Oh my!...  There's nothing like a "physicist" with his dander up, is there!  Unfortunately, Mr. Physicist has already exposed himself in his first paragraph as being a person who isn't even qualified to be commenting on this web page.  His statement about the "pedestrian little introduction to statistics is embarrassing enough" shows his lack of intelligence, since that little "introduction to statistics" is rock solid mathematically.  Any high school kid could have told Mr. Physicist that.  Maybe he should have asked one that also passed a freshman physics course as a refresher for his claimed profession.

One of the reasons that working scientists (and many others) do not participate in public debate is that we are constantly confronted by nonsense like this. When we label it as valueless, we are immediately challenged to “prove” that it is not true………I would rather wade in a cesspool, so boots on and here we go…..

More insight into the mentality of Mr. Physicist - who believes his opinion is gospel and should not be "challenged to prove that it is not true".  Well he came to the wrong place because we clearly stated on this web page that we have no regard for those who claim their "opinion" has any value.  Apparently Mr. Physicist has never heard the expression that "opinions are like assholes - everybody has one".  We offer only solid calculations that can be verified by anyone smart enough to punch the proper buttons on a calculator.  It looks like we are about to find out if Mr. Physicist is that smart.

Let’s get the diatribe over with first. If you are convinced by the arguments in this document you clearly do not understand even the basic concepts of statistics. Even if you don’t know ANYTHING about statistics (and lots of well educated people don’t), you should read through this and say to yourself, “This is garbage”. The argument is clearly wrong, on its face. If you don’t reach that conclusion pretty quickly, you are not thinking clearly. I note that several previous posters have already figured out that this is nonsense. If, on the other hand, you actually believe the ridiculous analysis, I probably won’t change your mind because you don’t think clearly enough to be educated (or listened to for that matter).

At this point we have to wonder exactly who it is that does "not understand even the basic concepts of statistics" since Mr. Physicist had a hard time understanding our example of the basic odds of a dice roll and called it "embarrassing" in his first paragraph.  Statistics don't get much simpler then the examples we presented, and we wonder why Mr. Physicist or anybody else thinks interjecting their "opinion" into such basic math would even make the slightest difference.

Now, to the details of just how bad the analysis is. Ziggy and others have already pointed out that without knowing the frequency of occurrence for each possibility (e.g. the different styles for the letter “a”) the analysis is meaningless. The author claims that this is a 3-to-1 probability, but in fact the probability depends on the actual frequency of occurrence. Almost everyone I know mixes the styles of the letter “a” when they print (we have even joked about it when we were reviewing handwritten red-lines to drawings). If most people (say 80%) mix the styles for the letter “a” (exhibit 3) then the probability is not 3-to-1, it is 1.25-to-1. Similarly, switching between block and cursive letters (exhibit 1) and connecting the letters “t” and “h” (exhibit 6) are extremely common characteristics and certainly will not generate anything approaching the 2-to-1 odds the author claims.

Well, it certainly didn't take Mr. Physicist very long to fall flat on his face.  He mentions "frequency of occurrence", but Mr. Physicist apparently isn't smart enough to understand that "frequency"  and "occurrence" are two completely different statistics. 
Here's the flaw in his logic: If we make the statement that the odds it is either raining or snowing in Boulder, Colorado right now are 3-to-1, that results in a factual statistic that can't be denied.  It's either raining, snowing, (or neither of the two) which means there are only 3 possible answers.  However, Mr. Physicist believes those 3:1 odds will change if we take into consideration the fact that it only rained 4 out of 30 days during this same month in Boulder last year.  Sorry Mr. Physicist, but the "frequency of occurrence" has nothing to do with the odds when there are only three possible answers to a given situation.

Mr. Physicist makes the same foolish mistake when he claims that "connecting the letters “t” and “h” are extremely common characteristics and certainly will not generate anything approaching the 2-to-1 odds".  A person writing text either connects the two letters, or they don't.  That's 2-to-1 odds in anybody's book, the same way as a coin toss is only 2:1 odds--heads or tails. 
Mr. Physicist must also believe that making a statement that a coin toss landing on heads "is an extremely common characteristic" can change the odds which are fixed at 2-to-1 and always will be.  We sure hope this guy doesn't do any serious gambling or we'll find him in Las Vegas wondering why he lost his whole paycheck by betting the numbers field on a roulette table.

Not including the frequencies of occurrence is bad enough, but there is a MUCH worse flaw built into the analysis. The author multiplies the various probabilities he generates. This makes the implicit assumption that the probabilities are independent. For example, “exhibits” 2, 6, 7, and 9 all deal with connections between letters. These are NOT independent probabilities and are instead STRONGLY (positively) correlated. That is, people who connect the letters “t” and “e” (exhibit 2) will be MUCH more likely to connect the letter “e” with the following letter (exhibit 7) as well. By treating these “exhibits” as though they are separate and multiplying the probabilities, the author magically generates a 16-to-1 (2x2x2x2) probability for the characteristic “tends to connect certain letter pairs” which is likely to be 25% of the population. This collapses a 16-to-1 probability to a 4-to-1 probability and we have taken the author’s original estimate of 10,000-to-1 down to something like 500-to-1, and we are just getting started with the long list of problems in the analysis.

