Methodology
Metric glossary
Definitions for provider statistics and the derived metrics calculated by OpenBox. Derived values are estimates and are labeled accordingly.
Game flow
Win probabilityThe provider's estimate of each team's chance to win after a play.
Excitement indexThe sum of absolute win-probability changes across a game. More swings produce a higher value.
Leverage playA play ranked by the size of its win-probability movement.
Lead changeA crossing of the 50% win-probability line, excluding exact ties.
Win probability added (WPA)The change in the acting team's win probability across one play, in percentage points. Used for the most-important-plays list and per-player WPA.
NHL win probabilityThe NHL feed does not publish win probability, so OpenBox estimates it from the score and time remaining with a normal approximation to the remaining goals. It is a model, not a market price.
Kalshi market priceThe price of a Kalshi binary contract in cents per $1 (0–100¢), which is the market-implied probability of that outcome. Bid/ask are the live order book; last is the most recent trade. Kalshi is a prediction market, so no de-vigging is applied.
Postgame win expectancyHow often a team that produced this game's underlying performance would be expected to win, ignoring the final score. Football uses net EPA, basketball shot-quality expected points, baseball BaseRuns, hockey and soccer expected goals.
Postgame input marginsThe per-input football comparison shown under the expectancy bar: success rate, explosive rate, explosiveness, points per opportunity, field position, havoc and turnovers. Descriptive, not a fitted weight model.
Expected winsOn team pages, the sum of a team's postgame win expectancy across its games. Compare it with actual wins to spot luck.
Football
Expected points (EP)Estimated points expected from a down, distance, and field-position state.
Expected points added (EPA)The value of the resulting state minus the pre-play state, adjusted for turnovers and scores.
EPA per play / dropback / rushTotal EPA divided by scrimmage plays, dropbacks (passes, sacks, scrambles) or designed runs.
Success rateA play gaining 50% of the needed yards on 1st down, 70% on 2nd, or 100% on 3rd and 4th. Touchdowns succeed; turnovers fail.
ExplosivenessAverage EPA on successful plays: how much a team gains when it does succeed.
Explosive plays / rateRuns of 12+ yards and passes of 16+ yards, and their share of scrimmage plays.
Early-down EPA / late-down successEPA per play on 1st and 2nd down, and success rate on 3rd and 4th down.
Standard / passing downsStandard downs are 1st down, 2nd down with 7 or fewer yards to go, and 3rd/4th down with 4 or fewer. Passing downs are the complement.
Middle 8The final four minutes of the second quarter plus the first four minutes of the third: the part of the game where clock and possession strategy changes most.
Line yardsRushing yards credited to the offensive line: losses count 120%, the first 4 yards 100%, yards 5–10 50%, nothing beyond 10.
Highlight yardsRushing yards beyond what the line is credited with, attributed to the ball carrier.
Opportunity rate / stuff rateShare of runs gaining 4+ yards, and share stopped at or behind the line of scrimmage.
Power runA run on 3rd or 4th down with 2 or fewer yards to go. Conversion is gaining the needed yards.
Havoc rateTackles for loss, passes defended, interceptions and forced fumbles, divided by the opponent's plays.
Havoc EPA lostExpected points lost on plays where the offense commits a sack, turnover or tackle for loss.
Scoring opportunityA drive that reaches the opponent's 40-yard line. Points per opportunity measures finishing.
Available yardsYards gained on drives divided by the yards available from each drive's starting spot.
Drive success rateShare of scrimmage plays in a drive that meet the success-rate definition.
Scripted drivesThe opening possession of each half, when a staff has had time to plan the first series.
Target matrix / rush matrixPass targets by short/deep and left/middle/right, and runs by left/middle/right. Direction and depth are parsed from the play description; attempts without a direction are excluded.
Usage / opportunityOpp = carries + targets. Tgt% is a player's share of team targets, FD% his share of team first downs, and 3D± third downs converted minus expected at the yards to go.
Five factorsBill Connelly's framework — efficiency, explosiveness, field position, finishing drives, and turnovers — shown with offense, defense, and margin for a team season.
Fourth-down decision modelCompares the actual 4th-down choice with the expected-points value of going, kicking a field goal, or punting. Conversion, field-goal and punt probabilities are approximations.
Team projectionA pregame estimate from the difference in net EPA/play between the two teams, converted to a margin and win probability with a normal curve. Directional, not a betting model.
Expected turnovers / turnover luckHalf of all fumbles plus 22% of passes defended; luck is expected minus actual turnovers.
Basketball
Estimated possessionsFGA + 0.44 × FTA − offensive rebounds + turnovers.
Offensive rating (ORtg)Points scored per 100 estimated possessions. Net rating subtracts the opponent's.
Four factorsEffective FG%, turnover rate, offensive rebound rate and free-throw rate: the four levers of possession efficiency.
Effective field-goal percentage (eFG%)Field-goal percentage adjusted to give made threes their extra value.
True shooting percentage (TS%)Scoring efficiency that includes two-pointers, three-pointers, and free throws.
Shot quality (xPts)Each field-goal attempt valued at the league make rate for its zone (rim, short mid-range, long mid-range, three), from ESPN shot coordinates.
Shot-makingActual field-goal points minus shot-quality expected points.
Usage rateEstimated share of team possessions a player used while on the floor.
Clutch pointsPoints in the final five minutes of regulation or overtime with the margin within five.
Game scoreA compact estimate of a player's box-score contribution; useful for comparison, not a complete impact model.
Baseball
Run expectancy (RE24)The change in expected runs across a plate appearance (from the 24 base-out states) plus runs scored. Batters get credit; pitchers get the negative.
BaseRunsRuns a lineup would be expected to score from its hits, total bases, walks, home runs and outs, regardless of sequencing.
ISO / BABIPIsolated power (SLG − AVG) and batting average on balls in play.
CSW rateCalled strikes plus whiffs, divided by pitches.
Whiff rateSwinging strikes divided by swings.
FIPFielding-independent pitching: (13 × HR + 3 × BB − 2 × K) ÷ IP + 3.10.
Hockey
Expected goals (xG)Each unblocked attempt valued with a distance/angle model, adjusted for shot type, rebounds and power plays.
Corsi / FenwickAll shot attempts, and unblocked shot attempts. Shares show each team's portion.
High-danger chanceAn unblocked attempt worth at least 0.15 xG.
PDOShooting percentage plus save percentage; 100 is neutral and extremes tend to regress.
GSAxGoals saved above expected: xG faced minus goals allowed.
Hockey usageTime on ice, shifts, faceoff percentage, hits, blocks, giveaways, takeaways, and power-play goals for each skater.
Soccer
Expected goals (xG)Estimated from the commentary for every attempt: location, header or foot, penalty, set piece and fast break.
Non-penalty xGxG excluding penalty kicks.
Big chanceA shot worth at least 0.30 xG.
Expected assists (xA)The xG of shots a player set up with the final pass.
General
LiveThe event is currently in progress according to the upstream status feed.
Team comparisonPaired values published by the provider; the bars show relative magnitude, not percentile rank.
UnavailableThe upstream event or league feed did not publish enough data for that view.