Is this possible???

[Hanover1]

I am currently in debate with a professed handicapper that has posted 80 (count them..) 80 show bets that he has won in a row. Is this possible? What are the odds?
I took the posture that he was past posting a tad to continue the thread he started.
He claims a great ROI, with the aid of rebates. If he hit 80 in a row, rebates or not, he has done well.
My issue is with the probability (using raw statistics...) of the streak ever occurring.
I come here posing this, as I consider you guys the best of the lot on the internet when it comes to analytical approaches. Is the dude blowing smoke?? What are the odds of this occurring? Rare or impossible?

[MNslappy]

lol, if he started playing a $2 parlay he would be worth trillions and trillions.

btw I don't believe that number for a single second, are you sure he didnt say "18" and you heard him wrong?

[Hanover1]

lol, if he started playing a $2 parlay he would be worth trillions and trillions.

btw I don't believe that number for a single second, are you sure he didnt say "18" and you heard him wrong?

Strike me dead, he quotes, and swears, and in fact posts this publicly, to be fact, using entries of the day, and payoffs, as proof. I am extremely suspicious.
I suspect he is approaching a point to where he will attempt to sell picks, ala. others I am aware of, and this boasting is the hook.
Still curious-is this probable/possible? (80 in a row...)

[GameTheory]

Strike me dead, he quotes, and swears, and in fact posts this publicly, to be fact, using entries of the day, and payoffs, as proof. I am extremely suspicious.
I suspect he is approaching a point to where he will attempt to sell picks, ala. others I am aware of, and this boasting is the hook.
Still curious-is this probable/possible? (80 in a row...)If his overall hit rate to show is 90% or higher, than it enters the realm of possibility that he will have a streak like that every once in a great while, but it is still extremely unlikely.

[woodtoo]

Sure,anything is possible,but if he doesn't post all plays before hand,I don't believe him,truth is in the pre-post pudding:D

[CBedo]

If he had a 95% strike rate, a run of 80 would happen 1 out of 60ish times...
If it dropped to 90% strike rate, a run of 80 would happen 1 out of 4,577 times.

I once hit 17 straight show bets at Bay Meadows, betting every race, and can't remember ever hitting another one, lol.

Possible, but improbable.

[fmolf]

If he had a 95% strike rate, a run of 80 would happen 1 out of 60ish times...
If it dropped to 90% strike rate, a run of 80 would happen 1 out of 4,577 times.

I once hit 17 straight show bets at Bay Meadows, betting every race, and can't remember ever hitting another one, lol.

Possible, but improbable.
every time i go to the track with my one friend we do a showpool,although we bet every race we have only picked all nine or ten races twice in thirty years!...I wouldask him if he picks every race...at which tracks he pleays and what his roi is?....oh and would he be kind enough to lay out his days plays ahead of time for us skeptics?

[fmolf]

If he had a 95% strike rate, a run of 80 would happen 1 out of 60ish times...
If it dropped to 90% strike rate, a run of 80 would happen 1 out of 4,577 times.

I once hit 17 straight show bets at Bay Meadows, betting every race, and can't remember ever hitting another one, lol.

Possible, but improbable.
every time i go to the track with my one friend we do a showpool,although we bet every race we have only picked all nine or ten races twice in thirty years!...I would ask him if he picks every race...at which tracks he plays and what his roi is?....oh and would he be kind enough to lay out his plays for today before post time for us skeptics?

[Jay Trotter]

Just as an FYI...........

....The Ultimate Parlay Challenge has completed 33 contests since January 2007. Starting with a $5.00 bet, the record is twenty-one successful plays in a row with a total of $342.20. This was completed by a very conservative Show player.

The best ever parlay was three plays for a total $2,276.90! Obviously, not Show bets.

I would suggest your friend should be playing our contest and taking some "free money" off the table.

My two cents,

Trotter:ThmbUp:

[Hanover1]

So I am to conclude, before others weigh in here: 80 in a row=fantasy? It would seem so....

[Hanover1]

Based on these premise, a one time streak that arrived at 80 seems an improbability, meaning unlikely.

[mistergee]

it would also be nice to know besides the above mentioned. how long it took to get to 80 plays

[Jay Trotter]

The UPC runs six week contests and nobody has gone the full six weeks without a miss! Even if you're only betting odds on favorites it is very unlikely to run a parlay up to 80 plays.

Another way of looking at this would be to look at the many Showvivor Contests held by various tracks. These contests can have up to 5,000 players and I don't believe there has been a single contest that ended up going to the end! (I don't know this for a fact though)

[JustRalph]

Invite him to Saratoga in August. We could have used him last year :lol:

[Hanover1]

From what I can tell, the guy supposedly plays small harness tracks, with small pools, and plays the chalks to show...
Just checked, and he was accused of "redboarding", whatever that is....

[Jay Trotter]

From what I can tell, the guy supposedly plays small harness tracks, with small pools, and plays the chalks to show...
Just checked, and he was accused of "redboarding", whatever that is....

I would say the odds of hitting 80 consecutive plays are about the same as someone having over 1,300 posts and not knowing what "redboarding" is! Now I think you're just pulling our legs.

[Stillriledup]

For S and Giggles, i put "80 in a row" into the google search.

Your friend doesnt happen to play hoops for the Uconn women's baseketball team, eh?

:D

[Hanover1]

I would say the odds of hitting 80 consecutive plays are about the same as someone having over 1,300 posts and not knowing what "redboarding" is! Now I think you're just pulling our legs.

No leg pulling here my friend. My post are/were of the horse nature, and seldom of the gambling interests. I have since learned that redboading is akin to past posting results. Hope this clears things up for you.
I can give you the info via PM if you want to see this action firsthand......

[Jay Trotter]

No leg pulling here my friend. My post are/were of the horse nature, and seldom of the gambling interests. I have since learned that redboading is akin to past posting results. Hope this clears things up for you.
I can give you the info via PM if you want to see this action firsthand......

I was just being funny.......or at least trying! My kids don't get my humor either! :faint: Just funnin' around!

[Hanover1]

I was just being funny.......or at least trying! My kids don't get my humor either! :faint: Just funnin' around!
Its all good.....

[PhantomOnTour]

I once made 14 successful win wagers in a row over a two day period (last 7 one day and the first 7 the next day) during an EXTREME speed bias at Fairgrounds.
No joke.

[Hanover1]

I once made 14 successful win wagers in a row over a two day period (last 7 one day and the first 7 the next day) during an EXTREME speed bias at Fairgrounds.
No joke.
Your thoughts on 80 in a row to show?

[PhantomOnTour]

Your thoughts on 80 in a row to show?
I'll never say never, but 80 is a huge number...I'm skeptical and, like others, I want a lil more info. Esp the period of time over which this streak was achieved.

[chickenhead]

Sometime between the first horse race ever and when the sun burns out -- someone or more than one will or has hit a ridiculous number of show bets in a row.

This is the internet so I don't believe it. But sure, I believe it in theory.

[Johnny V]

I too am somewhat skeptical of the claim. If even there was 95% hit rate on the bet the odds of winning even 50 in a row would be roughly less than a 7% chance I would think if I am estimating it correctly. It could theoretically happen I guess, but winning 80 in a row is a stretch to me.

[cj's dad]

I did some rough probable odds using an 8 horse field in each of the 80 races and the fact that there are 3 possibilities to show being W-P-S.

