Two Plus Two Poker Forums Coin flips only 40% likely to flip heads after a heads.
 Register FAQ Search Today's Posts Mark Forums Read TwoPlusTwo.com

 Notices

 Science, Math, and Philosophy Discussions regarding science, math, and/or philosophy.

 09-29-2015, 01:50 PM #2 plaaynde Poker Historian     Join Date: Jan 2010 Location: Rebel of Eden Posts: 17,833 Re: Coin flips only 40% likely to flip heads after a heads. Looks like a Monty Hall version. And it's retrospective, not prospective like coming happenings are. Title change suggestion: Coin flips only 40% likely to have been flipped heads after a heads. I see it like they "waste" the times there has been more than one heads immediately before, when your criterion is to look only at the last heads. Last edited by plaaynde; 09-29-2015 at 02:17 PM.
 09-29-2015, 02:06 PM #3 masque de Z Carpal \'Tunnel     Join Date: Aug 2009 Location: Stanford, CA USA Posts: 9,533 Re: Coin flips only 40% likely to flip heads after a heads. In a 4 hand sequence like HHHH do they count the heads after heads appearing 1 time or 3? HHHH HHHT HHTH HHTT HTHH HTHT HTTH HTTT THHH THHT THTH THTT TTHH TTHT TTTH TTTT i see 12 vs 12 here in the 14 qualifying sequences. What are they talking about? Also why 4 and not 16 or 128 sequences? I still do not see how the people need to do the statistics to arrive at 40%? Last edited by masque de Z; 09-29-2015 at 02:14 PM.
09-29-2015, 02:29 PM   #4
Aaron W.
Carpal \'Tunnel

Join Date: Sep 2002
Posts: 29,348
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by masque de Z In a 4 hand sequence like HHHH do they count the heads after heads appearing 1 time or 3? HHHH HHHT HHTH HHTT HTHH HTHT HTTH HTTT THHH THHT THTH THTT TTHH TTHT TTTH TTTT i see 12 vs 12 here in the 14 qualifying sequences. What are they talking about? Also why 4 and not 16 or 128 sequences? I still do not see how the people need to do the statistics to arrive at 40%?
http://andrewgelman.com/2015/07/09/h...is-a-hot-hand/

Code:
```rep <- 1e6
n <- 4
data <- array(sample(c(0,1), rep*n, replace=TRUE), c(rep,n))
prob <- rep(NA, rep)
for (i in 1:rep){
}

> print(mean(prob))
[1] NaN

> print(mean(prob, na.rm=TRUE))
[1] 0.41```

09-29-2015, 02:42 PM   #5
PairTheBoard
Carpal \'Tunnel

Join Date: Dec 2003
Posts: 8,545
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by masque de Z Also why 4 and not 16 or 128 sequences? I still do not see how the people need to do the statistics to arrive at 40%?
You don't. Statistics were done with basketball players to prove that the "Hot Hand" theory was bogus. Those statistics found that a player shooting after making a shot was no more likely to make the next one than usual for him. The idea here is that if they took those statistics the way the coin flips are being done here then those basketball statistics actually proved the "Hot Hand" theory rather than disproved it.

PairTheBoard

09-29-2015, 02:48 PM   #6
PairTheBoard
Carpal \'Tunnel

Join Date: Dec 2003
Posts: 8,545
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by masque de Z Also why 4 and not 16 or 128 sequences?
Can we produce a closed form solution for this probability for the arbitrary case of a sequence of n coin flips? Does the probability go to 50% in the limit? Why is it clear that all the sequences with at least one head in the first n-1 flips are equally likely? Can anyone explain where the bias is coming from?

PairTheBoard

09-29-2015, 02:50 PM   #7
egj
grinder

Join Date: Sep 2004
Posts: 673
Re: Coin flips only 40% likely to flip heads after a heads.

I think the WSJ reporters blew it in their summary of the finding:

Quote:
 The new paper, written by Joshua Miller and Adam Sanjurjo, begins with a riddle. Toss a coin four times. Write down what happened. Repeat that process one million times. What percentage of the flips after heads also came up heads?
