Pitcher strikeout props, both sides priced

Strikeout props are the closest thing to a fairly priced player prop measured so far. On 2,061 props the over side returned about -4.5% per bet, outside its noise floor for the first time - the sign and, narrowly, the size are supported; the ordering between lines still is not.

Written by Jesse, NegativeEV. Last updated 2 August 2026.

• Sample: 2,061 pitcher strikeout props
• Window: 2026-04-19 to 2026-07-31
• Priced at: best posted price on each side
• Measured: 2026-08-02

Realized EV per bet	Figure
Over, all lines	-4.5% (± 2.2, z = -2.1)
Over 3.5	-4.8%
Over 4.5	-6.0%
Over 5.5	-6.6%
Under, all lines	-3.6% (± 2.1)

The 4.5 and 5.5 lines carry most of the board on a given night, and that is where the over came back worst. Each individual line cut still sits inside its own noise floor, and the thin 2.5 and 6.5 buckets bounce positive on samples too small to mean anything, so the ordering between lines is not established by this window either.

Strikeout props sit near even money, outside the longshot band where margins run thickest, which is the structural difference between this market and anytime home run props at about -23% over the same window.

Method

Every archived pitcher strikeout prop is graded against the pitcher's real strikeout count in that game from the mapped play-by-play corpus, at the best posted price on each side, and reported as realized win rate and expected value per bet, overall and by line bucket. Props for games missing from the corpus are excluded rather than defaulted.

Inside the noise floor

One cut in this census clears its own noise floor. On 2,061 props the over across all lines is -4.5% ± 2.2 (z = -2.1), which separates from zero for the first time on this market. Nothing else here does: the per-line over cuts (-4.8% on the 3.5s, -6.0% on the 4.5s, -6.6% on the 5.5s) each sit inside their own floor and overlap each other, so the ranking between lines is not established, and the thin 2.5 and 6.5 buckets bounce positive on samples too small to read. The under side across all lines is -3.6% ± 2.1, which covers zero, so it is an estimate rather than a measured cost. The large-sample prop censuses - anytime home runs and batter total bases - remain the sharper measurements on this board.

Written up for readers: https://negativeev.com/about/strikeout-props All findings: https://negativeev.com/research