My math, if correct, comes to a probability of 4000/1.

Of course the possibility of extraordinary events is not factored in.

[ranchwest]

Who cares? Even if you hit 80 in a row, just a few losses bring you back to even.

[TrifectaMike]

Who cares? Even if you hit 80 in a row, just a few losses bring you back to even.

How would you know that to be true? What is the basis for your statement?

Mike

[Robert Fischer]

I am currently in debate with a professed handicapper that has posted 80 (count them..) 80 show bets that he has won in a row. Is this possible? What are the odds?...

He claims a great ROI, with the aid of rebates. If he hit 80 in a row, rebates or not, he has done well.
My issue is with the probability (using raw statistics...) of the streak ever occurring.
I come here posing this, as I consider you guys the best of the lot on the internet when it comes to analytical approaches. Is the dude blowing
smoke?? What are the odds of this occurring? Rare or impossible?

It appears quite possible. All this other "did he fib" stuff is a separate less interesting issue, that you would have to do your Columbo investigation. You would have to do your MOnk impression, and we can't really guess at that. As far as ROI , I am in a brain freeze but some one on this board can answer if they are motivated enough.
95% strike rate , per WIN getting 2.10back for his $2
do the ROI FORMULA, then mess around with small rebates and see if it is great ROI
maybe i'll crunch it myself if bored
If he had a 95% strike rate, a run of 80 would happen 1 out of 60ish times...
CBedo can you post/link the formula? I am assuming you have it correct. I am familiar with the formula but haven't looked at it in a while to be sure. thanks.

I took the posture that he was past posting a tad to continue the thread he started.
I am assuming that you witnessed or can plainly see (timestamp) that the posts were late? If so continue reading this paragraph. If not turn to page 44. LOL just kidding. Several things can cause past posting in a blog situation. Horseplayers that demand a certain value(irrelevant in this case), and horseplayers who demand a post parade look are two of the most common reasons. One of the reasons it's hard to brag up your picks if you plan on betting late money. Especially if you plan on your brag-ups being free of any late errors that could prevent your real wager.

Based on these premise, a one time streak that arrived at 80 seems an improbability, meaning unlikely.
It isn't that unlikely. If he plays consistently(frequency and number of racing dates, nothing to do with strike rate) often, and if CBedo has the correct formula, (which is very likely) he could have such a streak maybe every several years with avg "luck". When/if Cbedo posts the formula, you can try it with different runs, and it may not be that uncommon to go for 20 or more.

I would ask him if he picks every race
Nope. He demands in the least: supertrainers whose horse has a top jock and lays over a weak field. Usually harness.

lol, if he started playing a $2 parlay he would be worth trillions and trillions.Actually, a $2 parlay @ 80 in a row is worth $78.15 in this case.

Just checked, and he was accused of "redboarding", whatever that is....REDBOARDING is bragging up your picks AFTER the race, usually when you didn't post before the race. However if a bunch of people aren't sure about the math behind what he did and simply didn't believe him, they could falsely accuse him of redboarding. You would have to verify it within the context.


hopefully helpful n didn't butcher the fformat. eyes a bit scrambled

[snoadog]

A 99% strike rate has almost a 50/50 chance of achieving 80 in a row.

[Gapfire]

A 99% strike rate has almost a 50/50 chance of achieving 80 in a row.

.99 to the power of 80 = 0.447

[ranchwest]

How would you know that to be true? What is the basis for your statement?

Mike
If someone hits 80 in a row, they are probably at short prices. About a year ago there was a guy who hit an extremely high percent to demonstrate this was possible, but nearly all were at very short prices. He missed only a very few, but was about break even last time I looked. I think that was on the HTR board.

[TrifectaMike]

If someone hits 80 in a row, they are probably at short prices. About a year ago there was a guy who hit an extremely high percent to demonstrate this was possible, but nearly all were at very short prices. He missed only a very few, but was about break even last time I looked. I think that was on the HTR board.

Thanks.

Mike

[castaway01]

If he did hit 80 in a row, I'll book the next 80. If he hits another 80, then I'll just give him my money and say, "Here, bet whatever you want and send me the profits once a month." :lol:

[BCOURTNEY]

.99 to the power of 80 = 0.447


In this case because all cases are considered to be wins the formula reduction you used is fine. However the expanded and correct form should be used in practice. I was just hoping that you knew you did reduction. You cannot use (strike rate)^(cases of win) when cases of win are not 100%

n= trials, w = wins, p = prob
[n!/(w! * (n-w)!] * (p^w) * ((1-p)^(n-w))

Interestingly, if you want to determine the level of whether or not the claim is smoke. You just need to know the long term strike rate. This percentage alone dictates the truth factor.

For example, I would believe that it would be 93 times more difficult for someone with a 99% strike rate to get 75 of 80 wins in a row than someone with a 96% strike rate. Pretty counter-intuitive yes? In fact, in the 75 of 80 example, the person with a strike rate of about 94% has the best chance of doing a 75 of 80 streak. Neat eh?

What is the historical strike rate claimed on average? Need that and the number of races, and this issue of truth gets vetted very quickly.

[TrifectaMike]

In this case because all cases are considered to be wins the formula reduction you used is fine. However the expanded and correct form should be used in practice. I was just hoping that you knew you did reduction. You cannot use (strike rate)^(cases of win) when cases of win are not 100%

n= trials, w = wins, p = prob
[n!/(w! * (n-w)!] * (p^w) * ((1-p)^(n-w))

Interestingly, if you want to determine the level of whether or not the claim is smoke. You just need to know the long term strike rate. This percentage alone dictates the truth factor.

For example, I would believe that it would be 93 times more difficult for someone with a 99% strike rate to get 75 of 80 wins in a row than someone with a 96% strike rate. Pretty counter-intuitive yes? In fact, in the 75 of 80 example, the person with a strike rate of about 94% has the best chance of doing a 75 of 80 streak. Neat eh?

What is the historical strike rate claimed on average? Need that and the number of races, and this issue of truth gets vetted very quickly.

Pretty counter-intuitive yes? Rather obvious I'd say. Unless one intuitively believed that a value of a win is more relevant that a loss in a probability framework.

Mike

Mike

[BCOURTNEY]

Tell me the number of total wagers made including the 80 in the row. I will tell you if it is a false claim and with what likelihood. Is this a 80 streak out of 500 wagers, 300, 200, 100, 81? The average hit or strike rate claimed can be used to further validate this, it's not required, but would definitely answer the question.

Thanks.

[Gapfire]

In this case because all cases are considered to be wins the formula reduction you used is fine. However the expanded and correct form should be used in practice. I was just hoping that you knew you did reduction. You cannot use (strike rate)^(cases of win) when cases of win are not 100%

n= trials, w = wins, p = prob
[n!/(w! * (n-w)!] * (p^w) * ((1-p)^(n-w))

Interestingly, if you want to determine the level of whether or not the claim is smoke. You just need to know the long term strike rate. This percentage alone dictates the truth factor.

For example, I would believe that it would be 93 times more difficult for someone with a 99% strike rate to get 75 of 80 wins in a row than someone with a 96% strike rate. Pretty counter-intuitive yes? In fact, in the 75 of 80 example, the person with a strike rate of about 94% has the best chance of doing a 75 of 80 streak. Neat eh?