The quantity described here is the conditional probability of heads after heads and is just 0.5 as intuition suggests.

Where you get into trouble is if you do the following:
1) Take a million fair coins.
2) Flip each four times
3) For each coin, compute the relative frequency of heads-after-heads in that sequence of four flips.
4) Compute an equally weighted average of those one million values from step 3.

If you do this, you might expect to get something close to 0.5, but you won't.

09-29-2015, 04:12 PM   #8
PairTheBoard
Carpal \'Tunnel

Join Date: Dec 2003
Posts: 8,545
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
Originally Posted by egj
I think the WSJ reporters blew it in their summary of the finding:
Quote:
 The new paper, written by Joshua Miller and Adam Sanjurjo, begins with a riddle. Toss a coin four times. Write down what happened. Repeat that process one million times. What percentage of the flips after heads also came up heads?

The quantity described here is the conditional probability of heads after heads and is just 0.5 as intuition suggests.

Where you get into trouble is if you do the following:
1) Take a million fair coins.
2) Flip each four times
3) For each coin, compute the relative frequency of heads-after-heads in that sequence of four flips.
4) Compute an equally weighted average of those one million values from step 3.

If you do this, you might expect to get something close to 0.5, but you won't.
Right. And the WSJ description was at best vague. However, further down they give the book's description which is more clear. And I think if you put it in the terms I did (see below) you get the counter intuitive flavor of it.

This is the description from the book (see last WSJ paragraph from OP)
----------------
Their breakthrough is the surprising math of coin flips. Take the 14 equally likely sequences of heads and tails with at least one heads in the first three flips—HHHH, HTHH, HTTH, etc. Look at a sequence at random. Select any flip immediately after a heads, and you’ll see the bias: There is a 60% chance it will be tails in the average sequence.
------------------

And my OP description of the probability in question:

"this is saying that if you flip a coin 4 times and IF there is at least one heads in the first 3 flips, THEN

P(randomly picked heads in the first 3 flips is followed by a heads) = 40% "

PairTheBoard

 09-29-2015, 04:31 PM #9 plaaynde Poker Historian     Join Date: Jan 2010 Location: Rebel of Eden Posts: 17,833 Re: Coin flips only 40% likely to flip heads after a heads. Only thing you need to know is it's 0.5 probability for heads when you are about to flip, no matter what's happened before. Question that and you question all of reality. The coin doesn't have a memory. How is it NOT supposed to be 0.5? All the abracadabras can kiss my ass. If and if and if, select this and that from the past, show some athlete concentrating on throwing it in. He is not the one to change the natural laws, doesn't matter how tough and impressive he looks. Know I'm saving the already saved here, but just in case... Last edited by plaaynde; 09-29-2015 at 04:43 PM.
09-29-2015, 05:03 PM   #10
egj
grinder

Join Date: Sep 2004
Posts: 673
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 And my OP description of the probability in question: "this is saying that if you flip a coin 4 times and IF there is at least one heads in the first 3 flips, THEN P(randomly picked heads in the first 3 flips is followed by a heads) = 40% "
No, I don't think this is accurate. Here are the 14 sequences with at least one heads in the first 3 flops:

HHHH
HHHT
HHTH
HTHH
THHH
HHTT
HTHT
HTTH
THHT
THTH
TTHH
HTTT
THTT
TTHT

In those 14 sequences there are 24 heads in positions 1...3. Of those 24 heads, 12 are followed by a heads and 12 are followed by a tails. Exactly as you'd expect.

To get the puzzle, roughly speaking, you need to compute an average for each sequence and then an average of those averages.

09-29-2015, 05:07 PM   #11
PairTheBoard
Carpal \'Tunnel

Join Date: Dec 2003
Posts: 8,545
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by plaaynde Looks like a Monty Hall version. And it's retrospective, not prospective like coming happenings are. Title change suggestion: Coin flips only 40% likely to have been flipped heads after a heads. I see it like they "waste" the times there has been more than one heads immediately before, when your criterion is to look only at the last heads.
