Learn

Sports Statistics Glossary

Every statistic and term this site displays, defined in plain language — what it measures, and the misreading worth heading off. 40 entries.

Statistics & Method

Sample Size#: How many opportunities a statistic is built from — at-bats, attempts, games. It sets how much the number can be trusted.; A small sample does not produce a slightly wrong number; it produces a very wide range reported as one number.; What Sample Size Does to a Batting Average →
Confidence Interval#: The range of true values consistent with what was observed. A .300 average over 40 at-bats has a 95% interval of roughly .158 to .442.; The point estimate is the middle of the interval, not a measurement of the truth.; What Sample Size Does to a Batting Average →
Regression to the Mean#: The tendency for extreme results to be followed by less extreme ones, because extremes are partly luck and luck does not repeat.; It is not a force and nobody is 'due'. It also does not mean a player got worse — it means the extreme reading was an overestimate.; Regression to the Mean →
Base Rate#: How often something happens normally, with nothing special going on. The comparison point that makes any other number meaningful.; Most impressive-sounding sports facts are ordinary base rates described dramatically.; Base Rates →
Mean vs Medianalso: Average#: The mean adds everything and divides; the median is the middle value. When the mean sits well above the median, a few big results are carrying it.; Why Per-Game Averages Mislead →
Standard Deviation#: A measure of spread — how far a typical result sits from the average. Two players with the same average and different standard deviations are different players.; Why Per-Game Averages Mislead →
Rate Stat#: A statistic with a denominator — per carry, per plate appearance, per 36 minutes. Measures quality per opportunity.; Rate Stats vs Counting Stats →
Counting Stat#: A running total — home runs, receptions, tackles. Measures production, and therefore rewards opportunity as much as skill.; League leaders in counting stats are often the players given the most work.; Rate Stats vs Counting Stats →
Correlation#: Two things moving together. It does not say which one moves the other, or whether a third thing moves both.; Teams that rush for 150 yards win most of the time — because leading causes running, not the reverse.; Correlation, Causation, and Sports Narratives →
Survivorship Bias#: Drawing a conclusion from only the cases that made it to the end, while the ones that dropped out are invisible.; 'Players who start hot finish strong' — players who start cold get benched and never generate a second half to count.; Hot Streaks and Survivorship →
Placebo Testalso: Control group#: Running the same analysis somewhere the effect should not exist. If the result shows up there too, the original finding is not what it claimed to be.
Split#: A statistic computed over a slice of a season — versus left-handers, at home, in September. Every split is a small sample by construction.; Splits with a physical mechanism (platoon) are more trustworthy than splits without one (month, versus one team).; Splits, and When They Are Real →

Markets & Odds

American Oddsalso: Moneyline odds#: The price format used by US sportsbooks. Negative numbers show how much must be risked to win $100; positive numbers show the win on a $100 stake.; How Odds Actually Work →
Implied Probability#: The probability a price corresponds to. Negative odds: odds / (odds + 100). Positive odds: 100 / (odds + 100). So -150 implies 60%.; Implied probability includes the book's margin, so the two sides of a market add to more than 100%.; How Odds Actually Work →
Holdalso: Vig, Juice, Margin, Overround#: The amount by which a market's two sides add to more than 100%. It is the operator's built-in margin and it is present in every price.; A gap between observed hit rate and implied probability is usually the hold, not a mispricing.
Closing Line#: The final price before an event starts. Widely used as a reference point because it reflects all information available at kickoff or first pitch.
Propalso: Player prop, Proposition market#: A market on an individual outcome within a game — a player's hits, receptions, or points — rather than on the game result.
Milestone Ladder#: A set of nested one-sided markets on the same player, such as 1+, 2+, and 3+ hits. Each level is a subset of the one below it.; Because the levels are nested and one-sided, their implied probabilities must never be added together.
Line Movement#: A change in a market's price between when it was posted and now. Movement can come from news, from money, or from an operator adjusting its own exposure.; A large move on a batter prop a few hours before first pitch is most often lineup confirmation, not information about the outcome.
Parlayalso: Accumulator#: A single wager combining several outcomes, all of which must occur. The hold compounds with each leg added.; Parlay Math →