What is the historical strike rate claimed on average? Need that and the number of races, and this issue of truth gets vetted very quickly.

There is no need for "n" here. We are talking about one 80 win streak, and calculating the odds of whether or not this man is telling the truth is very simple.

[TrifectaMike]

There is no need for "n" here. We are talking about one 80 win streak, and calculating the odds of whether or not this man is telling the truth is very simple.

This enter discussion is nonsensical. The probability after the fact, if documentation is available is either 1 or 0.

If we start with no documentation or we ask an academic question prior to any selections then the question is relevant otherwise it's mute.

Mike

[Gapfire]

This enter discussion is nonsensical. The probability after the fact, if documentation is available is either 1 or 0.

If we start with no documentation or we ask an academic question prior to any selections then the question is relevant otherwise it's mute.

Mike

But, there is a chance that the man is telling the truth. So it is not a "moot" point.

[TrifectaMike]

But, there is a chance that the man is telling the truth. So it is not a "moot" point.

I agree. That is my point.

Mike
Moot becomes mute, when you do too much acid.

[BCOURTNEY]

Why would you be dismissive? The "n" provides a measure of the truth. 80 hits in a row is impressive in some cases of "n" and not in others. If the "n" lies in a particular range, the entire claim can be dismissed. If we have "n" we can determine a very accurate approximation of the validity of the given claim.
The question from the OP was to determine whether the claim was true.

There is no need for "n" here. We are talking about one 80 win streak, and calculating the odds of whether or not this man is telling the truth is very simple.

[TrifectaMike]

Why would you be dismissive? The "n" provides a measure of the truth. 80 hits in a row is impressive in some cases of "n" and not in others. If the "n" lies in a particular range, the entire claim can be dismissed. If we have "n" we can determine a very accurate approximation of the validity of the given claim.
The question from the OP was to determine whether the claim was true.

He is saying that n = 80.

Mike

[Gapfire]

Why would you be dismissive? The "n" provides a measure of the truth. 80 hits in a row is impressive in some cases of "n" and not in others. If the "n" lies in a particular range, the entire claim can be dismissed. If we have "n" we can determine a very accurate approximation of the validity of the given claim.
The question from the OP was to determine whether the claim was true.

Because the value of n in this case is unknown. We are talking about one man who purports to have a streak of 80.

[BCOURTNEY]

Does the person claim an average strike/hit rate of 100% as well?

The 80 in a row .. He does not seem to be claiming n=80.

This is unless of course the claim is also these are the only 80 wagers that have or will ever be made using that selection method. I do not believe someone with an edge played 80 times only and ever, then quit and decided that they were going to retire the only "first time 80-in-a-row winner-time-to-quit now person".

The potentially impressive part of the claim suggests that more than 80 wagers occurred AND 80 in a row happened, it is also what makes this interesting or not.

If someone just told me they witnessed 100 coin flips in a row resulting in all heads I would not even blink. If they said it was out of X number of trials, I would blink depending on X. I believe there is a difference, seems that when
claims get made ... circumlocution starts happening.


He is saying that n = 80.
Mike

[BCOURTNEY]

Without the value of N the claim is not interesting at all, because it cannot be evaluated or measured for significantly. I just walked outside and saw a license plate of ABC123, what are the ODDS of that?! Given all the millions of license plates in my state? Same thing, not interesting. N gives the context of the claim and removes the fishing trip story fluff. If the fish is still big after knowing N then kudos to my new hero?

Because the value of n in this case is unknown. We are talking about one man who purports to have a streak of 80.

[TrifectaMike]

Does the person claim an average strike/hit rate of 100% as well?

The 80 in a row .. He does not seem to be claiming n=80.

This is unless of course the claim is also these are the only 80 wagers that have or will ever be made using that selection method. I do not believe someone with an edge played 80 times only and ever, then quit and decided that they were going to retire the only "first time 80-in-a-row winner-time-to-quit now person".

The potentially impressive part of the claim suggests that more than 80 wagers occurred AND 80 in a row happened, it is also what makes this interesting or not.



If someone just told me they witnessed 100 coin flips in a row resulting in all heads I would not even blink. If they said it was out of X number of trials, I would blink depending on X. I believe there is a difference, seems that when
claims get made ... circumlocution starts happening.

I don't disagree, but read Post 1.

Mike

[BCOURTNEY]

Thanks Mike. I missed the rebate portion. The person bets frequently.
Is N > 110? Most probably. If so, I dismiss the claim of 80 hits in a row.
I don't dismiss the possibility 80 bets in a row resulting in profit with rebates.

Who do we need to talk to get N? Or is that part of the secret selection and wagering method .. the number of bets made?


I don't disagree, but read Post 1.
Mike

[Hanover1]

After clearing the cobwebs reading the heavy calculating of the probabilities, I can offer this to perhaps clarify....
The guy posts a pick-it hits. No other betting wagers are offered, i.e. other picks in a bundle. No evidence of the 80 occuring as part of some other betting pattern. Just a straight up "here is my selection".
The streak is contained somewhere in the time span of 18 months. The guy has over 3300 posts in this time span, but I have not deduced the content. His claim is a simple one...."I hit 80 in a row on this board, and see my posts for proof". Still a pill of horse size to swallow......

[BCOURTNEY]

I like statistical puzzles, can you link the post or series if it is public?

After clearing the cobwebs reading the heavy calculating of the probabilities, I can offer this to perhaps clarify....
The guy posts a pick-it hits. No other betting wagers are offered, i.e. other picks in a bundle. No evidence of the 80 occuring as part of some other betting pattern. Just a straight up "here is my selection".
The streak is contained somewhere in the time span of 18 months. The guy has over 3300 posts in this time span, but I have not deduced the content. His claim is a simple one...."I hit 80 in a row on this board, and see my posts for proof". Still a pill of horse size to swallow......

[Hanover1]

I hope this does not get deleted, as I don't want to break any rules....but you can log onto Chicago BarntoWire.com and look under the "lockmiester" thread. The guy posts as Sam.....

[snoadog]

Who do we need to talk to get N? Or is that part of the secret selection and wagering method .. the number of bets made?

It was an 80 bet parlay, therefore w = n. As soon as the first loss occurs the parlay is over. Also the formula you cite gives the probability of exactly w occurences over n trials not w occurences in a row over n trials.

[chickenhead]

what is under discussion here so far is the probability of this particular guy hitting 80 in a row. If we operate under the assumption (play along) that anyone that hits 80 in a row will post "Hey I hit 80 in a row" on a message board -- calculating the likelihood includes everyone "like him" over the period he's been doing what he's been doing.

Lets make up some variables to get an idea of the ballpark..

The Year, 2000, we begin. 500 guys in a room. They are of identical ability, and record identical hit rates. They place exactly 1000 bets a year each. Fast forward to the year 2010, 10,000 show bets per person later.

What is the required hit rate for the the likelihood to be greater than 50% that one or more of those people experienced an 80+ race hit streak?

[BCOURTNEY]

The formula cited is correct. Wins occurring in a row inside any N number of trials. n this case they are being claimed to be same thing, I'm offering that they are not. This is a rebate player, and likely frequently bets, n is not 80, and if it's higher than say 110, the claim is most likely a fishing story. I don't see the how the wager being "a parlay" is relevant to the claim of the streak at all.

