Featured An examination of the counterfeit slab epidemic. Scope and advice.

You are using the wrong analogy. Dice don't work in this instance because no matter how many times I roll them, 100 or a million, there will always be 6 probable outcomes.
In the case of the fake slabs. The number of outcomes is diminished everytime I choose a slab.

 

Log in or Sign up to hide this ad.

The lottery analogy doesn't work either because everytime a ticket is sold, the number of possible outcomes increases. Totally opposite of what is happening with the coins.

 

Even if this were the case (which I have explained why it isn’t) you would need around 800 slabs to be ensured that that just one is a fake.

At a large coin show, there might be 500 tables. Say each table has 200 slabbed coins. That is 10000 slabbed coins (gross underestimate). If just 2 were fake, that would make it a 0.02% chance of a fake slab. You would need to randomly buy at least half of the slabs at the show (5000 in this case) at random to be statistically certain to get a fake slab.

 

12.5% of 1/100 is far from certain.

 

If you had a pile of slabs of which 0.125% were fake, and they were evenly distributed, you would need to pick out 800 slabs to be reasonably certain you picked out a fake

 

Yes exactly. Every time I pick out a slab, if it isn't a fake, the chances of the next one I pick being fake increases. If I pick all the real slabs then every choice I make after that is an absolute certain that it will be fake.

 

No... this assumes an even distribution and that they are mutually inclusive. There is no guarantee that there will be a fake in 800, there is merely a probability of .125% that you'll get one. That is an extremely low percentage. That means that you can buy slabs with a 99.875% certainty that they're real. We've already agreed that this is a guess number anyway, but the point I'm trying to make with the statistics is that it's not nearly the threat the OP tried to make it sound like and we don't need to be telling newcomers not to buy PCGS.

 

I'm working on Global Cooling, the priority/fad/scam of the 1970's.

 

Let's fiddle while Rome burns.

 

There is no absolute guarantee unless all are chosen correct. But the assumption that probability never approaches 1/1 is not correct in this instance. If I were to choose one and put it back before choosing another, then the chances would never approach 1/1

 

The number of slabs you're picking is negligible compared to the total number of slabs, so there's no significant difference between "with replacement" and "without replacement".

 

That's an assumption that I am the only one picking. Devide the number of available slabs by the number of collectors picking through them.

 

It might not be the right one, but not for the reason you provided. The jackpot is always the same chance for each and every ticket, no matter how many are sold.

If there is a finite % of slabs that are assumed to be fake, that is constant, then the probability for having a fake in a pool of 100 is the inverse of the probability of pulling a real one 100 times in a row. ASSUMING a perfect random drawing, all the good and bad ones have an equal chance of being pulled out of the pool.

So the inverse of .99875 x .99875 times itself 98 more times? I am just thinking out loud and could be complete off base...

 

I don't think you're looking at the statistical problem correctly, but stats isn't my strongest subject, so I'm not the one to do the professorial thing here. (And I'm in a rush...)

 

The possibility of some person winning, increases with the number of people in the game. The other possiblity is that no one wins which decreases with every sequence of numbers chosen or each person entering the game. Your individual chances per ticket doesn't change because there is only one drawing.

 

This thread has more tangents than a circle

 

12.5% of 1% is .00125 or .125%. He is saying 12.5% chance that 1 coin in 100 is fake.

I have seen this decimal misunderstanding when using percentages on the radio and in political debates a lot lately. Right or wrong, I blame it on the reliance on calculators rather than an understanding of the math.

Biden meant 150,000 and said 150,000,000. Rush had a guest that said 600,000,000 when he meant 60,000,000.
Bernie doesn't even do calculations so he can keep out of trouble.

Tim calculates death rates using different basis than the CDC. The calculations are good but use different basis and are therefore not compatible as a comparison.

I'm old enough and gifted enough in math that it seams obvious to me, but I recognize things have changed.

Most of these errors occur because decimals are moved too few or too many places while converting from a raw number to percentage. The others are simply measuring different ways which are not comparable.

ie COVID-19 can be calculated with deaths per resolved cases and you get close to 10%. That's scary, but flu is calculated by deaths per estimated total cases. It's about 0.1% or 1 in 1,000. I saw one person say it was deaths per total population, which makes no sense at all.

Because unreported cases and unresolved cases are not known, every estimate has fundamental flaws in determining a basis in calculating the rate in percentage.

So be cautious, but don't be scared to death by those who don't understand what they're reporting.

 

Thank you.

 

Well not exactly. There is no certainty that any are fake or that more than 1 are not fake. It is probable, not certain. And all that is based on a poor estimate of the percentage of fakes in the total population which is highly uncertain.

 

The population of genuine and fake slabs increases every day.