Handwriting analysis: pure hooey or just 67% hooey?

(If this is in the wrong forum, please move.)

The idea came to me while watching a handwriting expert on pawn stars. They looked at a couple of autographs and claimed they were fakes based on comparisons to known samples of the celebrity’s actual signature. So, i thought, if i ever became famous, maybe i should change my signature at say 40 years of age, then 50, then 60, etc so that my autograph from certain eras would be more or less valuable because of rarity.

So, i sat down and wrote my name 6 times and discovered a huge problem: my own handwriting doesn’t match. Only two of the six samples looked even remotely similar. Based on what the handwriting experts on tv said, I’m actually forging my own signature!

So, my question: sign your name ten times. How many of them match?

I repeated the experiment. Out of ten signatures, 2 pairs of 2 last names looked very similar, but when first name was included, none of them matched.

My signature isn’t exactly the same, but I recognize it from 25 years ago.

I thought this was going to be about analyzing somebody’s personality through a handwriting sample: that I feel is pure hooey (or at least 99% hooey).

But comparing handwriting samples for authenticity? I think that’s pretty accurate. The OP was probably self-conscious when he wrote his samples and that exaggerated the differences. But I think normal handwriting falls into a narrow range because it’s an unconscious habit. So one sample would resemble another.

Well, the person on tv talked about angles of letters, sizes of loops, shapes of certain letters, and heights based on relative position. By that criteria, none of my samples matched.

There are forensic standards published for graphanalysis, the kind of comparisons used in court.

Graphology, on the other hand - determining personality types and so on - is pure woo.

I don’t think it can all be hooey (except for the personality analysis side). I can easily tell which of several colleagues wrote a note in a logbook we all use from just the handwriting. Of course, no one is trying to imitate another’s handwriting, AFAIK. I also know from experience that I can’t credibly imitate another’s handwriting. Maybe, at a stretch, I could do a good enough job to pass a casual glance, but not a side-by-side comparison. In fact, I could probably tell a forgery by someone of my own skill from the handwriting of someone I’m familiar with even without a direct comparison. I’m sure that with practice, anyone could become much better at imitating handwriting that I am. But if the experts are claiming they can never be fooled, they’re idiots (or think their audience are). OTOH, I’m sure that with practice, anyone could get better than I am at spotting forgeries. How the best forgery spotters compare to the best forgers, I have no idea. It wouldn’t surprise me at all if the experts are vastly overconfident of their own ability to detect forgery, but that’s not the same as saying it can’t be done at all.

I just did 4 signatures. There are some differences, but far more similarities. (I’m not sure how to rate that on the scale of the poll!)

I think if I asked someone else to forge my sig next to the four I did, you would easily tell the difference, even though mine aren’t exactly identical. I sometimes write my SO’s name in cards, and I sorta try to do his writing. Even though I think I technically do some bits the way he does, it looks nothing like his writing.

I’m sure that would be different for an experienced forger.

Hmm, now I’m interested. I might go and ask him to forge my signature, see how he does. I’ll report back. I know I won’t be able to do his, it’s really pretty and swirly and complicated, like my papa’s. Wonder what that says about me…

Heh i think I’ve written my name like a hundred times since posting this…

I think i can see what the handwriting expert was talking about. Even though all my signatures are different, certain things are remarkably consistent. I think with enough samples, I could reliably identify about six things I do when I sign my name that never change.

I wonder if this is going to change over the years? I mean, in 50 years will handwriting comparisons have any value? Most people don’t write things by hand much anymore compared to pre-email days, and certainly not compared to pre-telephone days. If you do something day in and day out you get into certain habits and patterns - I don’t think most people handwrite much on a daily basis, not even their signature. Certain professions require it of course, but in many cases even those have signature stamps available.

How often in a week do you actually sign your name on a piece of paper? Signing credit card machines doesn’t count - those never track well.

I have come to think a lot of this is highly dependent on the person being trained in writing with pens.

I grew up typing. Right from the beginning, I had computers - old PC XT’s back in the 80’s and so forth. I’ve never learned penmanship. Holding a pen/pencil remains awkward to me. My writing with a pen or pencil looks like the scrawling of a child. When I am called upon to sign things, even on the rare occasions when I bother to put effort into it, I can find few, if any similarities between different instances of my ‘signature’. Most of the time I don’t even bother; my ‘signature’ is a funny scribble and that’s pretty much it.

Therefore, I suspect handwriting analysis works fine…for people who have trained their hands and muscles to do writing regularly. They certainly gain patterns and habits within their writing that can’t be avoided. But for someone like me, who picks up a pen maybe once or twice a month, only when absolutely needed, and never spent years doing physical writing with a pen, I suspect I never built up those motor reflexes, patterns, and habits.

Several times a week–contracts, checks, and paper credit card slips usually. The main difference between my signature now and 25 years ago is that it’s gotten more and more simplified. That is, the letters after the initial capitals of my first and last name have become more gentures indicating the general shape of the letters within. You can maybe surmise that my first name is “Pete” or “Peter,” but my last name is indecipherable.

But when I do write out my name semi-legibly so that each letter is formed, it still ls recognizable as “my signature” as compared with my hasty signature. They both have very identifiable traits.

When I was a kid I thought “I’m going to grow up to be an artist someday, so I had better perfect my signature. Every artist has a good or memorable signature.”

So now I have a signature that reads very similar every time as I practised it a lot. It only looks different when I feel rushed (signing a cc receipt for instance). I imagine people who do not practice their signature and do not sign things often will have an inconsistent signature. Celebrities need to sign a lot of stuff, so their signatures will quickly become consistent.

Usually part of what determines a forgery is that the person who learns to forge a signature has to “learn” the loops and doesn’t do them naturally. So the pressure will be different at different parts of the letters - usually a more uniform pressure for the forgery because the forger is going slow and trying to match, whereas the natural signature will have varied pressure from just flowing out of a person’s hand.

I had eight out of ten, but I am a practiced calligrapher, so I might pay more attention when I’m signing something.

12.73% hooey.

Practice does seem to have an effect. Also another thing i found was hand position. My signatures became more consistent when i moved the paper up instead of my hand down.

Edit: side note: the key is consistency. Even if i made my signature smaller or bigger, i probably would see the same things. For example, the H in my name, i always make the left bar shorter than the right bar, no matter what, across all different types of signatures.

If you want to extend the sample rate, I can look at signatures from various points in my life, and they changed almost entirely several times from a childhood scrawl, to an almost copperplate and back to an adult careless scrawl.

Signatures vary, but there are certain elements in common to all of them. Mine are always slightly different, but it’s pretty clear that they’re all my signature.

Someone who has a valuable autograph likely signs their name way more often than average though. So for the case of the experts on Pawn Stars it might be relatively easy to tell if a signature was actually done by the famous person in questions.

67% of the time, it works every time.

I think it’s reasonable that some sort of computer imaging could identify someone’s signatures, but I don’t know about all the “the rakish angle of the “i” and the violent dot means that this guy likes to screw cantaloupes” business. That stuff seems like voodoo and/or woo to me.