Or maybe the "Two Child" (Boy or Girl) paradox
------------------
Mr. Jones has two children. The older child is a girl. What is the probability that both children are girls?
Mr. Smith has two children. At least one of them is a boy. What is the probability that both children are boys?

Gardner initially gave the answers 1/2 and 1/3, respectively; but later acknowledged that the second question was ambiguous.[1] Its answer could be 1/2, depending on how you found out that one child was a boy. The ambiguity, depending on the exact wording and possible assumptions, was confirmed by Bar-Hillel and Falk,[3] and Nickerson.[
-------------------------------

PairTheBoard

09-29-2015, 06:03 PM   #12
Aaron W.
Carpal \'Tunnel

Join Date: Sep 2002
Posts: 29,348
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by egj To get the puzzle, roughly speaking, you need to compute an average for each sequence and then an average of those averages.
HHHH = 3/3 = 6/6
HHHT = 2/3 = 4/6
HHTH = 1/2 = 3/6
HTHH = 1/2 = 3/6
THHH = 2/2 = 6/6
HHTT = 1/2 = 3/6
HTHT = 0/2 = 0/6
HTTH = 0/1 = 0/6
THHT = 1/2 = 3/6
THTH = 0/1 = 0/6
TTHH = 1/1 = 6/6
HTTT = 0/1 = 0/6
THTT = 0/1 = 0/6
TTHT = 0/1 = 0/6

((6+4+3+3+6+3+3+6)/6)/14 = 34/84 = 40.4%

09-29-2015, 07:26 PM   #13
PairTheBoard
Carpal \'Tunnel

Join Date: Dec 2003
Posts: 8,545
Re: Coin flips only 40% likely to flip heads after a heads.

PairTheBoard -
------------------
And my OP description of the probability in question:

"this is saying that if you flip a coin 4 times and IF there is at least one heads in the first 3 flips, THEN

P(randomly picked heads in the first 3 flips is followed by a heads) = 40% "
-------------------------------

Quote:
 Originally Posted by egj No, I don't think this is accurate. Here are the 14 sequences with at least one heads in the first 3 flops: HHHH HHHT HHTH HTHH THHH HHTT HTHT HTTH THHT THTH TTHH HTTT THTT TTHT In those 14 sequences there are 24 heads in positions 1...3. Of those 24 heads, 12 are followed by a heads and 12 are followed by a tails. Exactly as you'd expect.
Yes but that's not P(randomly picked heads in the first 3 flips is followed by a heads) in a random sequence containing at least one head in the first 3 flips.

I think the advantage to my description is it states it as a straight forward probability which can be computed as Aaron W. shows above.

Here's another description.

Take a coin. Flip it 4 times and record the result. If the first 3 flips are tails flip it another 4 times. Repeat until you get a sequence with at least 1 head in the first 3 flips. The sequence may have 1, 2, or 3 heads in the first 3 flips. Randomly pick one of them. Then look at the following flip. The probability that flip is a heads is 17/42.

That's what I mean by

"this is saying that if you flip a coin 4 times and IF there is at least one heads in the first 3 flips, THEN

P(randomly picked heads in the first 3 flips is followed by a heads) = 40%" (approximately - as in the article)

Note: In the OP I gave the exact calculation of 17/42

PairTheBoard

09-29-2015, 07:53 PM   #14
masque de Z
Carpal \'Tunnel

Join Date: Aug 2009
Location: Stanford, CA USA
Posts: 9,533
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Code: ```> rep <- 1e6 > n <- 16 > data <- array(sample(c(0,1), rep*n, replace=TRUE), c(rep,n)) > prob <- rep(NA, rep) > for (i in 1:rep){ + heads1 <- data[i,1:(n-1)]==1 + heads2 <- data[i,2:n]==1 + prob[i] <- sum(heads1 & heads2)/sum(heads1) + } > print(mean(prob, na.rm=TRUE)) [1] 0.4664823```

Length 16 sequence

Quote:
 Code: ```> rep <- 1e6 > n <- 64 > data <- array(sample(c(0,1), rep*n, replace=TRUE), c(rep,n)) > prob <- rep(NA, rep) > for (i in 1:rep){ + heads1 <- data[i,1:(n-1)]==1 + heads2 <- data[i,2:n]==1 + prob[i] <- sum(heads1 & heads2)/sum(heads1) + } > print(mean(prob, na.rm=TRUE)) [1] 0.4919289```
Length 64 sequence

So the lesson is to stop doing stupid things and consider the entire sequence.

