David
Feel free to sue statistics ay way you choose, most people do
Fact is I will go to my grave not knowing if the signature is 100% authentic
I was pretty much there with the ball authenticity, and my ??? was whether Ruth was a "clubhouse" or not. A somewhat simpler problem to solve in my minds, as a good forgery is likely to have the basic characteristics of the original, but a clubhouse has specific "features" which may not be common and which I was trying to identify. I was looking for specifics matches to a "known" good signature, and trying to maximize the "matches" This is why I paid attention to the horizontal, the straight b, t, the epsilon e, the loops and teardrops etc. Not trying to prove A= Truth, but basically following the "zeroth Law of Thermodynamics" (If A=B and B=C, then A=C) so my work was not to establish he entire spectrum of Ruth, just the 1929.
As luck has it, there is a lot of common details on bigtrain, which, also helps me "zero in" on the "thin" areas of mine, so i do get some benefit from evaluating the better signature. If you have the time, pls compare one to the other and you will see a high degree of correlation.
I'll save the confidence level for a future date, suffice to say I like to use a Weibull distribution with different weigh factors for experts and non experts (hence JSA, PSA get a higher weigh) but, it is very helpful to have an extra dozen knowledgeable people contribute. After all, the future liquidity market will not be driven by "truth", but by the perception of the truth. If everyone believes an outcome to be the truth, it will become the truth....
As you may know by now, it is hard to find 2 PhDs to agree on anything, and, when they do, thy will agree for different reasons, just the nature of the beast.
__________________
Bernie
|