Warriors' 0-3 and Denver's 23-Point Comeback: Two Numbers That Contradict Each Other

In Phoenix on Saturday night, three seats on the Golden State Warriors bench were empty by design. Stephen Curry, Draymond Green and Brandin Podziemski — all three out, all three on the same list, all three under the same decision. The scoreboard recorded a nine-point loss and the record ticked to 0-3. The number that had every headline by morning was 0-3. The number that stopped me was nine — the margin produced without three rotation players, on the second night of a road back-to-back, across 48 minutes.

That margin says little about the team's true level and a great deal about its load-management doctrine. What happens in Denver on Monday will be the product of that doctrine.

I still have the Chennai Slam chart — 1,180 possessions from 2026, hand-coded off a buffering web stream, one row per pass. That spreadsheet taught me what every season repeats: the scoreboard is not the final truth, it is the cheapest available summary.

Context: what this preview contains, and what it does not

The material in front of me is a Thomson Reuters wire-service game preview, datelined October 26, year not stated. My first task is always the same: count what is missing. There is no offensive rating here, no defensive rating, no net rating, no true-shooting figure, no salary number, no front-office sourcing. What exists is volume — scores, points, comebacks, and quotes attributed to two head coaches.

Every conclusion below therefore has to be read at that level. Where I cannot build the number, I will write 'data pending verification.' That is not a weakness in the copy. It is the method.

The season identifies itself from internal evidence: David Adelman is named as Denver's head coach, Podziemski sits in Golden State's rotation, and October 26 falling on a Sunday is consistent with a Monday game. That points to the opening of 2026-26 — medium-high confidence, not absolute confirmation.

Now the two teams. Denver beat New Orleans 141-135 in overtime after trailing by 23 on its own floor. Nikola Jokić posted 35 points, 15 rebounds and 10 assists — the 199th triple-double of his career. Jamal Murray scored 42. Golden State blew a late lead in its home opener, then lost by nine in Phoenix. Cameron Johnson has a sprained right ankle and is listed doubtful.

That is the entire ledger a wire service hands me. I will not ask for more, and I will not be satisfied with less.

Core analysis: three numbers, three different truths

First number — nine.

Reading a nine-point loss in Phoenix as proof of a team's level is the mistake every October repeats. Golden State played its home opener on Friday, then travelled to Phoenix on Saturday: the second night of a road back-to-back. Curry, Green and Podziemski were held out of exactly that game. Resting three rotation players together is not accident, it is a statement of priority: one October road game was judged less valuable than Monday in Denver and the season that follows.

The crucial point is that the rest is now level — both teams played Saturday — but the travel and altitude are not. Golden State goes Phoenix to Denver, roughly 1,100 metres of elevation, on the second leg of a trip. Denver has been home since Saturday. That differential lands disproportionately on a 37-year-old primary ball handler. No official rating exists for it, but what is visible can be counted: how long Curry's legs hold the same speed in the third quarter.

Second number — the home opener.

The loss carrying real tactical information is not Phoenix, it is the one at home. The game in which the core was available and a late lead was surrendered. That is the signature of a very specific illness: half-court, late-clock creation. Motion offence looks beautiful until a defender switches; after the switch the only question left is who holds the ball with eight seconds on the clock and what he builds. Nothing in this preview suggests Golden State has installed an alternative for that possession. In my own working language I call it the absence of a Plan B — the most expensive absence there is, because it only surfaces in April.

Third number — 141.

Those 141 points were produced across 53 minutes. Overtime volume does not translate into regulation capacity; treating it as if it does is a translation error, and the most common one in this sport. Likewise, trailing by 23 at home cannot be the defensive signature of a contender. The comeback is proof of resilience; the 23-point hole is proof of an intensity deficit. Both come out of the same game, and holding both together is the job. The piece that reports only the comeback reports half a fact.

A fourth number nobody counted — seven.

Denver rookie Brazile scored seven of his 17 in overtime. That is not garbage-time output. It is a five-minute, high-leverage window, and trusting a first-year player inside it is the clearest statement a head coach can make. Adelman made it. If it holds, Denver's playoff rotation gains an option that did not exist when the season opened. That is the kind of signal I find more interesting than the score.

Yaxel Lendeborg: the number doing the most lying

18.7 points and 8.3 rebounds across three games. On the surface, a superb rookie start. Beneath the surface, three distortions are compounding.

First, the three-game sample. Second, a 0-3 team with no established hierarchy, where shots do not distribute so much as scatter. Third and heaviest: at least one of those games was played with the primary and secondary usage options both unavailable. His 26 points in Phoenix were a team high — in the same game Curry, Green and Podziemski sat. His usage rate is artificially inflated.