Also to be fair i define as hot hand only something that is at least HH already and then we test it. Then i go to even HHH etc...

And you better believe it in real life, where you do not have fair coins and correlations exist, there is a hot hand real effect. Stocks have that also. You play different if you have the attitude of a winner than if you are trying to stop the madness or in between. Something is working when you have a streak sometimes, you have improved the chances to be successful even a bit more.

In poker you must have seen for example that the best players destroy the others with streaks when lucky (it builds on their luck). They tilt opponents to death if they get lucky in a sequence of all ins (even better if its bad beats) and then boom they explode and find themselves pushing all in at flops 100bb deep with only top pair after a reraise or pushing all in preflop with TT etc.

Last edited by masque de Z; 09-29-2015 at 08:11 PM.

09-29-2015, 07:58 PM   #15
Aaron W.
Carpal \'Tunnel

Join Date: Sep 2002
Posts: 29,348
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by masque de Z Length 64 sequence
That's to be expected. As the length increases, the bias decreases. But even at length 64, the calculation leads to a nearly 1% bias.

 09-29-2015, 09:59 PM #16 PairTheBoard Carpal \'Tunnel   Join Date: Dec 2003 Posts: 8,545 Re: Coin flips only 40% likely to flip heads after a heads. It might be easier to see what's going on for the case n=3. Below are the 6 equally likely sequences with at least 1 head in the first two flips. The probability shown next to each is the likelihood a randomly chosen head from the heads in the first 2 flips is followed by a head. The probability at the bottom is the total probability for that event. HHH 2/2 HHT 1/2 HTH 0/1 HTT 0/1 THH 1/1 THT 0/1 ---------- 1/6 (1 + 1/2 + 1) = 5/12 = 41.7% apx So the bias actually increases from n=3 to n=4. Notice the 2 heads followed by heads in HHH are give equal weight as the 1 head followed by a head in THH. PairTheBoard
 09-29-2015, 10:16 PM #17 PairTheBoard Carpal \'Tunnel   Join Date: Dec 2003 Posts: 8,545 Re: Coin flips only 40% likely to flip heads after a heads. The only way this could debunk statistical evidence against say Kobe Bryant's "Hot Hand" would be if the statistics randomly selected 1 shot Kobe made in each game and looked at the following shot. It seems unlikely they would have done it that way. Almost certainly they would have just looked at every shot Kobe made all season and checked to see if he made the following shot. This would not be relevant to that statistical method. And as egj has pointed out, coin flips done that way would show 50%. Of course both cases above restricted to non last shots. PairTheBoard
09-29-2015, 11:02 PM   #18
TomCowley
Pooh-Bah

Join Date: Sep 2004
Posts: 5,645
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by PairTheBoard The only way this could debunk statistical evidence against say Kobe Bryant's "Hot Hand" would be if the statistics randomly selected 1 shot Kobe made in each game and looked at the following shot. It seems unlikely they would have done it that way. Almost certainly they would have just looked at every shot Kobe made all season and checked to see if he made the following shot. This would not be relevant to that statistical method. And as egj has pointed out, coin flips done that way would show 50%. Of course both cases above restricted to non last shots. PairTheBoard
This. These guys deserve a hot hand- and by that I mean a fireplace poker with the end shaped like a hand- shoved right up their asses.

