The Value of a Possession: Points Per Possession and Why Per-100 Wins

Basketball, stripped to its frame, is a game of trades. You get the ball; you try to score; you hand it back. Each team gets roughly the same number of those turns, and the winner is simply whoever squeezes more points out of them. Once you start thinking in possessions instead of points-per-game, the whole sport snaps into focus — and a lot of numbers you thought you understood turn out to have been lying to you.

What a possession is worth

A possession is one team's turn with the ball: it starts when you gain control and ends when you give it up — a made shot, a missed shot the other team rebounds, or a turnover. The currency that matters is points per possession: total points scored divided by the number of possessions used. League-average sits a bit above one point per possession — call it the neighborhood of 1.1 to 1.15 in the modern game. That's the going rate. Every possession is, on average, worth a little more than a point.

That average is the yardstick I measure everything against. A possession that nets you a layup (2 points) or a three (3) is a clear win over the 1.1-ish baseline. A possession that ends in a turnover is worth exactly zero — not a missed-shot zero, where at least the ball went up and might have been rebounded, but a stone, total zero. Hold that thought, because it's why turnovers hurt more than they look like they should.

Pace versus efficiency: the lie in per-game stats

Here's the trap. A team scores 120 points a night and you think, great offense. But points per game is the product of two completely different things: how many possessions you get (pace) and how much you score per possession (efficiency). A team can hit 120 by being genuinely efficient, or by playing fast and cramming so many possessions into 48 minutes that even mediocre per-possession scoring piles up to a big number.

Those are not the same team, and conflating them is one of the oldest mistakes in reading a box score. A fast team and a slow team can have identical offenses on a per-possession basis while their per-game point totals look a full continent apart. Pace inflates or deflates every raw counting stat indiscriminately — points, rebounds, assists, all of it. I dug into exactly this in pace and possessions: the 120-a-night team might just be running, not cooking.

Why per-100 framing fixes it

The cure is to stop measuring per game and start measuring per possession — or, because decimals per single possession are awkward to read, per 100 possessions. That's all offensive and defensive ratings are: points scored and allowed per 100 possessions. The "100" is just a readable scale; 1.12 points per possession becomes a tidy 112. I walk through the mechanics in offensive and defensive rating, but the why is the part that matters here. Per-100 framing strips pace out of the comparison entirely. It puts the sprinting team and the grinding team on the same footing and asks the only fair question: when each of you has the ball, what do you do with it?

This is why I almost never quote a team's points per game as evidence of anything. It's a number contaminated by pace. Offensive rating is the clean version — the same statistic with the tempo washed out. Once you've made the switch, going back to per-game feels like reading a thermometer that hasn't been calibrated.

How turnovers and offensive boards bend the count

Now the subtle part, the thing that separates a casual read from a real one: possessions are not handed out in equal, fixed batches. Two things move the count, and both are levers a team can pull.

Turnovers don't just waste a possession — they're the worst possible use of one. A missed shot still has value baked in: the ball is live, your big might tip it back, the possession isn't necessarily dead. A turnover surrenders the ball with zero points and, often, zero chance at an offensive rebound — and frequently it gifts the other team a transition possession, which is the most efficient kind there is. So a turnover is worse than a brick twice over: you scored nothing, and you may have handed the opponent a above-average possession on the other end. That's why turnover rate is one of Dean Oliver's four factors, weighted heavily. Every one you commit is a possession converted to a guaranteed zero.

Offensive rebounds do the opposite — they buy you extra cracks at the same possession. Grab your own miss and the possession continues; you've effectively manufactured a bonus attempt the other team never got. This is why the bookkeeping uses "possessions" carefully: a trip with three offensive rebounds and four shot attempts is still one possession in the accounting, which is the right way to count it, because efficiency should reward turning one turn with the ball into multiple scoring chances. The crash-or-retreat math behind that is its own balancing act — I get into it in offensive rebounding and second-chance points — but the principle is clean: an offensive board is a stolen possession, taken from no one.

Put those together and you see why two teams that ran the same number of plays can use a different number of possessions, and why the per-possession lens is the honest one. Win the turnover battle and the offensive glass and you've quietly tilted the possession count in your favor before a single shot is graded.

It all rolls up into net rating

Here's where the possession view pays off in one number. Subtract a team's defensive rating from its offensive rating — points scored per 100 minus points allowed per 100 — and you get net rating, the cleanest single read on team quality there is. It's the scoreboard margin, pace-neutralized, expressed per 100 possessions. A net rating of +6 means that for every 100 times each team has the ball, you come out six points ahead. Over a season, net rating tracks winning about as tightly as any one number does, which is exactly why I trust it.

And every thread in this piece feeds it. Score above the ~1.1 baseline and your offensive rating rises. Hold opponents below it and your defensive rating falls. Protect the ball and crash the glass and you wring extra value out of the possession count on both ends. Net rating is just the possession framework, fully assembled: the sum of doing more with your turns than the other guy does with his. That's the whole game, and it always was — possessions are simply the unit that lets you finally see it.

The takeaway

Stop counting points per game and start counting points per possession. An average turn with the ball is worth a touch over a point; everything good a team does is an effort to beat that number on offense and suppress it on defense. Pace inflates the raw totals and per-100 framing washes it back out. Turnovers torch possessions; offensive rebounds mint new ones. And when you stack it all up, you get net rating — the one number that knows the difference between a team that's actually good and a team that just plays fast.