How we count

Every number on the site is computed by us from the match reports, with the same rules in every league. These are the rules.

Standings: Two points for a win, one for a draw. Teams are ordered by points, then goal difference, then goals scored. Where the federation breaks ties by head-to-head (a Luxembourg table says so under it), teams level on points are first ordered by the games among them. A play-off table includes the points carried over from the regular season, as the federation counts them. Cups and knockout rounds have no table. Only matches the source marks as played count; a postponed match never shows a score.
Form: The last five played matches of a team, the latest first: W a win, D a draw, L a loss. Only matches of the league shown count.
Shooting percentage: Goals divided by shots, from the lineup of each match. 7m throws are included in shots and goals; the 7m column shows them separately. To rank among the leaders a player needs at least 4 shots per round played (24 after six rounds).
Save percentage: Saves divided by shots faced, per goalkeeper, from the match lineup. To rank among the leaders a goalkeeper needs at least 10 shots faced per round played (60 after six rounds).
Attack: One possession of the ball. Sportinfocentar marks the event that ends a possession and how long it lasted; we count those. The average attack is the mean length in seconds.
Attack efficiency: Goals divided by attacks. 56% means the team scores on 56 of 100 attacks.
Attack and defence per 50 attacks: Tempo is a team's attacks per match. Attack is goals scored per 50 own attacks, defence goals conceded per 50 of the opponent's attacks, net the difference. A fast team plays more attacks, so it scores and concedes more goals per match; per 50 attacks that no longer counts. A match counts when the attacks of both teams are recorded, so only in leagues with advanced statistics. Averages pool every counted match. Colours show a team's place from 0 to 100 among the teams compared; across leagues a team needs 3 counted matches.
Four factors: After Dean Oliver's four factors of basketball, in handball form: field-shot percentage (every shot but the 7m throws, blocked shots included), lost balls per 50 attacks, 7m throws won per 50 attacks, fast breaks (their share of all shots and how many are scored), and goals scored within 10 seconds of the opponent losing the ball. Attack counts the team's own attacks; defence counts the same numbers for its opponents against it. Only matches with counted attacks; a measure the league does not record is left out.
Goalkeeper: saves above expected: Every shot on target a goalkeeper faces (goals and saves, 7m included) expects the save rate of its court zone in this league this season (the same zone rates as the HandHub index). Saves above expected = saves minus those expected saves. Goals conceded per 50 attacks while he was in goal: an opponent attack belongs to the goalkeeper on its finishing shot; an attack ending in a lost ball, a miss, a block or an empty-net goal belongs to the goalkeeper of the nearest shot faced in the same half of that match.
Shots against the league: For each court zone of the shot map: the team's goals over shots against every shot of the league from that zone. Hot means at least 5 percentage points better than the league, cold at least 5 worse, and a zone needs 10 shots of the team for either. The shot mix is the share of the team's shots taken from each zone, next to the league's.
Close finishes: Attacks that start in the last five minutes of regular time (55:00 or later) while the score is within two goals. Goals per 50 such attacks for the team and for its opponents, against all such attacks in the league; the shooters are every shot of the team in those attacks.
First attack of the match: The team's first attack of each match: the player who ended it (the shooter, or the player who lost the ball) and how it ended. The season share is the first attacks that ended in a goal, against every team's first attacks in the league.
Attacks a player more or fewer: Goals per 50 attacks split by the numbers on court when the attack ends, from the 2- and 4-minute suspensions as in the game situation (a suspension counts its full length). The comparison is every attack of the league with a player more, or with a player fewer.
Who finishes the attacks: Every attack of the team goes to the player who ended it: the shooter or the player who lost the ball; an attack that ends with a 7m throw goes to the player who won the 7m when the source names him. Share is his attacks over all the team's attacks; success is the share of his attacks that ended in a goal. Players with an attack in fewer than 3 matches are not listed.
Game situation: A player more, a player fewer or equal numbers on court, derived from 2-minute (120 s) and 4-minute (240 s) suspensions and the second of each goal. Two suspensions at once count twice. Empty-net play is not tracked yet.
Quarters: Goals by 15-minute period of regulation time: 0–15, 15–30, 30–45, 45–60 minutes, from the second of each goal.
Average per game (match overview): In the average row of a team's match overview every number is rounded to a whole goal (a half rounds up) and the difference is taken between the rounded numbers, so 10.5:6.2 shows as 11:6 +5.
Timeout swing: For a team: goals for minus goals against in the five minutes after its own timeouts, minus the same balance in the five minutes before them, summed over the season. A positive number means timeouts helped.
Timeout on the first attack: For each timeout: did it work? Where the source marks where every attack ends (the leagues with shot spots), the calling team’s first attack after the timeout decides: a goal is a success; a miss, a save, a block, the post or a turnover is not. When a 7m is earned in that attack, the 7m throw decides, however long it takes. Stoppages in between (a goalkeeper change, an injury) do not matter. In the other leagues: the team scored the next goal within 45 seconds, or that goal was its 7m, or it came after the team’s missed 7m. A timeout the source records twice (same team, same second) counts once. "+3 min" is that team’s goals minus the opponent’s in the three minutes after the timeout. "Close game" keeps only timeouts called with the difference between -2 and +2.
