Game Score Calculator — GmSc

John Hollinger built Game Score to answer the bar-room question — “how good was that performance, really?” — with one honest number from the box score. Enter a player’s line, and this calculator returns the Game Score (GmSc): 10 is an average night’s work in the league, 20 is excellent, 40 is a performance people remember by its date.

Hollinger's Formula, Term by Term, Explained

Game Score reads like an accountant's summary of a basketball night: PTS + 0.4·FG − 0.7·FGA − 0.4·(FTA − FT) + 0.7·ORB + 0.3·DRB + STL + 0.7·AST + 0.7·BLK − 0.4·PF − TOV. Its genius is what it charges. Scoring counts in full, but every attempted field goal costs 0.7 — the price of the possessions consumed — while each make refunds 0.4, so efficient shooting is rewarded at the level of arithmetic rather than rhetoric: a two-point attempt's expected value runs 2.4p − 0.7 at shooting probability p (points plus make bonus minus attempt cost), crossing zero near 29% — generous in isolation, which is precisely the point: the formula rewards taking decent shots and saves its punishment for the volume that drags a line's efficiency, through the raw count of attempts, rather than for ordinary misses. Offensive rebounds (0.7) are priced above defensive ones (0.3) because they manufacture new possessions; assists and blocks take 0.7; steals a full point; turnovers cost a full point and fouls 0.4. No minutes, no pace, no opponent — deliberately: it is a per-game production index, the box score's own opinion of itself.

Worked Example: A Star's Solid Night

The defaults run a strong all-round line: 25 points on 9-of-18 shooting, 6-of-7 from the line, 2 offensive and 6 defensive rebounds, 5 assists, a steal, a block, 3 fouls, 3 turnovers. The arithmetic: 25 + 3.6 (makes) − 12.6 (attempts) − 0.4 (the one missed free throw) + 1.4 + 1.8 (rebounds) + 1.0 (steal) + 3.5 (assists) + 0.7 (block) − 1.2 (fouls) − 3.0 (turnovers) = 19.8 — a tick below the "excellent" threshold of 20, which the calculator's band reports faithfully. The teaching value of the example is the sensitivity: remove two turnovers and the same night is a 21.8; take the same 25 points on 30 shots instead of 18 and it falls toward 11. Game Score's whole philosophy is in those two moves — the counting stats flatter volume, the coefficient structure invoices it.

The Scale of Greatness, Calibrated

The benchmarks anchor the intuition: league-average production sits near 10 per game (a full rotation's worth of average nights), an All-Star season averages near 20, and single-game immortality begins around 40 — a club whose famous members include the 100-point game's modern reconstructions and the great Finals performances, with the very highest recorded figures in the mid-40s to low-50s depending on the stat-keeping era's completeness. Two disciplines keep the number honest: Game Score is not pace- or minutes-adjusted (a 40-minute starter's 18 and a sixth man's 18 in 22 minutes are equal here and unequal everywhere else), and it knows nothing of defence beyond steals and blocks. Hollinger built it as a conversation tool with a statistician's conscience — one number, honestly derived, with its blind spots printed on the label, which this page's FAQs continue.

Fantasy Basketball and the Metric's Second Career

Game Score moonlights in fantasy sports, where its single-number summary prices players for daily formats and season-long trades — and its biases become strategy: it favours efficient high-usage stars over accumulators, punishes the empty-calorie volume scorer your league-mates overpay for, and ignores minutes so completely that a savvy manager pairs it with a per-minute lens before trading a starter for a bench spark. The formula's transparency is the fantasy virtue as much as the analytic one: every coefficient is on this page to argue with, which is more than the projection systems offer. A number you can recompute is a number you can reason about — Hollinger's bar-room gift to both sports.

