Why a Projection Moves a Full Strikeout When the Lineup Posts
At noon a starter projects for 6.1 strikeouts. At 5:40 the official lineup drops, nobody got hurt, nothing dramatic happened, and the same projection reads 7.0. That is not the model being indecisive. It is the model finally getting the one input it had been guessing at all day.
Every Morning Projection Is Built on an Assumption
A strikeout projection is not a property of the pitcher. It is a property of the pitcher and the nine hitters he is scheduled to face. Until the official card is posted, roughly two hours before first pitch, that second half is an estimate — usually the most recent lineup that team used against the same handedness, adjusted for known rest days.
That estimate is decent. It is right about six or seven names out of nine most days. But the two it misses are frequently the two that matter, because managers do not swap players at random. They rest the regular who has started twelve straight, they sit the left-handed bat against a tough lefty, and they slot a contact-first hitter into the top of the order. Every one of those choices pushes on the strikeout math in the same direction at the same time.
Both terms on the right depend on who actually plays. That is the whole story, and the rest of this post is the three separate ways a posted lineup rewrites them.
1. Who Is in It
A team's aggregate strikeout rate is not a fixed number. It moves several points depending on which bench pieces are in the order. A regular who strikes out 18% of the time being replaced by a bench bat at 29% shifts the whole lineup's rate by more than a point on its own — and teams rarely rest just one player.
Projected lineup K rate vs RHP: 22.8%. Blended with the pitcher's own 28% K rate, the model lands at a matchup rate of 25.5% across an expected 24 batters faced.
24 × 25.5% = 6.1 projected strikeouts.
The card posts. Two regulars are resting on the back half of a road trip and a rookie is hitting eighth. Lineup K rate vs RHP is now 26.1%, matchup rate 28.1%, and a whiffier lineup means slightly more traffic-free innings, so expected batters faced ticks to 25.
25 × 28.1% = 7.0 projected strikeouts. Nine tenths of a strikeout, no injury, no news.
This is the single biggest reason a projection that looked marginal at lunch looks clean at dinner, and it is invisible if you only ever check once.
2. The Handedness Mix
Most pitchers have a real platoon split, and for some it is enormous. A left-hander who strikes out 31% of left-handed hitters and 22% of right-handed hitters is a completely different projection depending on how many lefty bats the opposing manager decides to run out.
| Posted lineup vs this LHP | Weighted K rate | xK over 24 BF |
|---|---|---|
| 4 LHB / 5 RHB (projected) | 26.0% | 6.2 |
| 2 LHB / 7 RHB (posted) | 24.0% | 5.8 |
| 6 LHB / 3 RHB (posted) | 28.0% | 6.7 |
Nearly a full strikeout of spread between the two extremes, driven entirely by a decision made by someone in the other dugout at 4pm. Against a pitcher with a small platoon split this barely registers. Against a specialist with a nine-point gap, it is the difference between a comfortable read and a coin flip — which is exactly why checking the split before the lineup posts is guesswork dressed up as analysis.
Worth knowing: handedness cuts the other way too. When a manager stacks righties against a right-handed specialist to avoid the platoon disadvantage, he is often trading strikeout avoidance for weaker contact. The strikeout projection drops while the hits projection stays flat.
3. Batting Order — Quietly the Largest Effect on Hitter Props
For hitters, the identity question is usually settled — you know your guy is playing. What you do not know until the card posts is where. And batting order is not cosmetic. It is a direct multiplier on plate appearances, which is the raw material of every counting stat.
| Lineup slot | Avg PA per game | P(1+ hit), .240 per-PA hitter |
|---|---|---|
| 1st | 4.65 | 71.7% |
| 2nd | 4.55 | 71.0% |
| 4th | 4.34 | 69.5% |
| 6th | 4.13 | 67.9% |
| 8th | 3.92 | 66.2% |
| 9th | 3.81 | 65.2% |
Leadoff to the nine-hole is about 0.85 extra plate appearances. On a 1+ hits read that is six or seven percentage points of probability. On the rarer outcomes it is proportionally bigger: a home run is a roughly 4% per-PA event, so 0.85 extra trips is close to a 20% relative increase in the chance of one clearing the fence. A hitter moving from sixth to second is not a small tweak — it is the same skill with 12% more opportunity.
The market moves standard lines on this too, but not always immediately, and not always by the full amount. The gap between the model's revised number and a line still built on yesterday's batting order is one of the more reliable disagreements to look for in the two hours after cards drop.
What Does Not Move
It is worth being precise about the boundary here, because overreacting to a posted lineup is its own mistake. These inputs are settled well before the card and do not change when it lands:
- Park factors — fixed by venue, and stable across multi-year samples.
- Umpire assignment — published in the morning, and its strike-zone effect is already baked into the noon projection.
- Weather — refreshed on its own schedule, unrelated to the lineup.
- The pitcher's own form — his rolling rates are what they are.
So if a projection moves at 5:40, it moved for a lineup reason. That makes the change easy to reason about instead of mysterious.
The One Trap: Posted Is Not Final
A lineup can post and then change. Late scratches happen during batting practice, and a name that appeared on the card at 4:50 is occasionally gone by 6:30. The model re-reads the official card on a schedule rather than freezing on the first version it sees, so a scratch is picked up — but there is a window where a projection reflects a lineup that is about to be edited. If you are working a matchup close to first pitch, it costs nothing to reload.
Bottom line: a projection built on a projected lineup is a draft. The same projection after the card posts is the real one. If you only look at a matchup once a day, look at it after the lineups land — the number you would have used four hours earlier is frequently off by nearly a full strikeout, and the reason is never hidden.
How to Use This on PropPrizm
- Matchup cards show whether a lineup is confirmed or projected. Treat everything on a projected card as provisional.
- The opposing lineup's K rate and handedness split are shown on the pitcher card — that is the term doing most of the work in the strikeout math.
- Batter cards show the lineup slot. If it changed from the player's usual spot, the plate-appearance assumption changed with it.
- Late afternoon is the highest-information window of the day. Before that you are analyzing an assumption.
Open any matchup on the dashboard before lineups post, note the projection, then check it again once the card is confirmed. The ones that move the most are the ones where the lineup was doing the most work all along. PropPrizm is a statistical tool for informational and entertainment purposes and does not guarantee outcomes.