It was an 80 bet parlay, therefore w = n. As soon as the first loss occurs the parlay is over. Also the formula you cite gives the probability of exactly w occurences over n trials not w occurences in a row over n trials.

[BCOURTNEY]

Chickenhead,

I like the idea, I think a problem is that these monkeys smashing bets on the keyboards are not infinitely wealthy... Do our monkeys need to make money?

what is under discussion here so far is the probability of this particular guy hitting 80 in a row. If we operate under the assumption (play along) that anyone that hits 80 in a row will post "Hey I hit 80 in a row" on a message board -- calculating the likelihood includes everyone "like him" over the period he's been doing what he's been doing.

Lets make up some variables to get an idea of the ballpark..

The Year, 2000, we begin. 500 guys in a room. They are of identical ability, and record identical hit rates. They place exactly 1000 bets a year each. Fast forward to the year 2010, 10,000 show bets per person later.

What is the required hit rate for the the likelihood to be greater than 50% that one or more of those people experienced an 80+ race hit streak?

[chickenhead]

Chickenhead,

I like the idea, I think a problem is that these monkeys smashing bets on the keyboards are not infinitely wealthy... Do our monkeys need to make money?

No, it doesn't really matter for this. Let's assume they are $2 bettors, they risk only $2000 per year. Their bankrolls are easily funded, profit and loss are inconsequential to them. They just enjoy making show bets as a form of entertainment. We don't really know how good or bad they need to be yet....that's our question.

[BCOURTNEY]

Hanover1,

Ok I read a large number of the posts. The content seems intended to point others to lose money. Misinformation is a useful strategy in some games. Delusion is harmful to the individual and sometimes others. That is the most objective assessment I can make. To answer your original question, it is a false claim, nothing to see here, move along etc.

I hope this does not get deleted, as I don't want to break any rules....but you can log onto Chicago BarntoWire.com and look under the "lockmiester" thread. The guy posts as Sam.....

[snoadog]

The formula cited is correct. Wins occurring in a row inside any N number of trials

A trivial example: N=3, W=2, P=.5. Probability of 2 wins in 3 trials is 0.375 according to the formula. This is correct. Now, If you write out the win/loss trials.

1)WWW
2)WWL
3)WLW
.
etc.
.
8)LLL

You will see there are 8 different trials each with a prob. of 0.125. (0.5 to the power of 3, each) But only 2 of the trials have 2 wins in a row and that is a prob. of .250 not .375 as the formula predicts. The formula accounts for ALL occurences 2 wins...even WLW which is not 2 wins in a row.
You are ignoring this fact.

[BCOURTNEY]

I assume this is important to you as you have taken issue with my
formula -- it is both correct and sound. Left out of your analysis the variable dependence introduced as soon as it is declared that order matters in looking
for the 2 win sequence. I really hope your pulling my leg friend. If not here
are some sites you might find helpful in understanding this. Pascal's triangle comes to mind also ...

http://en.wikibooks.org/wiki/Probability/Combinatorics
http://en.wikibooks.org/wiki/Probability/Probability_Spaces

Here's a link to "The Wizard of Odds" web site.
It's written by a professional actuary who strangely has the exact same
formula ( it's not a secret ):

http://wizardofodds.com/askthewizard/probability-coins.html

A trivial example: N=3, W=2, P=.5. Probability of 2 wins in 3 trials is 0.375 according to the formula. This is correct. Now, If you write out the win/loss trials.

1)WWW
2)WWL
3)WLW
.
etc.
.
8)LLL

You will see there are 8 different trials each with a prob. of 0.125. (0.5 to the power of 3, each) But only 2 of the trials have 2 wins in a row and that is a prob. of .250 not .375 as the formula predicts. The formula accounts for ALL occurences 2 wins...even WLW which is not 2 wins in a row.
You are ignoring this fact.

[snoadog]

I have no issue with the binomial probability distribution formula. Of course it is correct and sound but it does not predict the probability of w wins in a row, it predicts the probability of w wins whether or not they are in a row.

I don't need a formula to figure out how many W's in a row there are in N.

There are 2 rows of 3 in 4 N: (4-3+1)
There are 3 rows of 3 in 5 Ns: (5-3+1)
There are 6 rows of 5 in 10 Ns: (10-5+1)
There are 31 rows of 80 in 110 Ns: (110-80+1)

It's trivial.

Also, a parlay is a special case of the binomial probability distribution where W = N.

[PaceAdvantage]

I hope this does not get deleted, as I don't want to break any rules....but you can log onto Chicago BarntoWire.com and look under the "lockmiester" thread. The guy posts as Sam.....Why would providing a direct link to the subject of this thread be breaking the rules? I am assuming of course that this Sam fellow is who you were referring to in your original post...

[BCOURTNEY]

Tried to help. You forgot the "columns" in concert with your "row" calculations, they matter and cause exclusions. You should revisit this as some point. I'm done. Good luck.

I have no issue with the binomial probability distribution formula. Of course it is correct and sound but it does not predict the probability of w wins in a row, it predicts the probability of w wins whether or not they are in a row.

I don't need a formula to figure out how many W's in a row there are in N.

There are 2 rows of 3 in 4 N: (4-3+1)
There are 3 rows of 3 in 5 Ns: (5-3+1)
There are 6 rows of 5 in 10 Ns: (10-5+1)
There are 31 rows of 80 in 110 Ns: (110-80+1)

It's trivial.

Also, a parlay is a special case of the binomial probability distribution where W = N.

[BCOURTNEY]

You are doing this --> Probability of an Independent Event
You need to be doing this --> Probability of Dependent Events

http://www.wikihow.com/Find-Mathematical-Probabilities

I have no issue with the binomial probability distribution formula. Of course it is correct and sound but it does not predict the probability of w wins in a row, it predicts the probability of w wins whether or not they are in a row.

I don't need a formula to figure out how many W's in a row there are in N.

There are 2 rows of 3 in 4 N: (4-3+1)
There are 3 rows of 3 in 5 Ns: (5-3+1)
There are 6 rows of 5 in 10 Ns: (10-5+1)
There are 31 rows of 80 in 110 Ns: (110-80+1)

It's trivial.

Also, a parlay is a special case of the binomial probability distribution where W = N.

[JohnGalt1]

I am currently in debate with a professed handicapper that has posted 80 (count them..) 80 show bets that he has won in a row. Is this possible? What are the odds?
I took the posture that he was past posting a tad to continue the thread he started.
He claims a great ROI, with the aid of rebates. If he hit 80 in a row, rebates or not, he has done well.
My issue is with the probability (using raw statistics...) of the streak ever occurring.
I come here posing this, as I consider you guys the best of the lot on the internet when it comes to analytical approaches. Is the dude blowing smoke?? What are the odds of this occurring? Rare or impossible?

No where do you state how many horses he bet per race.

Howard Sartin advocates betting 2 horses to win (odds high enough).

What if in some races he bet 2 or even 3 in a race.

And if he bet 3 in a race and they finished 1-2-3, would that count as one race in the streak or 3 since he would've got paid for all three bets.

If it is found that he bet one horse per race then I find this streak improbable, not impossible.

But without full information all we've done is have fun? speculating.

[TrifectaMike]

Initially when I read this post, I chuckled, because I thought it was trivial.

Let's accept the question at face value, and provide some answers.

