Nothing moves a baseball line like a change at starting pitcher. Not weather, not lineups, not the money. If you only learn to watch one thing on a morning board, watch the probables.
A "probable" is a projection, not a promise
Books post their opening prices against the listed starters. That listing is a team's stated intention, several hours or a day out, and it changes for ordinary reasons: an injury, an illness, a rainout the day before that shifts the whole rotation, a manager deciding to skip a turn.
When a starter is scratched, the line doesn't adjust — it gets re-priced from scratch. Moves of 30–60¢ are normal. A true ace being replaced by a spot starter can be worth a full 100¢ by itself.
Most books void the bet — which is the point
By default, most sportsbooks offer what's called listed pitchers protection: your moneyline bet is void and refunded if either listed starter doesn't start. Rules vary by book, and it's worth reading yours rather than assuming.
That protection is why an early bet on a great pitcher rarely leaves you stuck with a bad one. But it has a second consequence people miss: because those bets void, the market effectively re-opens the moment the change is announced. The new price isn't a nudge off the old one. It's a fresh number reflecting the new starter, posted into a market where the previous action has been erased.
Why the move is rarely symmetric
A starter change alters two things at once, and they don't move together.
It changes the side — who is more likely to win. It also changes the total, because a replacement starter usually means more expected runs and often a longer outing for the bullpen. A downgrade at one starter can push the moneyline 40 cents and the total half a run in the same few minutes.
The handedness matters too. Our packets record each starter's throwing hand alongside the name, because a swap from a left-hander to a right-hander re-prices the opposing lineup's platoon advantages even when the two pitchers are otherwise comparable.
We record probable starters and their handedness from the MLB Stats API at generation time, and we timestamp every starter update our watcher detects during the day. What we do not capture in the live feed is the reason for a change, the lineup posting time, or in-game injury news — those read as N/A in our packets rather than being guessed at.
Records, starters, and venue are retrieved when the packet is generated, not at first pitch. On a late scratch, that's a real limitation and we'd rather state it than paper over it.
Projected, confirmed, scratched
Each starter on our cards carries a confirmation status. The three states matter in different ways:
| Status | What it means | What the price is doing |
|---|---|---|
| projected | Team's stated intention, not yet locked | Opening number is against an assumption — softest price of the day |
| confirmed | Locked in for this start | The largest single source of uncertainty is now resolved |
| scratched | Will not start | Bets void at most books; line re-prices from scratch |
This is the one case where you can price a player
Most of the time, you can't isolate what the market thinks any single player is worth. Too much moves at once.
A starter scratch is the exception. Everything else about the game is held constant — same teams, same park, same day, same lineups — and one input changes. The difference between the consensus before the announcement and the consensus after it is about as close to a clean measurement of a player's market value as baseball offers.
That's worth watching over a season. A pitcher whose scratches consistently move the line 45 cents is being priced very differently from one whose scratches move it 12, regardless of what the two men's ERAs look like.
A useful habit
When you see a large move on the morning board and no obvious lineup news to explain it, check the probables first. A quiet pitcher change explains more big morning moves than any other single cause — and it's the one cause you can verify in about ten seconds.
The reverse holds too. If the probables are unchanged and confirmed, and the line still moved forty cents, then something else is going on: money, or a book correcting a number it posted badly. Those are worth a different kind of attention.