 09-30-2015, 12:39 AM #20 PairTheBoard Carpal \'Tunnel   Join Date: Dec 2003 Posts: 8,545 Re: Coin flips only 40% likely to flip heads after a heads. Penn and Teller had a show some years back where they looked at a guy who claimed that if you spin pennies they end up tails more often than heads (or visa versa). P&T supposedly authenticated it. I tried it with both old and new type pennies and could not find the effect. I have to call B.S. PairTheBoard
 09-30-2015, 08:15 AM #21 FoldnDark Carpal \'Tunnel     Join Date: Jun 2009 Location: St. Louie Posts: 11,939 Re: Coin flips only 40% likely to flip heads after a heads. Craziness! I've always had a weird feeling about these sabermetric guys who try and boil everything in sports down to numbers. Does anyone here not believe in the hot hand in sports? If a guy is hitting a bunch of shots in basketball, you feed him the ball, right? If Big Pappi is crushing everything he sees, you walk him, right?
 09-30-2015, 10:13 AM #22 Kristofero banned     Join Date: Sep 2015 Location: [i]a priori nihilio[/i] UofGtnRekt. Posts: 1,317 Re: Coin flips only 40% likely to flip heads after a heads. PTB, The end of Contact, the novel... It's probably possible that over really long series (think 10^xxx) something might turn up that h/t is actually 40%. Does not mean it is, does mean it is depending on the length of the stretch. Nobody here really knows anything about probability or super-long series, at least not as much as they would like to. Things can and may look fixed but just might be, cobber.
09-30-2015, 10:41 AM   #23
Aaron W.
Carpal \'Tunnel

Join Date: Sep 2002
Posts: 29,348
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by FoldnDark Craziness! I've always had a weird feeling about these sabermetric guys who try and boil everything in sports down to numbers.
I don't have a problem with sabermetrics. At least, the ones who get it are able to accept that noise is a dominant factor in their analyses, and don't make those types of mistakes.

I believe the hot hand can be real. I don't believe that analyses so far correctly capture the concept. Because the operationalized definition is that any made basket (or whatever) increases the probability that the next basket will be made, it doesn't really match the experiential feel of having a hot hand.

There's clearly a psychological element in play, both in the offense and in the defense when a hot hand is identified. It's not like so-and-so just made a layup, so next time we should give him the ball because he's got a hot hand. We know that there's such a thing as "being in the zone."

Quote:
 Does anyone here not believe in the hot hand in sports? If a guy is hitting a bunch of shots in basketball, you feed him the ball, right? If Big Pappi is crushing everything he sees, you walk him, right?
People believe in lots of things and make strategic decisions based on those beliefs all the time. But that doesn't mean those beliefs are a reality. Nor does it mean that those beliefs aren't a reality.

09-30-2015, 11:29 AM   #24
FoldnDark
Carpal \'Tunnel

Join Date: Jun 2009
Location: St. Louie
Posts: 11,939
Re: Coin flips only 40% likely to flip heads after a heads.

Quote:
 Originally Posted by Aaron W. I don't have a problem with sabermetrics. At least, the ones who get it are able to accept that noise is a dominant factor in their analyses, and don't make those types of mistakes. I believe the hot hand can be real. I don't believe that analyses so far correctly capture the concept. Because the operationalized definition is that any made basket (or whatever) increases the probability that the next basket will be made, it doesn't really match the experiential feel of having a hot hand. There's clearly a psychological element in play, both in the offense and in the defense when a hot hand is identified. It's not like so-and-so just made a layup, so next time we should give him the ball because he's got a hot hand. We know that there's such a thing as "being in the zone." People believe in lots of things and make strategic decisions based on those beliefs all the time. But that doesn't mean those beliefs are a reality. Nor does it mean that those beliefs aren't a reality.
Yeah, I'm not ****ting on sabermetrics. There have been a lot of useful contributions from it. The increased use of shifts in baseball is a great example.

What I'm wary of are those people who would argue the stats are it and nothing else. For example, there are plenty of those guys who will argue "the zone" is imaginary. All you're seeing are purely natural runs of no statistical significance, just like if you flipped a coin and happened to get 6 heads in a row. What I believe is that humans are very complex machines in constant need of calibration and fine tuning. When someone, say an 80% free throw shooter, gets locked "in the zone," he is really a 90% free throw shooter (or some number higher than 80) for that period. Later, after he's cooled off, he may lose his fine tuning and be at or below 80% for a period.