Is this possible???
I am currently in debate with a professed handicapper that has posted 80 (count them..)
80 show bets that he has won in a row. Is this possible? What are the odds? Answer: Of course it is possible. Odds? I don't know. There are no failures only a streak of 80 successes.

I took the posture that he was past posting a tad to continue the thread he started.He claims a great ROI, with the aid of rebates. If he hit 80 in a row, rebates or not, he has done well. Answer: From a probability point of view, this is meaningless.

My issue is with the probability (using raw statistics...) of the streak ever occurring.I come here posing this, as I consider you guys the best of the lot on the internet when it comes to analytical approaches. Is the dude blowing smoke?? What are the odds of this occurring? Rare or impossible?
Answer: What statistics and what probabilities? None are provided.
And none can be inferred.
.
Essentially there is insufficient information, except to say "of course it is possible".

In an attempt by some, with good intentions, to provide an "answer", we get into "what if" situations. Along with the "what ifs" come the formulas, distributions, independence, dependence, order, unordered, etc.And none of the above has anything to do with original question, Is it possible? Answer: Yes, it is possible.

The majority of the posts relate to Binomial experiments and the Binomial distribution. So, let's first define them.

A Binomial distribution is obtained from a probability experiment called a binomial experiment. The experiment must satisfy these conditions:

1. Each trial can have only two outcomes or outcomes that can reduced to two outcomes. The outcomes are usually considered as a success or failure (win loss).

2. There is a fixed number of trials.

3. The outcomes of each trial are independent of each other.

4. The probability of success must remain the same for each trial.

You will notice that order of the success or failure is not included in the above conditions.

Let's perform a binomial experiment.

If we play three races and want to know the probability of winning two races and losing one race and we have a hit rate of 40%. This can be determined as:

P(winning 2 races and losing 1 race) = P(winning 2 races)*P(losing 1 race)

P(winning 2 races and losing 1 race) = (.4)(.4)(.6) = 9.6%

The probability is only correct if the win/loss is order specific (win-win-loss).

Now, if order is not specific, we would also have win-loss-win and loss-win-win. In this case the probability is 9.6% + 9.6% + 9.6% or 28.8%.

This leads to the binomial theorem: P(x=k) = n!/k!(n-k)! *p^k(1-p)^(n-k)

where
K= successes
n= number of trials
p= probability of success
!= factorial

The main point is that the binomial theorem is independent of order.

Mike

[BCOURTNEY]

Wow.

Initially when I read this post, I chuckled, because I thought it was trivial.

Let's accept the question at face value, and provide some answers.

Is this possible???
I am currently in debate with a professed handicapper that has posted 80 (count them..)
80 show bets that he has won in a row. Is this possible? What are the odds? Answer: Of course it is possible. Odds? I don't know. There are no failures only a streak of 80 successes.

I took the posture that he was past posting a tad to continue the thread he started.He claims a great ROI, with the aid of rebates. If he hit 80 in a row, rebates or not, he has done well. Answer: From a probability point of view, this is meaningless.

My issue is with the probability (using raw statistics...) of the streak ever occurring.I come here posing this, as I consider you guys the best of the lot on the internet when it comes to analytical approaches. Is the dude blowing smoke?? What are the odds of this occurring? Rare or impossible?
Answer: What statistics and what probabilities? None are provided.
And none can be inferred.
.
Essentially there is insufficient information, except to say "of course it is possible".

In an attempt by some, with good intentions, to provide an "answer", we get into "what if" situations. Along with the "what ifs" come the formulas, distributions, independence, dependence, order, unordered, etc.And none of the above has anything to do with original question, Is it possible? Answer: Yes, it is possible.

The majority of the posts relate to Binomial experiments and the Binomial distribution. So, let's first define them.

A Binomial distribution is obtained from a probability experiment called a binomial experiment. The experiment must satisfy these conditions:

1. Each trial can have only two outcomes or outcomes that can reduced to two outcomes. The outcomes are usually considered as a success or failure (win loss).

2. There is a fixed number of trials.

3. The outcomes of each trial are independent of each other.

4. The probability of success must remain the same for each trial.

You will notice that order of the success or failure is not included in the above conditions.

Let's perform a binomial experiment.

If we play three races and want to know the probability of winning two races and losing one race and we have a hit rate of 40%. This can be determined as:

P(winning 2 races and losing 1 race) = P(winning 2 races)*P(losing 1 race)

P(winning 2 races and losing 1 race) = (.4)(.4)(.6) = 9.6%

The probability is only correct if the win/loss is order specific (win-win-loss).

Now, if order is not specific, we would also have win-loss-win and loss-win-win. In this case the probability is 9.6% + 9.6% + 9.6% or 28.8%.

This leads to the binomial theorem: P(x=k) = n!/k!(n-k)! *p^k(1-p)^(n-k)

where
K= successes
n= number of trials
p= probability of success
!= factorial

The main point is that the binomial theorem is independent of order.

Mike

[Hanover1]

Why would providing a direct link to the subject of this thread be breaking the rules? I am assuming of course that this Sam fellow is who you were referring to in your original post...
I don't know how to post links is the problem, horsemen are not the brightest bulbs in the box, even though horses make us look good at times.
Yes, Sam is the poster moniker. Been in dutch a time or two here, and am trying to tow the line and contribute without the hassles is all.....

[highnote]

If, on average, he has a 90% chance of hitting the show bet then the probability of winning 80 in a row is 0.90 to the power of 80 = .000218 -- or about 2 times in 10,000.

Not outside the realm of possibility, but not likely, either.

Now sometimes his true chance of winning a shot bet is going to be higher and sometimes lower. Or maybe he waits for extreme favorites.

[Hanover1]

If, on average, he has a 90% chance of hitting the show bet then the probability of winning 80 in a row is 0.90 to the power of 80 = .000218 -- or about 2 times in 10,000.

Not outside the realm of possibility, but not likely, either.

Now sometimes his true chance of winning a shot bet is going to be higher and sometimes lower. Or maybe he waits for extreme favorites.

Not to say this is correct or otherwise, but it is a simple enough answer to my question that I posted. The other responses carried the questions I posed and infered other actions took place that I did not say took place.
In short-a simple answer to a simple question. Not "we need more information". Good one.....

[dylbert]

Hope this fella was playing both Tapizar & Comma at the Top today... that will surely STOP THIS THREAD and its utter nonsense!!!

[TrifectaMike]

Not to say this is correct or otherwise, but it is a simple enough answer to my question that I posted.

Why post the question, if you didn't want a correct answer. So, are you stating that you wanted an incorrect answer as long as it was simple?

Just f'king amazing.

Mike

[Gapfire]

Not to say this is correct or otherwise, but it is a simple enough answer to my question that I posted. The other responses carried the questions I posed and infered other actions took place that I did not say took place.
In short-a simple answer to a simple question. Not "we need more information". Good one.....

I gave you the same solution about 3 pages back.

[Hanover1]

Why post the question, if you didn't want a correct answer. So, are you stating that you wanted an incorrect answer as long as it was simple?

Just f'king amazing.

Mike

Yep, you guessed it-I wanted a lie. Just f'king amazing.

[highnote]

I gave you the same solution about 3 pages back.

Your variables were different, but your formula was the same.

I suppose something "clicked" with Hanover with the way he interpreted my explanation or the fact that he saw the solution presented differently.

That's why in college I used to go to the library and look at different math textbooks for the same type of problems. I would sometimes find that when I saw a procedure presented in a couple of different way it made more sense to me.

