid it down, and all the inroads he had made into Shulman's stack were lost, the older player fighting back as the Canadian threatened to overwhelm him.
The stacks? As they were at the outset of the contest - Negreanu 6.1 mill, Shuman 3.8 mill.
Game on!
Level
27
Blinds
30000/60000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

Negreanu Slipping a Little

Daniel Negreanu
I'll catch ya Barry, just wait.
The players are seated across the felt from each other and a small mountain of £20 bricks lies between them, the bracelet glimmering luxuriantly at the apex of its papery nest.
With the prize and riches lying there as a reminder of just what these two players are playing for, the opening stages of the heads-up have been carefully played by both players.
Negreanu is obviously the favorite as they stand, his experience and skill, not to mention his chip lead all edges which Shulman has to overcome.
So far though, he is more than holding his own.
In fact he has managed to eat slightly into Negreanu's lead. He's still ahead of Shulman but he has to turn the pattern round or he may risk losing out on that fifth bracelet.
Can Shulman take down his more acclaimed opponent?
Level
27
Blinds
30000/60000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

Bansi Ban-zai'd! (3rd)

Daniel Negreanu
I gotcha Praz!
Praz Bansi really needed to get back into this, so made a desperation ship of Q 2, only for Negreanu to find one of the worst hands for him to face - ace-queen.
Praz realised he was staring down the barrell and head hung he walked over and shook Negreanu's hand as Kid Poker prepared to deliver the killer blow.
Dominated, the crowd realised the English star would have to hit a deuce or hearts to stay in contention and loud, slightly inebriated shouts of "deuce, deuce, deuce" rang out through the casino.
As the rail watched on, their breath held, an inconsequential series of baby cards were peeled off, but unfortunately for Bansi, none of them were the deuce he needed, meaning Negreanu takes another victim, busting Praz Bansi.
Bansi wins £360,887 for his third place finish and receives a standing ovation from his multitude of supporters on the rail.
But now it's mano-a-mano for the title, Negreanu vs Shulman - Kid Poker holding a sizeable advantage.
Who will take it down?
Level
27
Blinds
30000/60000
Ante
10000
Average Stack
$5,010,000
Players Left
2
Tables Left
1
10/01/2009 (4 years ago)

Schulman Strikes Again

Praz Bansi
That guy's running so good!
Praz raised up the button with K-8 and Barry Schulman made the call in the big blind.
The flop came king-high, Barry checked and Praz led out, only for Schulman to check-raise all-in. Praz made the call and once again Shulma's tournament life was on the line.
This time, he was in good shape though, his king-jack connecting harder with the flop than Bansi's K-8.
Needing to hit an eight, Bansi watched the turn and river blank.
Huge "ahhhhhh" of sympathy from his supporters and he was left short with just 700k to his name, whilst Shulman has moved to over 4 million.
Level
27
Blinds
30000/60000
Ante
10000
Average Stack
$3,340,000
Players Left
3
Tables Left
1















