A grade is a summary of two things. A price is a summary of everything else.

This is the cleanest measurement on this index, because everything anybody normally blames for a price difference is held constant.

The pair

Lots 80550 and 80551 in the same Heritage auction on 25 August 2024.

Both described as a 1960 Ty Cobb single-signed baseball gifted to a Coca-Cola executive. Both graded PSA/DNA NM-MT 8. Both dated in Cobb's own hand.

Not the same day in different sales. Not the same grade on different objects. Consecutive lots, catalogued in identical words, in front of the same bidders, within a few minutes of each other.

The first made $21,600. The second made $13,200. A difference of sixty-four per cent.

It is not a one-off

Two other signers on this index show the same thing.

Roberto Clemente has three signed 1955 Topps rookies in the recorded results, every one carrying an identical label: card graded Authentic, autograph graded 8. They sold for $33,600, $48,000 and $75,000. The top is more than twice the bottom.

Gale Sayers has two 1966 Philadelphia rookies, both graded PSA VG-EX 4 on the card and PSA/DNA Auto 10 on the signature. Identical on both counts. They made $1,952 and $900.

Three signers. Three sports. Three price levels, from under a thousand dollars to over twenty thousand. Same result every time.

The hobby has a name for this

None of it is a scandal, and it is not a criticism of the graders. It is a well-understood thing called within-grade variance, and PSA says so themselves.

Their own article on pricing notes that there are still strength variances within each grade, and that a high-end example in a given grade may carry a premium for exactly that reason.

A grade is a band, not a point. Two cards can sit at opposite ends of the same band, and the label cannot tell you which end.

Collectors describe the effect as largest at the bottom of the scale. A PSA 1 might be a card that is mint apart from a pinhole, or it might be a wreck. The number is the same.

Somebody built a business on it

The clearest proof that this is real rather than noise is that a company now exists to address it.

MBA does not regrade anything. It reviews a card inside its existing holder and applies a second-opinion sticker to the case, marking examples it judges to be at the top of their grade.

A stickered PSA 10 has sold for around thirty per cent more than an unstickered one.

And the honest analysis of that result is worth repeating, because it refuses the easy conclusion. One reading is that the sticker added the value. The other is that the underlying card was already the stronger of the two and the sticker merely made that visible. Both can be true at once, and the market does not distinguish between them.

What is not in doubt is the premise: buyers will pay materially different sums for two objects the grading label describes identically.

Why it is worse on a signature

Everything above comes from the card world. Signed material has a problem the card literature does not address, and it is the reason this index keeps running into it.

There are two grades and neither of them is about the thing you are buying.

The card grade scores the cardboard. The autograph grade scores how the signature was written and how well it survived. Between them they say nothing about:

Where on the object the signature sits. How boldly it was written. Whether it crosses the subject's face or the sky above it. How it photographs. Whether the ink has that particular density that reads well across a room.

Those are the things that separate two otherwise identical items, and there is no field on any holder that records any of them.

On Cobb's two baseballs, one of them presumably looked better than the other in the catalogue. That is very likely the entire explanation, and nobody wrote it down.

What else moves it

Three factors that have nothing to do with the object at all.

Who is in the room. An auction price is the second-highest bid plus an increment. Two determined bidders produce one number and one determined bidder produces a different one. On a thin market that is most of the variance.

Order. There is a reasonable argument that a buyer who wants one Cobb baseball and secures it at lot 80550 does not bid on lot 80551. The second lot faces a smaller field than the first by definition.

And the photograph. Most bidders never see the item. They see a catalogue image, and a signature that photographs well is worth more than one that does not, irrespective of what a grader concluded with the object in hand.

What to do with this

Three things.

Stop treating a grade as a price. It is one input. On the evidence here it explains rather less of the price than most buyers assume, and on signed material it explains the least.

Look at the picture, not the label. The variables that separated the Cobb baseballs were visible in the catalogue and absent from the holder. That is true of almost every signed item.

And read a single comparable sale as one bidding contest, not a market level. If two identical objects can land sixty-four per cent apart in the same room on the same afternoon, then one past result tells you considerably less than it appears to.

The thing worth sitting with

We publish grades on every page of this index because they are the best shared language the hobby has, and because an item nobody has certified is a different proposition entirely.

But the Cobb pair is a useful corrective, and we would rather show it than hide it.

Two baseballs signed by the same man on the same day, given to the same person, graded identically by the same company, sold consecutively by the same auctioneer to the same room.

Sixty-four per cent apart.

Whatever made the difference is not printed on either holder, and nobody involved recorded it.

This index records documented sales for Cobb, Clemente and Sayers with both grades shown separately, so comparisons like this one can be made rather than assumed. You can read how we work the numbers out.

Sources