# Basketball analytics glossary

The terms behind our analysis and our EuroLeague pages, each with its formula and how we use it. Ratings are per 100 possessions unless stated otherwise.

## Possessions and ratings

### How are possessions calculated?

One team's turn with the ball. It ends in a shot attempt, a turnover or a trip to the free-throw line; an offensive rebound continues the same possession. Both teams have almost the same number in a game, so numbers per possession compare teams fairly while numbers per game reward pace.

Formula: `FGA − ORB + TOV + 0.44 × FTA`

The 0.44 × FTA term estimates free-throw trips from a box score. We count the trips themselves from the play-by-play.

[How we count possessions for the EuroLeague ratings](https://www.clutchdata.net/euroleague/net-rating#nr-method)

### Pace

Possessions per 40 minutes. A fast team does not have a better attack, only more chances, which is why we rate offense and defense per 100 possessions.

Formula: `40 × possessions / game minutes`

### How is offensive rating calculated?

Points scored per 100 possessions. The cleanest measure of how good an attack is, because it takes pace out.

Formula: `100 × points / possessions`

[EuroLeague offensive rating by team](https://www.clutchdata.net/euroleague/net-rating#season-ortg)

### Defensive Rating (DRtg)

Points allowed per 100 opponent possessions. Lower is better.

Formula: `100 × opponent points / opponent possessions`

[EuroLeague defensive rating by team](https://www.clutchdata.net/euroleague/net-rating#season-drtg)

### What is net rating?

Offensive minus Defensive Rating: how many points a team outscores its opponents by every 100 possessions. A team at +10 is ten points better per 100 possessions. Over a short stretch it says more about how good a team is than its record, which swings on a few close games.

Formula: `ORtg − DRtg`

[EuroLeague Net Rating, updated every round](https://www.clutchdata.net/euroleague/net-rating)

## Shooting

### Effective field goal percentage (eFG%)

Field goal percentage with each made three counted as one and a half made twos, since it is worth 50% more. A team that only shoots twos at 50% and one that only shoots threes at 33.3% both have an eFG% of 50%.

Formula: `(FGM + 0.5 × 3PM) / FGA`

### True shooting percentage (TS%)

Scoring efficiency including free throws: points per shooting possession, on the scale of a field goal percentage.

Formula: `PTS / (2 × (FGA + 0.44 × FTA))`

[EuroLeague advanced stats, updated every round](https://www.clutchdata.net/euroleague/advanced-stats)

### Four factors

The four things that decide a possession, after Dean Oliver: shooting (eFG%), turnovers (TOV%), offensive rebounding (ORB%) and getting to the line (FT rate). Each has an offensive and a defensive side. Our defensive points saved are built on three of them.

Formula: `TOV% = TOV / possessions · ORB% = ORB / (ORB + opponent DRB) · FT rate = FTA / FGA`

### Expected points per shot (xPPS)

What a model predicts a shot is worth from its location, distance and shot type. Comparing actual points with xPPS separates shot quality (are we getting good looks?) from shot making (are we converting them?).

Formula: `Diff = actual PPS − xPPS`

## Players and lineups

### What is PIR in basketball?

The EuroLeague's official player rating, shown in every box score. It adds the positive box-score events and subtracts the negative ones, each counted one for one.

Formula: `(PTS + REB + AST + STL + BLK + fouls drawn) − (missed FG + missed FT + TOV + shots blocked + fouls committed)`

Example: Elijah Bryant, Hapoel Tel Aviv 102–98 Real Madrid, round 3 of 2026-27: 22 points, 13 rebounds, 7 assists, 1 steal and 9 fouls drawn (+52), minus 8 missed field goals, 2 missed free throws, 1 turnover and 1 foul (−12). PIR 40.

**PIR vs Clutch Data Score**: PIR barely rewards efficiency: a missed shot costs one point and a made two earns two, so 1 of 3 on twos breaks even. Defense counts only through steals, blocks and rebounds. CDS weights scoring by efficiency and measures defense from on/off, so the two rankings can differ.

[EuroLeague MVP race by CDS](https://www.clutchdata.net/euroleague/mvp-race)

### Usage rate (USG%)

The share of a team's possessions a player finishes, with a shot, a trip to the line or a turnover, while on court. It measures role, not quality: a high-usage player carries the attack, efficiently or not.

Formula: `Player (FGA + free-throw trips + TOV) / team possessions, with the player on court`

[EuroLeague advanced stats, updated every round](https://www.clutchdata.net/euroleague/advanced-stats)

### On/off

The team's ratings with a player on court against with the player off. It captures what the box score misses, such as spacing, screens and help defense, but it also carries who plays alongside, so we read it for duos and trios too.

Example: Baskonia 2024-25: −7.0 Net Rating with Markus Howard on court and +3.0 with him off, a 10-point swing per 100 possessions. His teammates' archetypes explain most of it.

[The full case](https://www.clutchdata.net/methodology#case)

### Player archetype

An extreme playing style found in a league's data, such as Shooter Specialist or Rolling Big. A player is described as a mix of archetypes that adds up to 100%, not as one label, so a guard can be half High Volume Scorer and a third 3&D Wing in the same season.

Example: Dylan Osetkowski, Unicaja 2023-24: 57.0% High Usage Forward, 23.2% Role Guard, 15.2% 3PT Specialist, 4.5% Defensive Specialist.

[The method, from our MIT Sloan paper](https://www.clutchdata.net/methodology)

### Projection range

The 10th to 90th percentile of a projected stat: in eight cases out of ten the real number should land inside it. A narrow range means a reliable projection; a wide one means the destination league is hard to read from the player's current one.

## Game context

### Garbage time

The minutes of a game that is already decided, when benches empty and the effort changes. Leaving them out keeps blowouts from distorting efficiency. In our data a game is in garbage time with a margin of 20 points or more from minute 30, 15 from minute 35 or 10 from minute 38.

### Clutch time

The last five minutes of the fourth quarter and any overtime, with the score within five points. There are few such possessions per game, so clutch numbers need many games before they mean anything.

[EuroLeague advanced stats, updated every round](https://www.clutchdata.net/euroleague/advanced-stats)

## Clutch Data metrics

### Clutch Data Score (CDS)

Our one-number player rating: efficient offensive production plus defensive points saved, on one scale and cumulative over the season.

Formula: `CDS = EOP + DPS`

[EuroLeague MVP race by CDS](https://www.clutchdata.net/euroleague/mvp-race)

### Efficient offensive production (EOP)

Points and assists weighted by offensive efficiency, scaled so the league's total adds up to its points. Assists count more when they lead to made shots within 1.5 m of the basket.

Formula: `(AST factor × AST + PTS) × OE, with AST factor = 2 × the share of assists on made shots within 1.5 m`

### Defensive points saved (DPS)

Defense from on/off: the change in opponent eFG%, offensive rebound % and turnover % with the player on court against off, turned into stops and counted in the games the player played. Negative when opponents do better with the player on court.

Formula: `2 × defensive stops gained`

### Offensive efficiency (OE)

The share of the possessions a player finishes that end in a made shot or an assist. It can pass 1.00 when a player has more offensive rebounds than shots.

Formula: `(FGM + AST) / (FGA − ORB + AST + TOV)`

---

Clutch Data · basketball analytics consultancy · info@clutchdata.net · https://www.clutchdata.net/glossary