 09-30-2015, 12:06 PM #25 PairTheBoard Carpal \'Tunnel   Join Date: Dec 2003 Posts: 8,545 Re: Coin flips only 40% likely to flip heads after a heads. The key things to notice in P("a Heads" is followed by a heads) = 17/42 = apx 40% is how "a Heads" is selected and how in the method of selection the sequence HHHH counts no more toward the calculated probability than TTHH. PairTheBoard

 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is Off Forum Rules
 Forum Jump User Control Panel Private Messages Subscriptions Who's Online Search Forums Forums Home Links to Popular Forums     News, Views, and Gossip     Beginners Questions     Marketplace & Staking     Casino & Cardroom Poker     Internet Poker     NL Strategy Forums     Poker Goals & Challenges     Las Vegas Lifestyle     Sporting Events     Other Other Topics Two Plus Two     About the Forums     Two Plus Two Magazine Forum     The Best of Two Plus Two Marketplace & Staking     Commercial Marketplace     General Marketplace     Staking - Offering Stakes     Staking         Staking - Offering Stakes         Staking - Seeking Stakes         Staking - Selling Shares - Online         Staking - Selling Shares - Live         Staking Rails         Transaction Feedback & Disputes     Transaction Feedback & Disputes Coaching & Training     Coaching Advice     Cash Game Poker Coach Listings     Tournament/SNG Poker Coach Listings Poker News & Discussion     News, Views, and Gossip     Poker Goals & Challenges     Poker Beats, Brags, and Variance     That's What She Said!     Poker Legislation & PPA Discussion hosted by Rich Muny     Twitch - Watch and Discuss Live Online Poker     Televised Poker General Poker Strategy     Beginners Questions     Books and Publications     Poker Tells/Behavior, hosted by: Zachary Elwood     Poker Theory     Psychology No Limit Hold'em Strategy     Medium-High Stakes PL/NL     Micro-Small Stakes PL/NL     Medium-High Stakes Full Ring     Micro-Small Stakes Full Ring     Heads Up NL     Live Low-stakes NL Limit Texas Hold'em Strategy     Mid-High Stakes Limit     Micro-Small Stakes Limit Tournament Poker Strategy     STT Strategy     Heads Up SNG and Spin and Gos     Mid-High Stakes MTT     Small Stakes MTT     MTT Community     Tournament Events Other Poker Strategy     High Stakes PL Omaha     Small Stakes PL Omaha     Omaha/8     Stud     Draw and Other Poker Live Poker     Casino & Cardroom Poker         Venues & Communities         Regional Communities     Venues & Communities     Tournament Events         WPT.com     Home Poker     Cash Strategy     Tournament Strategy Internet Poker     Internet Poker         Global Poker         MPN – Microgaming Poker Network         BetOnline.ag Online Poker     Commercial Software     Software         Commercial Software         Free Software General Gambling     Backgammon Forum hosted by Bill Robertie.     Probability     Sports Betting     Other Gambling Games 2+2 Communities     Other Other Topics         OOTV         Game of Thrones     The Lounge: Discussion+Review     EDF     Las Vegas Lifestyle     BBV4Life         omg omg omg     House of Blogs Sports and Games     Sporting Events         Single-Team Season Threads         Fantasy Sports     Fantasy Sports         Sporting Events     Wrestling     Golf     Chess and Other Board Games     Video Games         League of Legends         Hearthstone     Puzzles and Other Games Other Topics     Politics and Society     Playground: Well Named Loves Social Science     History     Business, Finance, and Investing     Science, Math, and Philosophy     Religion, God, and Theology     Travel     Health and Fitness     Laughs or Links!     Computer Technical Help     Programming

All times are GMT -4. The time now is 10:39 AM.

 Contact Us - Two Plus Two Publishing LLC - Privacy Statement - Top