I studied statistics in college, but never understood the concept of odds, although I knew how to make them. Then when I started reading racing books I still didn't understand them, but I knew how to calculate them and calculate my payoffs.

Then one day I was reading a British book about on-track bookmakers and odds were defined as the ratio of the probability to losing to the probability of winning. I could understand that explanation. If a horse has a 2/3 chance of losing and a 1/3 chance of winning then his odds are 2-1. It's simple, but until I saw it explained like that it never really sunk in to my thick head in a way that was really useful to me.

Sorry. This is a longwinded comment. :D

[Zman179]

I certainly believe that hitting 80 show bets in a row is possible, considering that it seems he limited his plays to half-mile trotter tracks on favorites. I also believe that he'll most likely never do it again.

Anything is possible. One time on the craps table I rolled ELEVEN hard numbers in a row.

[TrifectaMike]

[QUOTE=swetyejohn].
Then one day I was reading a British book about on-track bookmakers and odds were defined as the ratio of the probability to losing to the probability of winning. I could understand that explanation. If a horse has a 2/3 chance of losing and a 1/3 chance of winning then his odds are 2-1. It's simple, but until I saw it explained like that it never really sunk in to my thick head in a way that was really useful to me./QUOTE]

Based on what you stated, what are the odds he wins his 81st bet?

Mike

[citygoat]

Who was covering his parlay bets after he hit 58 in a row? I didn't notice any bridgejumper pools that magnificent.

[Hanover1]

I certainly believe that hitting 80 show bets in a row is possible, considering that it seems he limited his plays to half-mile trotter tracks on favorites. I also believe that he'll most likely never do it again.

Anything is possible. One time on the craps table I rolled ELEVEN hard numbers in a row.

My dad lived in Vegas for many years, and I kept hearing tales of 28 straight passes on the craps table. An incredible run.....

[Hanover1]

Your variables were different, but your formula was the same.

I suppose something "clicked" with Hanover with the way he interpreted my explanation or the fact that he saw the solution presented differently.

That's why in college I used to go to the library and look at different math textbooks for the same type of problems. I would sometimes find that when I saw a procedure presented in a couple of different way it made more sense to me.

I studied statistics in college, but never understood the concept of odds, although I knew how to make them. Then when I started reading racing books I still didn't understand them, but I knew how to calculate them and calculate my payoffs.

Then one day I was reading a British book about on-track bookmakers and odds were defined as the ratio of the probability to losing to the probability of winning. I could understand that explanation. If a horse has a 2/3 chance of losing and a 1/3 chance of winning then his odds are 2-1. It's simple, but until I saw it explained like that it never really sunk in to my thick head in a way that was really useful to me.

Sorry. This is a longwinded comment. :D

Sums it up nicely....

[Track Collector]

[QUOTE=swetyejohn].
Then one day I was reading a British book about on-track bookmakers and odds were defined as the ratio of the probability to losing to the probability of winning. I could understand that explanation. If a horse has a 2/3 chance of losing and a 1/3 chance of winning then his odds are 2-1. It's simple, but until I saw it explained like that it never really sunk in to my thick head in a way that was really useful to me./QUOTE]

Based on what you stated, what are the odds he wins his 81st bet?

Mike

While people have a hard time wrapping their arms around this, having already won 80 in a row, the probability of winning the 81st is still the same as any one-race winning expectation. If a favorite wins 1/3 of the time, and the last 20 races have been won by the non-favorite, the favorite's probability of winning race# 21 is still 33%. Each race is still an independent event not influenced by previous results.

[Fastracehorse]

[QUOTE=TrifectaMike]

While people have a hard time wrapping their arms around this, having already won 80 in a row, the probability of winning the 81st is still the same as any one-race winning expectation. If a favorite wins 1/3 of the time, and the last 20 races have been won by the non-favorite, the favorite's probability of winning race# 21 is still 33%. Each race is still an independent event not influenced by previous results.

...it's like when people think a horse is due....
....this is called the 'Gambler's Fallacy'

fffastt

[TrifectaMike]

[QUOTE=TrifectaMike]

While people have a hard time wrapping their arms around this, having already won 80 in a row, the probability of winning the 81st is still the same as any one-race winning expectation. If a favorite wins 1/3 of the time, and the last 20 races have been won by the non-favorite, the favorite's probability of winning race# 21 is still 33%. Each race is still an independent event not influenced by previous results.

And what is the one-race winning expectation of the gentlemen in question in this thread. No fluff, just give me a number.

Mike

[thaskalos]

[QUOTE=Track Collector]

And what is the one-race winning expectation of the gentlemen in question in this thread. No fluff, just give me a number.

Mike92.5%

[TrifectaMike]

[QUOTE=TrifectaMike]92.5%

Now, do please tell me how you arrived at 92.5%.

Mike

[TrifectaMike]

If one were to list 80 winners in a row ( complete sequence -either true or not) someone should be able to tell me what the probability of winning the next bet would be. Surely someone can answer this question.

Mike

[snoadog]

I don't pretend to know the answer to this one, but my intuitive guess is that if this handicappng god has been no more or less luckier than usual then 80 in a row is likely the result of a 99.14% strike rate of.

.50 = x^80
crunch, crunch. x = .9914

[TrifectaMike]

I don't pretend to know the answer to this one, but my intuitive guess is that if this handicappng god has been no more or less luckier than usual then 80 in a row is likely the result of a 99.14% strike rate of.

.50 = x^80
crunch, crunch. x = .9914

Good man. Honest answer. You don't know.

Now, if I were to backup one race, and ask the same question after 79 races. What is the probability of winning the next race, I would expect the same answer. So, let's continually back step one race at a time, after 78, 77, 76, .... 1 race. I would still ask the same question and would expect the same answer. If that is the case, how can anyone draw any conclusions on 80 winners in a row without any context.

Mike

[thaskalos]

[QUOTE=thaskalos]

Now, do please tell me how you arrived at 92.5%.

MikeVery simple...

I am well acquainted with this guy that Hanover1 is talking about...so I called him up and asked him.

He told me that the winning % on his "key" show bets is...92.5%

[BCOURTNEY]

At the 92.5% hit rate given:

The chance of 80 of 80 hits in a row is : 1 in 511.28
The chance of 81 of 81 hits in a row is : 1 in 552.73


If one were to list 80 winners in a row ( complete sequence -either true or not) someone should be able to tell me what the probability of winning the next bet would be. Surely someone can answer this question.

Mike

[snoadog]

Now, if I were to backup one race, and ask the same question after 79 races. What is the probability of winning the next race, I would expect the same answer. So, let's continually back step one race at a time, after 78, 77, 76, .... 1 race. I would still ask the same question and would expect the same answer. If that is the case, how can anyone draw any conclusions on 80 winners in a row without any context.

Well okay... but I don't use the 99.14% to conclude he is a spectacularly unusual handicapper. I use it to conclude that 80 in a row is a spectacularly unusual event.

[BCOURTNEY]

I'm well convinced now that the majority of this message board is insane regarding probabilities. In the case of a 99.14% hit rate, if that streak of 80 wins in a row occurred WITHIN a larger n > 90 total bet sequence ( win or lose ) the odds of the 80 in a row occurring streak go from about ~ 1 in 500 to ~ 1 in 12,000,000 .. and further compounds. It wasn't an unusual event it simply did not happen.