Head to head: The team comparison: both teams are limited to the matches the filters keep (all / last 5 / home / away, one opponent, a date range), and the head-to-head tab to the matches between them among those. Wins, draws, goals, quarters, a player more / fewer and the players of those matches come from the same rows as each team page, so a team shows the same numbers on both.
7m against (stopped): The opponents' 7m throws against the team, from the opponents' lineups in the match reports, and how many of them were not scored: saved or missed, the sources do not tell them apart. The percentage is the stopped share.
Goal run: Three or more goals in a row by one team without an answer; the match page shows the score before and after each run.
Shot zones and goal mouth: Court zones are the shooting position the source records (wing, 6 m, 6–9 m, 9 m and beyond, by side; fast break and 7 m apart). The goal mouth is a 3×3 grid plus lob, for shots that ended as a goal or a save. Colours: 70% and above green, 40–69% amber, below 40% red.
Estimated positions: Croatia, Montenegro and North Macedonia send no playing positions. For their players we estimate one from where they shoot: at least six shots that tell a position (7m, fast breaks and long shots do not), more than half of them from one area (left wing, right wing, back court, line). A back is a left or right back when their back-court shots lean clearly to that side, otherwise a centre back. Goalkeepers come from the source. Checked against the real positions of the Serbian leagues: 82% exact, the misses almost all between left, centre and right back. Estimates are marked and are replaced as soon as a source sends a position.
Player search: One row per player and league. The thresholds compare the numbers as the table shows them: goals, assists and turnovers per match to one decimal, percentages whole. Players need 3 matches, and a percentage counts only above a minimum (20 shots for shooting %, 5 throws for 7m %, 50 shots faced for save %), unless the visitor sets another. A filter on a number some leagues do not record leaves those leagues out and says how many: shooting % and save % need shots (not in leagues with goals only), assists and turnovers the advanced level, the HandHub index shot zones. Estimated positions are included unless switched off, and marked.
HandHub index: One number from 50 to 100 for a player's whole contribution, on the model of the HBL Handball Performance Index. What goes in, per match: every shot (a goal adds 10 × (1 − p) + 1, a miss 1 − 10 × p, where p is how often shots from that zone score in this league this season; a 7m throw has no + 1), assist +4, earned 7m +4, earned 2-min +3, steal +8, block +7, turnover −8, 2-min suspensions −3, −4 and −6 (first, second and third in the match), red card −12 (not for the automatic red after a third 2-min), yellow card −1. Goalkeepers get 10 × (saves − expected saves), each shot on target expecting the save rate of its zone in this league, plus the same cards and suspensions. The idea: a goal from 9 m, where most shots fail, is worth more than one from 6 m, and a missed 7m costs more than a missed 9 m shot. The zone rates start from published international rates and move toward this league's own as its shots add up. Each match is put on a scale per role (field players, goalkeepers): 50 for the bottom 1 % of this league's matches this season, 100 for the top 1 % (5 % and 95 % while a role has fewer than 100 player matches); As in the Bundesliga's HPI, the scale shown is 50–100: a match above 100 shows as "100+" (its exact value still breaks ties on the MVP list) and one below 50 as 50. The season value is the average of the player's matches, each first held to 50–100 as the HBL does, so our season values compare with the HBL's; one decimal. To make the number easier to read, five bands in five colours: one match 90+ outstanding, 80–90 very good, 70–80 solid, 60–70 below average, under 60 weak; a season, which moves less, 80+ excellent, 75–80 very good, 70–75 solid, 65–70 below average, under 65 weak. The HBL publishes no bands; ours follow how often each value occurs there and in our leagues (about 3 % of matches reach 90, the middle match is about 70). Matches in which a player has no recorded action do not count; counters a source does not record count as 0. Only leagues that record where shots come from have the index. The MVP list ranks only players with an index in at least half of the rounds played so far, rounded up.
Source rating: The player and goalkeeper rating that Sportinfocentar computes; shown as the source's, never ours.
Referees: Per referee pair (referees officiate in pairs here; a referee with another partner is another pair): matches officiated, average total goals, average home margin, home win / away win / draw shares, and yellow cards, red cards, 2-minute suspensions and 7m throws in total and per match.
Sources and coverage: Serbia, Croatia, Montenegro and North Macedonia come from Sportinfocentar match reports with full play-by-play and shot positions; Slovenia comes from the RZS system with goals, shots, 7m, suspensions, cards and a play-by-play without shot positions. A block that lacks data simply does not appear. Ball photograph: "Handball the ball" by Armin Kuebelbeck, CC BY-SA 3.0, via Wikimedia Commons. Map data: Natural Earth.
Estimates and checks: Where a source does not mark the end of an attack (Slovenia), attacks are estimated from shots and labelled so. For every match the goals counted from the play-by-play must equal the official score; mismatches are reported and the official score is shown.