Closing Line Value in MLB Markets — A Research Memo

Executive Summary

This memo covers how closing line value behaves specifically in Major League Baseball
markets, and what that tells us about where a price is likely to be wrong. CLV measures the
gap between the price taken and the price the market settled on at close. It is a property of
an entry, not a grade assigned to a sportsbook, and the distinction drives everything below.

Baseball is an unusually informative place to study it. The 162-game season generates a large
sample quickly, individual games vary widely in scoring, totals move in quarter-run steps, and
lineup and bullpen news lands late in the afternoon relative to first pitch. That combination
means the distance between an opening number and a closing number is often substantial, and
frequently explainable.

What closing line value is, precisely

Take an over at 8.0 runs. The market closes at 7.5. The over was bought on a lower total than
the market's final answer, so that position holds positive closing line value. The game's
result is irrelevant to that fact and to its usefulness.

The metric works because the close is the most informed number of the day: by then the market
has absorbed confirmed lineups, reliever availability, weather, and every dollar that arrived
with a view. Consistently beating it means a position was opened before the market agreed with
the conclusion behind it, which is one of the few claims about process that can be tested
without waiting on a large sample of results.

Why MLB totals and MLB sides behave differently

Sample accumulation. A 162-game season with many games per day produces a large volume of
graded positions quickly relative to a low-frequency market, so a totals lane can be measured
on a forward sample sooner than almost anywhere else. Short seasons and schedule disruptions
do the opposite: fewer observations, wider swings, and far more caution required before any
number looks like a finding rather than a run of variance.

Park and weather effects are priced, but unevenly. Run-scoring environments shift with
wind direction, humidity and park dimensions, and those inputs are public and arrive close to
first pitch. Books generally have the park factors already in the number; what they are slower
to absorb is the specific combination — a hitter-friendly environment plus a confirmed weak
bullpen plus a wind blowing out. That is where a total can be demonstrably stale at open and
correct by close, which is exactly the pattern that produces positive closing line value on the
over. Coors Field is the obvious example of a park whose effect is well known and nonetheless
still requires the day's weather to be resolved before the number is right.

Late information on the side markets. Moneyline and run-line prices react strongly to
starting-pitcher confirmation and bullpen rest, both of which can be settled well after the
open. A side market that barely moves through the day is usually a market that already knows
something, not one that is being neglected.

Competitive structure. Totals are set and competed across many books at once, which makes
the close an efficient consensus. Sides and run lines in less liquid pairings can leave a
wider spread between the best and worst available price, which matters when measuring an entry
against a single book's number.

What we treat as a real finding

A median closing line value over one slate is an anecdote. The claim only becomes usable when
it holds across a large forward sample, computed the same way every day, against the close of
the same market at the same book.

Three habits quietly inflate it and are worth stating because they are common:

1. Defining "closing" after seeing the result. The close must be fixed in advance and held fixed.
2. Comparing an entry at one book against a close observed at another, which measures the
spread between books rather than any judgment being correct.
3. Grading against a line that never existed, reconstructed from what seemed likely.

Signal logs opening price, the model's own number and closing price for every graded position
so the measurement is reproducible rather than argued. Where the close is a proxy drawn from an
odds aggregator rather than a sportsbook-confirmed number, that is stated in the study rather
than smoothed over, and it is documented in the CLV Validation Study.

Where this is used

Median closing line value is published per model in the Daily Analysis alongside the
record it came from, so the two are read together rather than one being allowed to stand in for
the other. For the arithmetic on a single position, the free
CLV calculator takes an entry price and a closing price and returns
the value. For why the scoreboard is a weak teacher on its own, see
Why Win Rate Is a Terrible Metric for Serious Bettors and, for
what closing line value actually measures, the explainer
What Closing Line Value Actually Means.

Research only · estimates only · not betting advice.

---

Educational estimates only · Not betting advice · Past research ≠ future results.