Well okay... but I don't use the 99.14% to conclude he is a spectacularly unusual handicapper. I use it to conclude that 80 in a row is a spectacularly unusual event.

[speculus]

Here are my two cents:

The Mathematics Of Luck (http://prakashgosavi.blogspot.com/2009/06/mathematics-of-luck.html)

[BCOURTNEY]

Speculus,

Nice link. Linked article was good. His numbers are slightly off, I included charts for 10 and 11 in a row using my much questioned formula. Apparently, other crazy people are infected with it's use it also.

(Strike Rate) (1 in N Chance of Streak Start) (Avg. Bets Required)
0.9 3.186635545 14.18663555
0.8 11.64153218 22.64153218
0.75 23.67696885 34.67696885
0.7 50.57333107 61.57333107
0.6 275.6361948 286.6361948
0.5 2048 2059
0.4 23841.85791 23852.85791
0.35 103574.182 103585.182

(Strike Rate) (1 in N Chance of Streak Start) (Avg. Bets Required)
0.9 2.867971991 12.86797199
0.8 9.313225746 19.31322575
0.75 17.75772663 27.75772663
0.7 35.40133175 45.40133175
0.6 165.3817169 175.3817169
0.5 1024 1034
0.4 9536.743164 9546.743164
0.35 36250.96371 36260.96371

Here are my two cents:

The Mathematics Of Luck (http://prakashgosavi.blogspot.com/2009/06/mathematics-of-luck.html)

[PaceAdvantage]

Why post the question, if you didn't want a correct answer. So, are you stating that you wanted an incorrect answer as long as it was simple?

Just f'king amazing.

MikeYour attitude is just f'king amazing. Settle down Beavis.

[TrifectaMike]

Beavis here again (Dr. Beavis to PA)

The concept of probability (likelihood) is only useful for events that are yet to occur. Once they have occurred, the outcome is no longer probable. Probability of an outcome for a past event is no longer relevant since the outcome is certain.

Mike(Dr. Beav)

[snoadog]

I'm well convinced now that the majority of this message board is insane regarding probabilities. In the case of a 99.14% hit rate, if that streak of 80 wins in a row occurred WITHIN a larger n > 90 total bet sequence ( win or lose ) the odds of the 80 in a row occurring streak go from about ~ 1 in 500 to ~ 1 in 12,000,000 .. and further compounds. It wasn't an unusual event it simply did not happen.

Are you are trying to tell me that the likelyhood of 80 wins in a row out of 80 tries is greater than 80 wins in a row out of n>80 tries, despite the extra tries n>than 80 affords me??? Only common sense is required to discover the error if this logic.

[BCOURTNEY]

Yes, your common sense and intuition cloud your judgement, they are also wrong. Your insistence reminds me of someone seeing the Monty Hall paradox for the first time. If you want to be upset with intuition you can try it out also. This one has a history of dropping PhD's in it's graveyard.

http://en.wikipedia.org/wiki/Monty_Hall_problem

Cheers.

Are you are trying to tell me that the likelyhood of 80 wins in a row out of 80 tries is greater than 80 wins in a row out of n>80 tries, despite the extra tries n>than 80 affords me??? Only common sense is required to discover the error if this logic.

[snoadog]

I'm certain there are many instances in discussions of probability/statistics where you cannot trust your intuition, but this is not one of them.

[PICSIX]

I am currently in debate with a professed handicapper that has posted 80 (count them..) 80 show bets that he has won in a row. Is this possible? What are the odds?
I took the posture that he was past posting a tad to continue the thread he started.
He claims a great ROI, with the aid of rebates. If he hit 80 in a row, rebates or not, he has done well.
My issue is with the probability (using raw statistics...) of the streak ever occurring.
I come here posing this, as I consider you guys the best of the lot on the internet when it comes to analytical approaches. Is the dude blowing smoke?? What are the odds of this occurring? Rare or impossible?

Lets end this question once and for all. It's simple, you have two choices. Either he did it or he didn't--so the odds are 50/50. :lol: :lol: :lol:

[BCOURTNEY]

Conditional probability is the probability that something will happen given that something else has already happened e.g. picking the next winner, after just selecting one. These events must occur in sequence to keep with the in-a-row requirement, as such, they create a dependency. Don't confuse joint and conditional probabilities. A Venn diagram can be helpful in thinking about this. Your assumption of independence is flawed and therein lies the problem, but as long as your discussing this, I feel that you are beginning to question your intuition and I'm hopeful you can learn something. This doesn't mean you have to be "right", although it is preferable, you can continue to live in the "not even wrong" as my physics friends like to put it. We are on the same team fighting the probability machine friend.

I'm certain there are many instances in discussions of probability/statistics where you cannot trust your intuition, but this is not one of them.

[BCOURTNEY]

The resulting discussion of probability and the exposure of misunderstanding by so many people on the board suggests it's back to a 101 reading on joint and conditional probability.

Who cares about this guy now? He helped everyone learn how flawed their thinking was, and further entrenched those with gigantic egos whom have nothing to learn? So my thanks to Mr. 80/80. :ThmbUp:


Lets end this question once and for all. It's simple, you have two choices. Either he did it or he didn't--so the odds are 50/50. :lol: :lol: :lol:

[Coleman]

Lets end this question once and for all. It's simple, you have two choices. Either he did it or he didn't--so the odds are 50/50. :lol: :lol: :lol:

And the odds are close to zero that anything anyone has posted here is going to change his mind about whether he believes this guy is telling the truth or not.

[TrifectaMike]

Initially when I read this post, I chuckled, because I thought it was trivial.

Let's accept the question at face value, and provide some answers.

Is this possible???
I am currently in debate with a professed handicapper that has posted 80 (count them..)
80 show bets that he has won in a row. Is this possible? What are the odds? Answer: Of course it is possible. Odds? I don't know. There are no failures only a streak of 80 successes.

I took the posture that he was past posting a tad to continue the thread he started.He claims a great ROI, with the aid of rebates. If he hit 80 in a row, rebates or not, he has done well. Answer: From a probability point of view, this is meaningless.

My issue is with the probability (using raw statistics...) of the streak ever occurring.I come here posing this, as I consider you guys the best of the lot on the internet when it comes to analytical approaches. Is the dude blowing smoke?? What are the odds of this occurring? Rare or impossible?
Answer: What statistics and what probabilities? None are provided.
And none can be inferred.
.
Essentially there is insufficient information, except to say "of course it is possible".

In an attempt by some, with good intentions, to provide an "answer", we get into "what if" situations. Along with the "what ifs" come the formulas, distributions, independence, dependence, order, unordered, etc.And none of the above has anything to do with original question, Is it possible? Answer: Yes, it is possible.

The majority of the posts relate to Binomial experiments and the Binomial distribution. So, let's first define them.

A Binomial distribution is obtained from a probability experiment called a binomial experiment. The experiment must satisfy these conditions:

1. Each trial can have only two outcomes or outcomes that can reduced to two outcomes. The outcomes are usually considered as a success or failure (win loss).

2. There is a fixed number of trials.

3. The outcomes of each trial are independent of each other.

4. The probability of success must remain the same for each trial.

You will notice that order of the success or failure is not included in the above conditions.

Let's perform a binomial experiment.