Here goes Mr. Physicist off on some completely unrelated tangent.  But to be honest, he's not the only one to make the same mistake when they voiced their opinion.  He wants to include some kind of "national possibility study" in his calculations.  That's fine if he wants to do that, but that's a completely different analysis then we presented. His study would be an "average handwriting analysis" and we clearly stated that our web page is not a "handwriting analysis".  Here are two examples where Mr. Physicist and several other people missed the boat:

The statement that "These are NOT independent probabilities and are instead STRONGLY (positively) correlated" may be correct if we were doing a national handwriting study, but we are not.  We are dealing with a very specific writing sample which is printed.  Probably 98% of the ransom note is printed.  The writer of the ransom note attempts to print the context of the note, but slips up on occasion and forms connected script letters that happen to match Patsy Ramsey's normal writing characteristics.  The fact that the ransom note author is attempting to print his/her writing, and winds up connecting letters only by accident, makes each probability independent and not correlated to each other.  Each accident must be judged on its own merits given the intention of the author.

Additionally, let's consider Mr. Physicist's statement that perhaps it "is likely to be 25% of the population" that connects those letters when they write.  The analysis we presented was based strictly on known, hard-core visually verifiable evidence.  Mr. Physicist wants to introduce an unknown variable into the equation which would only result in a variable answer instead of a solid statistic.  If Mr. Physicist wants to play the game of "averages" fairly then he would have to include another statistic in his calculation.  That would be how many of his "25% of the population" who write with connected letters woke up with a dead child in their basement and a ransom note with those exemplars on the floor.   Those odds would cause a non-scientific statistical analysis, the type Mr. Physicist desires, to show the chance anyone other than Patsy Ramsey wrote the note are billions to one.

Finally, the “powerball” discussion at the end is too misguided to do anything but laugh (or maybe cry). Bluntly….the factor of 5281::1 is irrelevant to the discussion of Patsy Ramsey. She did live in Boulder, and her child did die……it doesn’t have a probability of 5281::1, it has a probability of 1::1, it is established fact.

What a shame... Now we have to add "reading comprehension" to Mr. Physicist's list of failures.  We clearly stated that the 5281:1 odds were that an intruder broke into the home and killed the child, not that a child would just happen to die in Boulder.  Surely many children have died in Boulder over the years, possibly from diseases and car accidents.  Those deaths have nothing to do with this analysis, but it's not surprising he brought them up - since nothing Mr. Physicist has written yet has anything to do with this analysis.

Now for a little probability discussion of my own. There are only two possibilities. The first is that the author of this nonsense actually believes it…..that possibility is just too sad to think about, people like that belong in a sheltered living environment, not running loose anywhere near a computer. The second (and I believe more likely) possibility is that the author knows better and is using “science” as a magic cloak to convince us of the rightness of his long-held beliefs. The scarey thing is that he believes that his pathetically juvenile “introduction” to probability makes him seem competent and helps to establish his credibility. Rather, the laboured discussion suggests to me that he has dredged up virtually everything he knows about probability only to deliver a discourse that should embarrass a bright high school sophomore.

What a great paragraph!  Mr. Physicist thinks our basic lesson in statistics is "pathetically juvenile" and would "embarrass a bright high school sophomore".  That's really funny coming from a guy who can't even understand that the odds on a simple coin toss are solidly fixed at 2-to-1 and a dice roll at 6-to-1.  If Mr. Physicist thinks that berating us will make his opinion sound important and cover the gross errors in his logic he is sadly mistaken.  His previous statements have shown that he is doesn't have a clue as to what our analysis is all about or how it relates to solid mathematical principals.

If the discussion interests you, go read a good book on statistics, just get a copy of whatever the local university is using in their Introduction to Statistics course. Then go check out every book on handwriting analysis in your local library….if there aren’t any, ask the librarian to help you find one and ask her to order it….I promise the librarian will be tickled pink that someone actually wants the library to order a book for them. There is an established body of work in handwriting analysis but the disagreement between the experts would seem to indicate that it is not a “hard” science.

Do those books explain what the odds are on a coin toss and a dice roll, Mr. Physicist, or did you miss that edition?  And why are you recommending people get a book on "handwriting analysis" when we clearly stated that is not what this web page is about?  It's now very obvious that Mr. Physicist's inability to understand this web page is completely due to a reading comprehension problem.

Finally, I have done my duty, I have stood up and labeled nonsense for what it is. I don’t have any intention of discussing the matter further except to say that this looks like science being used (pathetically badly) in service of a personal agenda. That offends me at the very core of my being and I have tried to do something about it. That is my public duty as a scientist.

Mr. Tipper
 

If Mr. Physicist Tipper's "duty" was to make himself look like a fool and cast a real shadow on his profession, there is no doubt he succeeded.   If all physicists have the same problem with reading comprehension and understanding the basic statistics of mathematical odds, it's a wonder the world is still in existence   We have to wonder if the name "Mr. Tipper" isn't some sort of inside joke.  Based on his inability to grasp even the most basic mathematical equations we can only deduce that he couldn't possibly be a "tipper".  This person would find it impossible to look at a service bill and figure out what the 15% gratuity would be.  First he would have to perform a nation-wide study on the tipping habits of people in his same age, race, and eye color catagory.

One last thing... Remember Mr. Physicist Tipper mentioned another person named "Ziggy" whose opinion he agreed with?  Well, below is a post by "Ziggy", who like Mr. Physicist Tipper, is a little too numb in the nuts department.  Ziggy also believes a coin toss has odds different than 2-to-1:

Posted by Ziggy on 03-Jan-02 at 09:41 AM:
 I think if you will do some research you will find that studies have been done concerning the coin toss and when a quarter is tossed, it most often lands on heads. Doesn't that make the odds different then? You have a better chance winning if you call heads. It was probably a government grant or something that funded that study, but it was interesting...