2014年3月2日星期日

Stop Acting Like a Birdbrain

I've been playing poker for longer than I've been a research psychologist. Many of my favorite moments at the table are ones that connect my two worlds.
Near the top on this list are those times where someone does something "superstitious."
They ask for a new setup after taking a really bad beat. Call an early-position raise with Q 9 because it's their "lucky" hand. Mutter something that sounds an awful lot like a prayer at a critical juncture (Jerry Yang could be heard intoning such during every all-in moment at the final table of last year's WSOP).
I'm calling these gestures "superstitions" because they are. According to a noted authority, a superstition is "a belief held in the absence of what one not holding that belief would consider to be adequate evidence to substantiate or support it." (Dictionary of Psychology, Penguin Books - and yes, I wrote it, go buy a copy marked cards lenses; I can use the royalties.)
The key is belief. If you believe a new deck will change your fortune, that Q 9 is really lucky for you, or that the chant will provide you with a much-needed suck-out, you're free to do so.
Daniel Negreanu
If you truly believe Q-9 is really lucky, Daniel Negreanu would head the list of people waiting to play you.
And if I, after logging all the data and applying what I know about the world around us, am not convinced of the legitimacy of the proposition, then it is a superstition.
You may be right. I may be wrong. We may both think we know; we don't always.
I'll bet you a rack of reds that many of you reading this think that the new setup request is silly, that Q 9 is not the best hand to call a raise with, but nonetheless suspect that Jerry Yang may have been on to something.
There is a deservedly famous experiment in psychology that tells us a lot about this. It was run with pigeons some years back by the (in)famous behavioral psychologist, B. F. Skinner.
It went like this - if you're thinking of bailing out on me now by clicking on the closest link, don't. This really is a column about poker.
Skinner took a group of his pigeons and put them in small boxes each equipped with a light, a food hopper and not much else.
Pigeons aren't the smartest of birds, and it's pretty tough to bore one. Skinner had spent years studying how we (pigeons, people, rats) respond to reinforcement. If a pigeon receives food following pecks at an illuminated disk, it quickly learns to peck at disks.
Mark Kroon
It's so weird. Every time I make this face, everyone folds.
No problem, you say. Pigeons can recognize the relationships between their actions and the arrival of food. Indeed, one can imagine a pigeon chirping away with one of its avian buds and saying something like, "You know, man, I've really got this guy conditioned. Every time I peck this key, he gives me food."
From the pigeon's perspective this seems right - though Skinner might demur. We can imagine him chatting with a colleague and saying something like, "You know, man, I've really got this bird conditioned. Every time he hits that disk, I give him food."
There's a bit of a clash here between interpretations. From the pigeon's point of view, the matter couldn't be clearer. He's pretty sure he's in charge and everything that's happened so far supports that view.
Skinner, on the other hand, also appears to be bankable. He's also pretty sure that he's in charge and firmly believes that he's running this show.
Anton Allemann
That seat looks so good to me.
Indeed, both have a virtual lock on reality here. Frankly, I'm not sure that these intellectually challenged bird brains can grasp concepts like "virtual lock," not to mention "reality," but that doesn't matter.
The reason for this epistemic twist is that, without further elaboration, each of them has a world view that accounts for all the data. You may know (or at least be damn sure) that Skinner's point marked cards of view is right but that, for the point I'm making here, is irrelevant.
These experiments just set the stage for the really interesting one. One day Skinner decided to program the food hoppers to deliver a couple of grains of food from time to time no matter what the pigeons did.
There was no requirement to peck at a disk or, in fact, to do anything. They could just sit on their feathered haunches if they wished and handfuls of grain would arrive.
Skinner put a half dozen pigeons in their boxes and went home. What do you think he found when he returned?
If you think he had six birds sitting on their haunches in front of the food hopper, you're as wrong as you can be.
Scotty Nguyen
Does whatever it takes to keep the Corona hopper filled.
In one box a pigeon was balancing on one leg and tilting sideways. In another the bird was walking around the cage holding its right wing extended. In the next the pigeon was bobbing its head up and down. In the next ... well, you get the picture.
What was going on? Let's take the pigeon perspective. Skinner closes the lids and leaves. After a bit, they start doing pigeon things, which include stretching their wings, standing on one leg, bobbing their heads, etc.
Now, suddenly, the food hopper fills up. "Hmm," thinks the bird, "I lifted my wing and food showed up. Think I'll try that again." And, indeed, so long as the bird keeps its wing up food does appear.
In his neighbor's box, of course, all this happened while that bird was bobbing its head - leading this beast of little brain to conclude, "Hmm, I've got this guy conditioned again. Now, if I keep bobbing and weaving he sends food down into the hopper."
The next time you even think about calling for a new setup, switching seats because the one you're sitting in doesn't seem to be attracting cards, requesting a table change because you've taken a couple of bad beats or replacing solid poker decision making with chanting and praying, I want you to take five, remember Skinner's experiments and stop acting like a birdbrain.