Game Score Across a Career: Peaks and Averages

Historians of the sport use the metric at two magnifications, and the calculator serves both: the single-game peak (the 40-plus nights that anchor a legend's highlight reel and survive in record books) and the career or season average (the measure of a player's ordinary excellence, where the all-time leaders cluster in the low-to-mid twenties per game). The two views correct each other's propaganda — a compiler with one immortal night is a story; a star whose average night clears 20 for a decade is an era. When the argument turns to the greatest performances ever, insist on knowing which magnification is on the table; this page computes the unit both are built from, and the FAQ's scale keeps the two from being confused at the bar.

The Formula's Confessed Blindness

One limitation deserves its own paragraph because it shapes every argument the metric starts: Game Score sees only what the box score records, and the box score is nearly blind to defence beyond steals and blocks — the league's best positional defender can post a modest GmSc while erasing the opponent's best scorer, whose own line the erasure quietly ruins. Gravity, screening, hockey assists and the fouls drawn are likewise invisible here. Hollinger knew it; the metric was offered as a production summary, not a player verdict. The calculator repeats the confession so the number is used as its maker intended — the first word in evaluating a performance, never the last — and the reader who quotes it with its blind spot attached will win more arguments than the one who quotes it as gospel.

Related Tools

More from the Sports Calculators hub: the Magic Number Calculator for the standings side of the season, and the Calories Burned Weight Lifting Calculator for the training room behind the box score.

Frequently Asked Questions

What is a good Game Score, then?

By the scale built into the calculator: 10 is an average performance, anything in the high teens is a strong night, 20+ is excellent, 30+ is the best performance in most buildings that evening, and 40+ is history. Season averages compress the same ladder: rotation players live at 7–12, stars at 18–24, and the MVP argument happens above 25. Judge a single game against the single-game ladder and a season against the season's — the formula is identical, the meaning of the average is what changes.

How is Game Score different from PER, Hollinger's other metric?

PER (Player Efficiency Rating, also Hollinger) is Game Score's rigorous older sibling: per-minute, pace-adjusted, league-normalised to 15.0 every season, and computed from a far longer formula including team context. Game Score deliberately drops all of that to stay computable from one night's box score — which is its entire purpose on this page. Use GmSc to evaluate a performance, PER to compare players across teams and eras; confusing the two is like judging a single at-bat with a career batting title's standards.

Why does Game Score punish missed shots quite so much?

Because possessions are the game's finite currency: a missed field goal usually ends one, and the 0.7 charge is the opportunity cost of the teammates' shots not taken. The make bonus (0.4) plus the points themselves mean an average shooter still profits from shooting — the formula is not an argument for passivity — but a volume scorer on a cold night can sink below a quiet efficient teammate, which most coaches' eyes already knew and the arithmetic merely refuses to forget. Free-throw misses cost less (0.4 each) because the possession often survives them.

Can Game Score be negative?

Yes — the band table in the calculator says so plainly, and short, turnover-strewn, foul-plagued cameos reach negative figures in real box scores. A negative GmSc is the formula's way of saying the minutes would have produced more with the player seated. It is rare across a full starter's workload, where the counting stats accumulate faster than the penalties, but for a disastrous five-minute stint it is not even uncommon — one more reason the metric earns its keep as a summary rather than a compliment machine.

Does Game Score work for other basketball leagues and eras?

The formula consumes any box score that records its twelve inputs, so WNBA, college (with its 40-minute games compressing the totals — expect the same scale but lower averages), FIBA and historical NBA lines all compute. Cross-era comparisons carry the usual warnings: pace differences inflate modern counting stats, older box scores lack recorded steals, blocks or turnovers (entering zeros understates those players), and the three-point line changed what a "make" is worth in flow if not in this formula. Within a game, a season, a team — its home ground — the comparisons are sound.

Why are offensive rebounds worth more than defensive ones?

Because of what each one buys: a defensive rebound ends the opponent's possession, which the defence as a unit accomplished and the rebounder finished; an offensive rebound creates a new possession from a dead one — worth, by possession value, more than twice the defensive act. Hollinger's 0.7/0.3 split is that economics rounded for the bar room, and it is one of the formula's quietly correct judgements: the glass-cleaner who extends possessions is manufacturing the scarcest resource in the sport, and Game Score pays him accordingly.