Why nobody can tell you a prospect's odds of reaching MLB
What it would actually take to say “73% chance” honestly, why no public prospect tool has it, and what to use in place of a number nobody can compute.
Key finding
A probability without outcome data and a backtest behind it is a ranking wearing a decimal point — and the decimal point is the dangerous part, because it gets believed.
Somewhere in every prospect discussion a number appears: 73% chance of reaching the majors, or +148% on cards like this. They are compelling, they are easy to repeat, and they are almost always invented. This is what it would take to produce one honestly, and what to do instead.
What a calibrated probability actually requires
“Calibrated” is not a compliment, it is a testable claim: of all the players a model called 70%, about 70% should have made it. Checking that needs three things almost nobody has.
- Outcome data. Thousands of past prospects followed to the end of their careers, not just the famous ones — the forgotten majority is the entire denominator.
- Survival rates by level and age. Published base rates for how often a 19-year-old in High-A reaches the majors at all. Without a base rate there is nothing to calibrate against.
- A backtest. The model run against players whose careers already happened, with the result published — including the years it was wrong.
A number produced without those is not a probability. It is a ranking that has been rescaled to look like one.
Why the decimal point is the dangerous part
“This player is a better bet than that one” is a claim you naturally hold loosely. “73%” is not: two digits of precision read as measurement, and measurements get acted on. The same ranking, expressed as a percentage, will move more money and survive more contradicting evidence — which is exactly backwards, because nothing about the underlying confidence changed when the units did.
The same problem, wearing a dollar sign
“+148% on cards like this” is the return version, and it fails the same test plus one more. It needs a defined universe of “cards like this,” every one of their prices then and now including the ones that went to zero, and a stated holding period. Survivorship alone will produce a triple-digit figure from a market that lost money: the cards nobody lists any more are the losses, and they are invisible.
What to use instead
- An ordering, held as an ordering. Ranked better or worse is a claim the available evidence can support. Treat it as a sort order, not a score out of a hundred.
- Level and age, read directly. Where a club assigned him and how old he is for that level carry more than any headline number — see reading a prospect's stat line.
- Comps, for the price question. Whether the player is good and whether the card is cheap are two separate questions, and the second is answered by sold comps, never by a projection.
- One question for any source quoting a percentage: what was it backtested against, and where is that published? A source that cannot answer is showing you a ranking.
What this site refuses to show, on the record
Basis & limits
What this is built on. The rules and figures this project's own identity and valuation engines enforce, plus the domain research behind them.
Where it stops. This is an explainer, not a study: it carries no sample size and makes no forecast. Figures that move in the real world — grading fees, print runs, marketplace behaviour — can date it; the updated line above marks the last material revision.
Methods are documented on the methodology page; sources and their limits on trust & data sources.
More on whether the player is worth owning
Keep reading
- Reading a prospect's stat line when you're buying the cardThe bridge between the baseball and the card: what a stat line can and cannot tell a buyer, and the context that changes its meaning entirely.
- How to read sold comps without fooling yourselfSold-only, delivered-price, exact-match discipline — the comp-reading rules our valuation engine enforces, written out for humans.
- When does a prospect card peak?The exit side of prospecting: why maximum hype and not maximum performance is the peak, and the signals that say a thesis is over.
- Why the same card sells for $40 and $90The spread between sales is information, not noise — and the median of six recent sales is a different kind of number from the average of two.
- What a 60 grade actually means50 is average and every ten points is a standard deviation — but the size of the crowd a player beat decides how high the number can honestly go.
- What actually moves a prospect card's priceStep changes, not drift — and the harder question of whether the market has already repriced the news you are acting on.
- Will this card ever be worth $1,000?Ask it twice — once with stardom assumed, once without — because the gap between the two answers is the useful part.
See this applied to real prospects: every player page shows live listings, sold evidence and the parallel ladder for one player, and the market board ranks what is mispriced right now.