FAQ
Plain answers, stated so they can stand alone. The current record: 3 hit · 4 miss · 9 open — the ledger has the receipts.
Ask us anything · andy@feverbaseball.com
The basics
- What is Fever Baseball?
- Fever Baseball is a baseball publication built on an instrument called the Future Value Radar (FVR). It finds mispriced baseball talent using batted-ball physics validated on eleven seasons of public MLB play-by-play, publishes every prediction on a public timestamped ledger that resolves by computation, and covers the baseball card market angle for every player it tracks. Everything refreshes nightly, automatically.
- Is it free? What's the catch?
- Free, all of it — the boards, the ledger, the daily Overnight, the player pages, the emails, the daily game. The business model is simple and disclosed: some card listings link to eBay through an affiliate program, labeled right where they appear — and every dollar the paper earns is published in public on The Till. Open books. The record itself will never sit behind a paywall, because a public record you have to pay to audit isn’t a public record.
- How can I support the paper?
- Four ways, all optional, all gathered on the support page — with the open books on The Till. Start your eBay card shopping from the paper’s link — it costs you nothing extra and pays a small commission. Drop something in the tip jar, one-time or monthly. Send bitcoin, if your binder sits next to a wallet. Or simply subscribe and share the site — attention is the currency a young paper needs most. Whatever comes in shows up in the public register, to the penny.
- What's your track record so far?
- The live count is printed on every page: 3 hit, 4 miss, 9 open. The site is young — the first call window closes in August 2026 — so the honest answer is that the record is mostly still being written. That’s exactly why every call was timestamped before the outcomes: when the results arrive, nobody has to take our word for anything. The ledger has all of them, with the date each one gets graded.
- How do I know you won't quietly delete a miss?
- The ledger is append-only: calls are never edited after filing, corrections are appended in public, and misses get stamped on the record with the same rubber stamp as hits. Every edition also lands in subscriber inboxes the night it prints — an archive we couldn’t quietly rewrite if we wanted to. Deleting a miss would cost more than eating it, which is the design, and the Ethos spells out what each of those rules has already cost us.
- Who makes this?
- A lifelong baseball fan and card collector, an engine built from scratch on public play-by-play data, and a stable of AI agents working the night shift — ingesting games, refreshing boards, printing the morning edition. One human sets the standards; the methodology page shows the receipts, including the findings that weren’t flattering. More on the about page.
The engine
- What is a call, and how does it resolve?
- A call is a falsifiable prediction filed with four parts: a metric, a threshold, a window, and a UTC timestamp. When the window closes, the engine computes the metric and stamps HIT or MISS — nobody argues, and nothing is edited after filing. A hit rate with the misses removed is marketing, not measurement. Every call we have filed is here.
- What is EV95?
- EV95 is a hitter’s 95th-percentile exit velocity in mph — how hard he hits the baseball when he really connects. It’s used instead of average exit velocity because top-of-range contact is the skill and mishits dilute the average. It is the most stable metric in the FVR engine (year-over-year correlation of 0.88), which makes it a better forecaster than any batting average. The One Chart shows you why.
- What is the luck gap?
- A hitter’s actual results on contact minus what the physics of his batted balls say those results should have been, adjusted for ballpark. Negative gap: the results are underpaying the contact — breakout shape. Positive gap: the results are borrowing — fade shape. The luck gap tends to close in both directions, which is what makes it a signal — and unlike contact quality, almost none of it survives to next season.
- Why isn't a star like Shohei Ohtani on the boards?
- Because the boards rank mispricing, not talent. A superstar whose stat line already pays him in full for his contact quality has no gap to bet on, so an honest mispricing detector says nothing about him. Stars like that live in the Index — full player pages with a computed explanation of exactly why no board flagged them.
- Why are there no MLB pitchers on any board?
- The stuff model that would rank them is stability-tested but not yet outcome-validated, and the house rule is that no signal ranks a board until it clears the same pre-registered out-of-sample bar the hitting signals did. It hasn’t — so there is still no pitcher board, and big-league arms appear only as tracked-but-ungraded Index pages. What HAS cleared the bar is a different arm signal, the ERA–FIP luck gap (the H-ARMGAP study): we now file regression calls on established arms by hand from it, and they resolve on the ledger like any other. The board rule stays; the calls are new.
- How is the engine validated?
- On eleven seasons of public play-by-play with a pre-registered protocol frozen before the test data was examined — one evaluation, published regardless of outcome. On the held-out seasons (721 qualified batter-seasons), the composite signal separated top from bottom quintiles by +1.15 annualized WAR. The methodology page also discloses a computation bug found during the work, and a null result: clutch hitting is not a repeatable skill.
The cards
- What does any of this have to do with baseball cards?
- A card’s price follows a player’s story, and this paper is in the business of getting to the story early: breakouts, milestone chases, and prospect debuts. Fever Baseball publishes its signals and milestone watches first, in public, with the timestamp to prove it — then links each player’s eBay listings so collectors can act on their own read. How much any of that actually moves a price is a claim we have not measured, and we will not make it on the pages that would profit from it being true.
- Should I buy cards because of one of your calls?
- No — buy cards because you love them. Nothing here is financial advice; collectibles are speculative and can lose value; and our own rule is that we take positions in a player’s cards only after the call is published, never before (rule 08, and the books are open). What the site offers is honest information, early — what you do with it is yours.
The paper & the games
- How often does everything update?
- Nightly, automatically. A pipeline ingests every MLB and Triple-A game, rebuilds the boards, refreshes all 120+ player pages, prints The Overnight with newspaper-style box scores, updates the milestone chases, and emails subscribers. Every page shows its as-of date. If you want to check our arithmetic yourself, the tables are at Open Data.
- What is The Card Back?
- The daily mystery-player game every collector already knows how to play: cover the front, read the stats off the back, name him. 3 guesses, a new card at midnight ET, same card for everyone, streaks saved in your browser, Prospect Issues on Saturdays. Free, and it is here.
- How do I follow along?
- Two free emails: the Sunday Radar Digest (the week’s drift, every call’s countdown, the Chase, the Wire) and The Overnight (last night’s baseball, graded, every morning after games). Sign up on the homepage. There’s also the Discord for arguing the boards, and the site posts on X.
- I found an error — what do I do?
- Tell us: andy@feverbaseball.com. If you’re right, the correction is appended in public and the original stays visible, struck through by honesty rather than deleted. Being auditable is the entire point of the operation.