Baseball

Batting Averagealso: AVG, BA#: Hits divided by at-bats. The oldest rate stat in the sport and one of the noisiest.; It stabilises very slowly. A .300 and a .270 hitter are still only about 1.5 standard deviations apart after 1,000 at-bats.; What Sample Size Does to a Batting Average →
On-Base Percentagealso: OBP#: How often a hitter reaches base per plate appearance, including walks and hit-by-pitches. A better measure of offensive value than batting average.
Slugging Percentagealso: SLG#: Total bases divided by at-bats. Measures power, not the ability to reach base.
OPS and OPS+#: OPS adds on-base percentage and slugging. OPS+ rescales it for park and league, with 100 set as league average.
Park Factor#: A multiplier for how much a stadium inflates or suppresses an outcome. 1.15 for home runs means roughly 15% more than a neutral park.; Noisy year to year, different by handedness, and an average across all weather conditions.; Park Factors →
Platoon Split#: The difference in a hitter's performance against left- versus right-handed pitching. One of the few splits with a genuine physical mechanism.; Splits, and When They Are Real →
ERA and ERA+#: Earned runs allowed per nine innings. ERA+ adjusts for park and league with 100 as average, so higher is better — the opposite direction to raw ERA.
WHIP#: Walks and hits allowed per inning pitched. A simple measure of how much traffic a pitcher allows.

Football & Fantasy

PPR#: Points Per Reception scoring, where each catch is worth a point. It raises the value of high-volume receivers and pass-catching backs.
ADPalso: Average Draft Position#: Where a player is typically drafted, expressed as an overall pick number rather than a round-and-pick.; ADP describes what drafters are doing, which is a measure of consensus opinion — not of a player's value.
Target Share#: The percentage of a team's passing targets that go to one receiver. A role measure, and more stable week to week than receptions.
Snap Share#: The percentage of a team's offensive plays a player was on the field for. The clearest available signal of role.
Yards Per Carryalso: YPC#: Rushing yards divided by attempts. A rate stat, and heavily influenced by a few long runs.; One 70-yard run moves a game's YPC enormously. Check the median carry, not just the mean.; Why Per-Game Averages Mislead →
Strength of Schedulealso: SOS#: A summary of the quality of opponents a team has faced or will face. Remaining schedule is usually more useful than schedule already played.; Early-season figures are unstable because opponent ratings are still moving.; Strength of Schedule →
Regressed Baseline#: Our Draft Kit figures: last season's per-game PPR production pulled toward the positional mean by how many games the player actually appeared in, with floor and ceiling from that player's own weekly variance.; This is NOT a projection model. It describes what already happened, adjusted for sample size. Rookies have no prior season and are shown as null rather than guessed.; Why Projections Have Ranges →

Basketball

Pace#: Possessions per 48 minutes. A team-level setting that changes how many chances every player on the floor gets.; Pace and Usage →
Usage Ratealso: USG%#: The share of a team's possessions that end with a given player shooting, drawing a foul, or turning it over. A measure of role, not of quality.; Efficiency usually dips slightly as usage rises, because the extra shots are the harder ones.; Pace and Usage →
Per-36 Minutes#: Production rescaled to a common 36 minutes of playing time, so players in different roles can be compared.; It assumes production scales linearly with minutes, which breaks down for players who only play short bursts.
True Shooting Percentagealso: TS%#: Scoring efficiency accounting for two-pointers, three-pointers, and free throws together. A fairer comparison than field-goal percentage.
Effective Field Goal Percentagealso: eFG%#: Field-goal percentage adjusted for the fact that a three-pointer is worth more than a two.

← Back to all lessons