I have no true shooting, no effective field-goal percentage, no plus-minus, no on-off. Here it must be said plainly: the data is insufficient. Anyone forming a judgement on 18.7 points without an efficiency figure is reading a scoreboard, not a game.

And Monday's assignment is a deliberately constructed stress test for this rookie. Denver's frontcourt — Jokić and Aaron Gordon — is the worst possible pairing for a rookie forward, because it sets two different examinations at once. Gordon tests strength and athleticism; Jokić tests processing speed, how fast you read where the help is coming from. One attacks the body, the other attacks the mind.

Steve Kerr's public line — that Lendeborg has to figure out how defences are playing him — is a coach pre-announcing an adjustment period. I do not think that is accidental. When a coach says publicly that a player must 'figure it out,' he is preparing the ground for an inefficient night. Not malice. Management.

And this is where an old rule of mine applies: name the role, name the mechanism, then name the number — never the reverse. Lendeborg's role right now is a cutter and second-chance creator who can finish over a smaller defender after a switch. His mechanism is cutting from depth and attacking the offensive glass. His number today is 18.7, and it may well be 12 once Curry returns. Both can be true. What cannot be true is the claim that three games constitute proof of anything.

Cameron Johnson's ankle: small injury, large geometry

A floor spacer going down looks like routine attrition. For Denver it is a change in geometry. Johnson is a perimeter threat, which means his presence keeps Murray's driving lanes open and forces defences to honour Jokić's handoffs. Without him the lanes compress, and in a compressed lane Murray has to spend more physical currency per possession.

That is precisely where Brazile's minutes suddenly matter. A rookie's minutes are usually filler; here they are structural necessity. The status is 'doubtful,' meaning the decision lands before tip-off — and that decision will reshape Denver's spacing.

Contrarian angle: 0-3 is not the truth either, and neither is 141

Here I have to turn on my own argument, because that is the only honest route available.

I just said 0-3 is a rest artefact. That is correct, and it is incomplete. Correlation and causation are different objects. The fact that three players sat does not tell us what the team looks like with them. It tells us only that on that specific night, a nine-point margin was a poor proxy for team level. If I dismiss 0-3 entirely, I commit the same error as the other side — just pointed the other way.

And the same discipline applies to Denver. Winning from 23 down has become a heroism story. On my sheet it is two things at once: the capacity to win in overtime, and the habit of falling 23 behind at home. The second is more stable than the first. Teams that win from comebacks tend to manufacture the need for comebacks. In the regular season that reads as thrill; in April it reads as fatigue.

The shared blind spot is that three October games are being treated as playoff-transferable. They are not. Denver is still gelling under Adelman; Golden State is integrating a rookie into a genuine role. Drawing playoff conclusions from a three-game sample is the exact error I made in Chennai, when I built a model from 1,180 possessions across 14 Chennai Slam games — a model that was worth nothing the following season. That chart outlived the league it was built to explain, and that is not a compliment, it is a warning.

The third blind spot is the cheapest one: the Lendeborg story. 18.7 points across three games is a very good headline and very bad evidence. If he shoots 2-for-6 on Monday, the story will not change — only the headline will. What will change is how many shots he generated and how many he passed up with Gordon's chest in front of him. That number no wire service provides. I have to count it myself, possession by possession, and log it in a spreadsheet the way I did in Chennai.

My own limit is equally clear here. I have NBA play-by-play. I do not have a possession-level record of this preview's underlying games, and I will not invent one to make the piece tidier. What I can write is this: the most reliable thing I know about this matchup is the public language of the two coaches — Adelman's trust in a rookie, Kerr's pre-declared adjustment period. Everything else is inference, and labelling inference as inference is the minimum entry requirement for this trade.

And I trust the possession count more than the final whistle.

Forward look: three things to watch on Monday

One — Denver's first quarter. If they repeat the New Orleans start, the 23-point hole is habit rather than accident, and their 141 stops being evidence of anything. If they open with defensive intensity across the first ten minutes, I will soften the comeback narrative.

Two — Lendeborg's first-half shot quality. Not points. Quality. How often he needed help against Gordon, how often he passed from the wrong spot. That is the dataset I will build, and it will define his role by December.

Three — Adelman's final five minutes. If Brazile is on the floor in that window again, an October experiment becomes a December policy.

And the thing that could falsify this entire piece is equally clear: if Golden State plays fully healthy on Monday and still has no half-court answer in the last four minutes, my rest-artefact argument collapses. In that case I will correct myself and publish the correction — because a chart that will not let itself be proven wrong is not a chart, it is advertising.