If we play three races and want to know the probability of winning two races and losing one race and we have a hit rate of 40%. This can be determined as:

P(winning 2 races and losing 1 race) = P(winning 2 races)*P(losing 1 race)

P(winning 2 races and losing 1 race) = (.4)(.4)(.6) = 9.6%

The probability is only correct if the win/loss is order specific (win-win-loss).

Now, if order is not specific, we would also have win-loss-win and loss-win-win. In this case the probability is 9.6% + 9.6% + 9.6% or 28.8%.

This leads to the binomial theorem: P(x=k) = n!/k!(n-k)! *p^k(1-p)^(n-k)

where
K= successes
n= number of trials
p= probability of success
!= factorial

The main point is that the binomial theorem is independent of order.

Mike

Have you guys lost your collective minds? This is simple probability 101. There are no joint, no conditional, no subjective, no objective or any other probability in this situation.

IT IS A SIMPLE PROBLEM EASILY SOLVED BY USING THE BINOMIAL THEOREM.

Read the post. Apply the theorem.

LET ME HELP YOU!

An example that anyone can understand.

An average basketball player makes 3/4 of his free throws, find the probability he makes the next 4 free throws.

P(4 successful shots) = 1*(3/4)**4 *(1/4)**0

P(4 successful shots) = 81/256 = .3164

Substitute whatever successes you want.

I have never heard such nonsense!

Mike (Dr. Beav)

[Coleman]

A better analogy of the problem posed might be:

A basketball player, who seems pretty good at free throws, says that he made 80 in a row. Do you think he's telling the truth?

[TrifectaMike]

A better analogy of the problem posed might be:

A basketball player, who seems pretty good at free throws, says that he made 80 in a row. Do you think he's telling the truth?

Coleman,

There are several things going on in this thread at the same time.

I didn't address the question of truthfulness. As I said earlier...

The concept of probability (likelihood) is only useful for events that are yet to occur. Once they have occurred, the outcome is no longer probable. Probability of an outcome for a past event is no longer relevant since the outcome is certain.

So, if his 80 in a row were observed, probability doesn't play a role. However, if the question is asked, Can he repeat it?... or the question is asked prior to taking his first shot, then it becomes a question of probability,

Mike (Dr. Beav)

[snoadog]

Careful Mike, somone is about to bring his Ns out again :)

[TrifectaMike]

Careful Mike, somone is about to bring his Ns out again :)

That's ok. A bad case of the N's is nothing compared to convincing some colleagues that confidence values are NOT probabilities, and they had doctorates in math. There are times when strangeness simply appears.

Mike (Dr. Beav)

[Hanover1]

A better analogy of the problem posed might be:

A basketball player, who seems pretty good at free throws, says that he made 80 in a row. Do you think he's telling the truth?

No.......

[Hanover1]

The resulting discussion of probability and the exposure of misunderstanding by so many people on the board suggests it's back to a 101 reading on joint and conditional probability.

Who cares about this guy now? He helped everyone learn how flawed their thinking was, and further entrenched those with gigantic egos whom have nothing to learn? So my thanks to Mr. 80/80. :ThmbUp:

It appears that what is obvious is quite often not what one actually sees, however in this case the Emperor indeed has no clothes......

[BCOURTNEY]

We know when we are considering 80 in a row and n > 80 trials -- conditional probability does absolutely play a role, and in the case of exactly n = 80 conditional probability does not play a role. Only in the case that n = 80 and the in-a-row = 80 does the binomial theorem account for all of the probability. My point is that even off hand analysis suggests that n is > 80, since the person claiming the hit is a rebate player. So binomial theorem is not what you use to perform analysis. However the formula I gave ALLOWS for the use of the binomial theorem in the SPECIAL case when in-a-row matches the number of n trials, and also allows someone to analyze in the real world if the in-a-row is possible of occurring in a larger series of n trials.

All of that being said are you suggesting then that the possibility of 80 in a row of 100 trials has the same likelihood of occurring as 80 in a row within 1000 trials, if the strike rate is fixed? This is not 80 of 100 or 80 of 1000, in a row.

Have you guys lost your collective minds? This is simple probability 101. There are no joint, no conditional, no subjective, no objective or any other probability in this situation.

IT IS A SIMPLE PROBLEM EASILY SOLVED BY USING THE BINOMIAL THEOREM.

Read the post. Apply the theorem.

LET ME HELP YOU!

An example that anyone can understand.

An average basketball player makes 3/4 of his free throws, find the probability he makes the next 4 free throws.

P(4 successful shots) = 1*(3/4)**4 *(1/4)**0

P(4 successful shots) = 81/256 = .3164

Substitute whatever successes you want.

I have never heard such nonsense!

Mike (Dr. Beav)

[BCOURTNEY]

Too much eating of chi-square without corresponding F-distributions, and exploring non parametric methods with data that is normally distributed .. mmm.

That's ok. A bad case of the N's is nothing compared to convincing some colleagues that confidence values are NOT probabilities, and they had doctorates in math. There are times when strangeness simply appears.

Mike (Dr. Beav)

[snoadog]

Okay I have been restraining myself long enough.

BCOURTNEY. In post 35 you cite the Binomial Formula:

n= trials, w = wins, p = prob
[n!/(w! * (n-w)!] * (p^w) * ((1-p)^(n-w))

In post 54 you state about it:

The formula cited is correct. Wins occurring in a row inside any N number of trials.

This statement is wrong, wrong, wrong. Go crack a stats 101 book just read the parts on combinations. It is clear, you are messed up about combination theory. If you know a bunch of statistical jargon but do not understand the binomial thereom what am I to think? And another thing, please work on making yours post intelligible

Are there more of you? Get them to put their money in the pools. Hurry!

[TrifectaMike]

Too much eating of chi-square without corresponding F-distributions, and exploring non parametric methods with data that is normally distributed .. mmm.

This is a case of "words gone wild". I suggest you either read my posts carefully or take snoadog's advice...learn a bit of combination theory. Then progress to binomial coefficients, and then on to the binomial theorem.

It might do you some good.

Mike (Dr. Beav)

[BCOURTNEY]

Mike, this is words gone wild. I think that I was being overly insistant that the non reduced form be used even in the case that the inr a row did equal n. My fear was that others would not understand the impact of additional trials and leave terms from their calculations. I liked the chi square post. F measure is important and AIC for model selection, logit boosting is yesterdays news.



This is a case of "words gone wild". I suggest you either read my posts carefully or take snoadog's advice...learn a bit of combination theory. Then progress to binomial coefficients, and then on to the binomial theorem.

It might do you some good.

Mike (Dr. Beav)

[mamaluke]

You had to make one show pick daily at Acqueduct Race track

Jon Hardoon from the Rogazin sheets won the last man standing daily show betting contest where there was going to be one winner.

Entry fee was $2 and contestants had to make one show bet every day at Acqueduct.A loss and you were eliminated

I think this was in or about 2006 and Jon Hardoon finally won after about 120 show winners

I dont remember the payout but I think it was over $10,000

He sure deserved it

[TrifectaMike]

Mike, this is words gone wild. I think that I was being overly insistant that the non reduced form be used even in the case that the inr a row did equal n. My fear was that others would not understand the impact of additional trials and leave terms from their calculations. I liked the chi square post. F measure is important and AIC for model selection, logit boosting is yesterdays news.

Ahh...now I understand your references.

Thank you.

Mike (Dr. Beav)