Twelve checks professional authenticators run, in the order they run them. Read it once before you buy or sell anything signed. Half the bad purchases in the autograph market would be avoided by applying just the first six.
An authentic signature is the product of years of muscle memory. The hand moves quickly and unevenly: faster on the long strokes, slower on the loops. Forgeries, even good ones, show a forced, careful uniformity because the forger is drawing the signature rather than signing it. Look at the long strokes especially. Real ones taper, accelerate, and sometimes skip slightly. Drawn ones are stable end-to-end.
Real signatures vary in pressure across a single stroke. Pen pressure peaks at the start of curves and direction changes, lightens through the middle of long strokes. Forgeries are typically painted at one steady pressure, partly because the forger doesn't trust themselves to vary it, partly because they're tracing. Under angled light or strong magnification, the ink depth on a real signature is uneven. A flat, uniformly dark trace is suspicious.
Where a pen stops, lifts, or pivots, a tiny bead of ink collects. This is one of the most reliable authenticity tells because it's automatic. The signer doesn't think about it. Forgers often skip these dots and pools because they're drawing fast or tracing. Magnify a signature 10x and look at every termination point: the dots on i's, the ends of crossed t's, the tail of the last letter. Their presence is good; their absence is a flag.
Real signatures don't pause. Forgers pause to think about the next stroke. The result is hesitation marks: tiny stutters, doubled lines, visible pen lifts in places where flow should be continuous. The most common location is in transitions between letters where the forger had to re-engage the pen. If you spot a hesitation mark where the writer's hand should have been moving smoothly, that's a problem.
Felt-tip pens didn't exist before the 1960s. Sharpies didn't exist before the 1970s. Ballpoints became common in the late 1940s. If you've got a 1920s autograph in Sharpie, you've got a fake. Same logic applies to paper. Pre-photocopier eras used different paper stocks, and authentic period paper shows specific aging patterns (foxing, brittleness, type-specific yellowing) that modern paper cannot replicate.
An autopen is a machine that holds a real pen and replicates a signature with mechanical precision. Public figures use them for mass mail. The tell: every autopen signature from the same template is identical. Real human signatures vary, even from the same person on the same day. To test, overlay two signatures from the same signer (you can scan and digitally overlay, or just use tracing paper). If they match millimeter for millimeter, it's autopen. Two real signatures from the same person never overlay perfectly.
Throughout history, busy people have had assistants sign for them. This was extremely common for politicians, executives, and celebrities pre-2000. Secretarial signatures aren't forgeries in the criminal sense (they're authorized), but they're also not the actual person, and the market values them at a fraction of the real thing. Spot them by comparing against verified exemplars from the same era and looking for systematic differences in letter formation that wouldn't happen by chance.
Most signers' signatures evolve over their career. Compare your piece against verified examples from the same approximate time period. A Babe Ruth signature from 1925 looks meaningfully different from one in 1947. A Beatles signature in 1962 looks different from 1969. If the signature on your piece matches the signer's late-career style but the item is dated early-career (or vice versa), something's off.
Where did this come from? Authentic items usually have a story: bought at a charity auction in 1987, signed in person at a card show in 2008, inherited from a collector grandparent. A signature with no provenance isn't automatically fake, but provenance dramatically affects both authentication confidence and value. Original receipts, photographs from the signing event, signed letters of authenticity from the original buyer. All of it matters.
Certificates of Authenticity are only as good as the issuer. A COA from PSA, JSA, or Beckett carries real market weight because their authentication teams have track records. A COA from "World Signatures Authentication LLC" you've never heard of is essentially worthless and is often a marker of a deliberate fake: bad actors print fake COAs to make fakes look legitimate. Look up the issuer. If they're not in the top tier of recognized authenticators, treat the COA as decorative, not evidential.
Look at what was signed on. Modern photo paper has specific characteristics that didn't exist before certain years. Trading card stock changes over time. Album jackets have known printing dates that constrain when they could have been signed. If a piece claims to be a 1955 photo signed in 1955, but the photo paper is clearly a 1980s reprint, the signature can't be earlier than the 1980s. This is one of the highest-leverage checks and one that AI scan apps cannot do.
For anything you think is worth real money (generally $200+ for sports, $500+ for entertainment, $1,000+ for historical), submit it to PSA, JSA, or Beckett. The fee is $20 to $50 per item, the turnaround is a few weeks, and a certified item from any of those three commands a 30-100% premium in resale. They get things wrong sometimes, but their batting average is dramatically higher than yours or any app's, and the market priced their stamp in years ago.
The app runs the first ten checks automatically. Use it as a first-pass filter, then apply the human checks on anything that clears.
