What a 60 grade actually means

Every prospect list you have ever read is written on the same scale. 50 is average, every ten points is one standard deviation, and 20 and 80 are the ends. A 60 hit tool is roughly one standard deviation better than the average professional; an 80 is the sort of thing two or three players in the sport have. The scale is old, it is universal, and it is why “60 power” communicates instantly to anyone who has read a scouting report.

It is also, as normally used, a projection — and that is the part worth understanding before you spend money on the back of one.

A scout's 60 and a measured 60 are different claims

When a scouting service writes 60 power, it is a statement about what the player will become: a human being watched the swing, formed a view about the body and the approach, and projected the tool forward to the major leagues. It is a forecast, and forecasts about twenty-year-olds are hard.

The grades on this site are a different thing entirely. They are a re-expression of a percentile against the player's real same-level cohort right now — not a projection of what he becomes. The transform is the normal quantile, which is the definition of the scale rather than a model of anything, so the grade adds no information to the percentile it came from. It just says it in the language the hobby already speaks.

Both claims are useful. They are not interchangeable, and a card priced as though a measured 65 were a projected 65 has been priced on a misreading.

The translation, in real numbers

Against a cohort of 120 same-level players, the percentile a player occupies converts like this:

Notice how quickly it compresses at the top. The distance from the 50th percentile to the 84th is ten points of grade; the distance from the 95th to the 97.5th is four. That is not a quirk of the implementation, it is what a normal distribution looks like — and it is why the gap between a 55 and a 60 is a much smaller achievement than the gap between a 70 and a 75.

Being the best in the room is not an 80

Here is the part almost nothing else does properly. The best player in any cohort sits at the 100th percentile, and the normal quantile of 1.0 is infinite. Something has to give, and the obvious fix — clamp it to 80 — would hand the top of a 35-man complex-league group the same grade as the top of a 400-player full-season level. Those are not the same accomplishment.

So the percentile is pulled in by half a cohort step before the transform, and the reachable ceiling then rises with the size of the pool the player actually beat:

Beating twenty-nine players is not evidence of an 80. It is evidence of a 71, which is still an excellent number and an honest one. When you see a huge grade attached to a player at a low level with a small cohort, that ceiling is doing work you should know about.

Below thirty peers there is no grade at all

A percentile is a real measurement with as few as five peers. A grade is not, and this site refuses to print one below thirty. The reasoning is precision: in a five-man pool, one playeris twenty percentile points, so the difference between a 45 and a 65 is one peer having a better fortnight.

“Better than 71% of nine players” is an honest sentence. “63 grade” is the same information wearing a uniform it has not earned, and it invites a reader to compare it against a 63 built on four hundred peers as though the two were the same. So the percentile still appears and the grade simply does not — a refusal, not a gap.

What this means when you are buying

• Ask what the grade is measured against. A 60 at a level with four hundred players and a 60 at a level with thirty-five are different statements about the world.
• A measured grade moves. It describes production now, so it changes as the season does. A card bought on a 68 that settles into a 52 was bought on a snapshot, not a projection — which is a different mistake from being wrong about the player.
• The compression at the top cuts both ways. A player sliding from a 70 to a 63 has fallen from the 97.5th percentile to the 90th, which sounds far less dramatic and is the more accurate description.

What it cannot tell you

It cannot see a swing, a body, an arm action or a temperament, and those are what a scout is paid for. It cannot project: nothing here says a 60 today becomes a 60 in the major leagues, and the honest base rates for that are covered in prospect odds and why they are guesses. And it cannot tell you a card is going to go up — a grade is a measurement of a player, and the price is a separate market with its own supply.