The Locked-In Index — measuring how well a hitter is seeing the ball

What the Locked-In Index is

The Locked-In Index (LII) is a 0–100 score for how well a hitter is currently seeing and striking the ball. Roughly 50 is league average, and the scale is built so a score reads like a percentile-flavoured rating rather than a rate stat. It is a form metric, not a talent metric: the question it answers is “how is this hitter swinging right now,” not “how good is this hitter.”

It is deliberately narrow. LII carries no park, no opposing pitcher, no weather and no game context — tonight’s whole picture is the Matchup Score’s job, which is designed to be read alongside it. Keeping LII narrow means a hot hitter in a bad spot and a cold hitter in a good spot are legible as different things instead of averaging into one muddy number.

The three components and their weights

LII is a weighted average of three components: Power at 45%, Contact Quality at 20%, and Plate Discipline at 35%.

Power is the mean of four Statcast z-scores: maximum exit velocity, average exit velocity, barrel rate, and hard-hit rate. Contact Quality is the mean of two: expected wOBA on contact and expected batting average. Plate Discipline is the mean of four, with three inverted because lower is better: chase rate (inverted), whiff rate (inverted), zone-contact rate, and strikeout rate (inverted).

Every input is z-scored against the current league distribution, so the number is relative to how the league is hitting now rather than to a fixed historical baseline. Each component is converted to a 0–100 subscore as 50 + z × 25, clamped to the scale, and the headline is the weighted average of those three subscores. That construction matters for reading it: the parts sum to the whole, so a breakdown always explains its own headline.

Why peak exit velocity carries so much of it

The single most useful discovery in building this metric was that maximum exit velocity is the strongest forward predictor of production among the inputs available, and it was entirely absent from the first version of the index. It enters through the Power component and is the main reason Power carries the largest weight.

The mirror-image finding was that expected statistics — the x-stats that dominated version one — carry very little incremental signal once batted-ball quality is already in the model. They survive as a small Contact Quality component rather than as the centre of the metric.

Why the weights were rebalanced

An earlier version weighted the components 65% Power, 8% Contact Quality, 27% Plate Discipline. Those weights were fitted, but they had two problems in practice. Power alone accounted for 84% of the headline’s variance, so clicking into a score showed three bars where one decided everything. More seriously, the metric had stopped behaving like a form metric at all: a hitter’s score varied almost as much between players as it did for the same player over time, which describes a talent board rather than a Locked-In Index.

The current 45/20/35 weights cut Power’s variance share from 84% to 58% and restore a genuine three-part breakdown. The cost was measured rather than assumed, on a cohort neither the original fit nor its test set had seen: about 0.006 of Spearman correlation with forward wOBA. On the statistic the live selectors actually use — the forward production of the top-ranked hitters each day — the change was not distinguishable from zero, and it was slightly ahead on the most recent season.

What the number cannot tell you

The weights are not load-bearing, and it would be dishonest to present them as precisely optimal. A full sweep of the weighting space across roughly 49,000 hitter-day observations found a broad plateau: 43 of 73 grid points sat within 0.010 Spearman of the best, spanning Power weights anywhere from 0.40 to 0.75. The specific numbers are a defensible point inside that plateau, chosen partly so the breakdown is readable, not a peak the metric balances on.

LII is also not a betting signal by itself. It describes current form; it says nothing about the price being offered, and a hitter can be thoroughly locked in at a number that is already fully accounted for in the market.