LOCK LABS — How It Works

How Lock Labs Works

Every pick is scored by a 7-component model. No gut calls. No hype. Just data.

Scrape Stats

→

Confirm Lineups

→

Score 7 Components

→

Rank & Tier

→

Publish Picks

points

Hit Rate

How often the batter is getting hits lately. Weighted toward the last 7-14 games to capture streaks, not season-long averages that mask cold spells.

points

Contact Quality

Line drive rate, hard-hit rate, and barrel percentage. A player hitting the ball hard is more likely to convert at-bats into hits, even if his average dipped temporarily.

points

Pitcher Matchup

ERA, WHIP, K rate, and batting average against filtered by batter handedness. A lefty slugger vs a right-handed pitcher with a 5.40 ERA is a different bet than vs a 2.80 ERA ace.

points

Lineup Position

Batting 1st-4th means more plate appearances and better pitches to hit. Lower lineup slots get fewer chances and often face relievers.

points

Park Factor

Coors Field is not Petco Park. We adjust for each ballpark run environment so a pick at a hitter-friendly park gets a bump.

points

Plate Discipline

Walk rate, strikeout rate, chase rate. Disciplined hitters see more pitches, get on base more, and are less likely to go 0-for-4 on three pitches.

points

Momentum

Recent trend direction: is the player heating up or cooling down? Captures short-term trajectory that the other components miss.

Tier System

S+

82+ pts

Elite locks

78–81 pts

Strong plays

B+

75–77 pts

Solid value

72–74 pts

Moderate edge

<72 pts

Low confidence

Beat the Streak (BTS)

Each day we pick 2 players most likely to record at least one hit. The BTS streak counts consecutive days where both picks hit. It's the hardest test of the model — picking individual games is easier than chaining perfect days. Our current system optimizes for floor (high contact, low K rate) over ceiling.

Bankroll Simulator

See what your balance would look like if you'd bet flat units on every Lock Labs pick.

Unit Size ($)

Juice (vig)

Simulation uses actual graded pick results from the database. Assumes flat-unit betting at the specified juice on every pick. Past performance doesn't guarantee future results.