More Weird Stats @AQU !!!
 

Found a truly weird stat at Aqueduct on a Sloppy surface in Claiming races on the Dirt

It seems horses that get tagged at 5/2 in the Morning Line in Claiming races on a Sloppy surface have been 0 for their last 13 going back to January 10th 2016

Interesting to note, the last time a 5/2 Morning Line Claimer won on a Sloppy surface was on December 2nd, 2015 exactly 2 years ago on this very day!!!

http://www.insidethenumbers.net/imag...aqubadstrk.png

Let's see if one of these 3 can break the jinx today!!!
My theory of 10 got smashed here but let's see if the streak continues 2 years to the day!!!

I am especially interested in Rodriguez's Horse #1 Transistor let's see if he can get the win today!!!

Racing was cancelled on December 2nd due to weather, the forecast predicts rain this coming Saturday

It should be interesting to see what the 5/2 Morning Line Claimers do should the Surface be listed as Sloppy!!!

Let's keep a sharp eye on this situation!!!

Fooled By Randomness
 

Please check out this book/concept.

If a six-sided fair die is thrown 60 times, except for ties, one number will have been thrown the most often. Let say the number "4" was thrown 9 times and 9 times is the greatest number of times thrown of all the numbers "1" through "6".

For the next 60 throws, would you expect the number of "4"s that appear to be: (1) less than 6 because now numbers other than "4" are due; (2) 6 times because that would be the expected value for a fair die; or (3) more than 6 because, even though the die is fair, "4" has been thrown more than its fair share thus far?

Fooled By Randomness
 

My post above should have read as follows... apologies!

Hey Wizard

I get it gambler's fallacy...there is no such thing as something being due to happen just because it happened a number of times consecutively or otherwise. It's the same probability no matter how many attempts are made.

Especially in the case of a fair coin or die, just because you get 15 heads in a row or in the example 14 number 4's in a row doesn't mean the other numbers are due or the 4 has a better chance because it is coming out more so than the other numbers

I like to track the short term events and in the short term I like to see what the trends are and go with or against what's trending at the time.

Like going with a hot shooter on a basketball team until he starts to miss or a running back that is destroying the defense until they figure out a way to adjust

That's what I like to look at so I started tracking the trends

How could you possibly know whether to go with or against a "trend"?

I don't know it's just a guess and I hope that I guess right based on what's happening

So in the case of the 4's in your example. There's no way I could possibly know ahead of time if there will be any more 4's that will occur but if I see that every time there are a big string of 4's they get cold after subsequent attempts I might go against the trend.

Or, if I have seen that 4's have been cold for a long stretch and they are starting to come out maybe I will take a chance on another 4 since they are starting to win.

In your example it's based on a "fair" die. However, we all know that when it comes to gambling "fair" gets very slippery. So, just in case it's not fair, if I have information that the 4's are winning more often than the other numbers it works out to an advantage.

Indeed

I don't know of any casinos that publish the result sets of the information related to their craps games or any other table gaming event. They don't want to take a chance on anyone knowing that the 4 hit 14 times regardless of any coincidence.

You better believe, they know exactly what the metrics are related to their games and they don't want anyone to do the due diligence and have that type of information readily available at their fingertips to make better guesses.

Why would that information do anyone any good for subsequent gamblers?

Maybe that information exposes them to something they don't want the public to know

I mean if the information is that inconsequential then why not publish it?

Why not make it available for everyone to see?

I think they don't want the public armed with any information that might hurt their pockets coincidence or not

That's my take

I genuinely mean no disrespect, but, IMHO, you just seem to be restating The Gamblers' Fallacy in various different guises as if it were actually True.

Hey Wiz,

No issue healthy discussion.

I'm just saying that the Gambler's Fallacy is based on a fair game (i.e. evenly weighted coin, dice not loaded, etc.)

Take the NY Pick3 State Lottery for example, you can go on the NY State lottery website download all of the numbers that came out and if the number 3 appeared more times than in others you can't make a case and say that it will continue to happen or that the other numbers are due.

There is complete transparency, the information is published freely online and they don't care that people know about it because the information won't increase anyone's chances of winning.

In those instances, one outcome has no effect on the other. The outcome of 14 number 4's in a row doesn't mean the 4 has more of a chance than the other numbers on the die.

What I'm saying, is that just in case the games aren't fair there may very well be a reason why the 4 is coming out more times than others maybe the dice are loaded who knows.

My point is that the most casinos as well as other gambling institutions are not transparent about their data and it makes me wonder about the fairness of their games

To that end, I track the data to see where it leads and if the game is not fair I believe that knowing the data gives you an advantage more so over someone that doesn't.

How could you know the difference between "not fair" versus randomness? Are you performing any statistical tests? Chi-Squared?

Hey Wiz

You bring up a good point I don’t have any statistical data regarding casino games but I do have a healthy amount of results data from NYC tracks

As far as I can know, a Chi Squared test checks to see if two variables are related in some population so education levels and incarceration or education levels and marital status so I’m not sure how that would apply to my data.

However, I do remember that there is some test that measures randomness I forget which one that is, I could perform a randomness test against my data and see what I get from the data.

Let me look into that.