Why NBA Cap Totals Exceed the Active Payroll

A fan adds up the names on a team’s salary page and gets $145,000,000. The summary above the table says $171,000,000. Before anyone accuses the front office of hiding another expensive center, there is a more useful question: do those two numbers include the same things?

A player list is a view of a roster. A cap total is a view of accounting entries. Their difference can explain an important roster constraint, reveal a missing category, or expose a stale source. The arithmetic alone cannot tell you which explanation applies.

Understanding that distinction changes how you evaluate a signing, a waiver, and a summer plan. It also gives you a practical way to challenge a salary page without assuming every mismatch is an error.

This guide builds that method around Team North and Team South, two fictional clubs. Every team amount, transaction, and example cap line below is hypothetical. None represents a current NBA team’s books or an available contract offer.

Start by identifying what each number measures

For this guide, active payroll means the cap charges attached to the current roster contracts included in a particular table. It does not mean a count of players available tonight. Before using an outside table, check its definition: salary paid, listed salary, and cap charge are labels with different possible scopes.

Total cap is especially dependent on the publisher. One page might total contracts and former-player charges. Another might include holds and other accounting entries. A third might offer several totals for different purposes.

The underlying NBA framework allows team salary to include former-player obligations, free-agent amounts, unsigned first-round picks, and incomplete-roster charges. Exceptions can authorize spending above the cap. Tax and apron calculations also have their own definitions. Those principles explain why one roster sum cannot answer every spending question. CBA, Article VII, Sections 2, 4 and 6

Treat a table’s labels as questions to resolve before reaching a conclusion. The first exhibit is a reading checklist, not a claim that every publisher uses these terms identically.

A label without a definition is a lead to investigate. It is weak evidence for a precise claim about what a front office can do.

Reconcile the difference before explaining it

North’s fictional page lists $145,000,000 in active contracts and $171,000,000 in total cap. Subtracting gives an unexplained difference of $26,000,000. That difference is the starting point of the investigation.

Suppose the underlying records supply the following categories. The example assumes each amount belongs to the same season and snapshot, and that the categories do not overlap.

The $26,000,000 gap is now explained: $8,000,000 plus $14,000,000 plus $4,000,000. Calling the entire gap dead money would overstate that category by $18,000,000. Calling it an arithmetic error would miss the fact that both totals reconcile exactly.

The phrase “other documented charges” is deliberately conditional. A real reconciliation needs the actual entries and their definitions. You cannot fill a missing bucket with whatever amount makes the final sum work.

Suppose instead the records explain only $24,000,000 of the difference. The remaining $2,000,000 stays unresolved. It might reflect an omitted entry, a timing difference, or a problem in the data, but assigning it a confident explanation before checking would turn arithmetic into speculation.

This is also why a complete-looking player table can be incomplete for your purpose. A list may correctly show every contract within its filter while leaving out entire categories that appear in the summary above it.

A waiver can change the roster without lowering the total

Guaranteed salary can remain after a waiver. A buyout can reduce the obligation; stretching eligible waived salary requires a separate election. An ordinary trade assigns the contract rather than automatically creating dead money. Those are different events and should be recorded separately. CBA, Article VII, Sections 4(a), 4(c) and 7(d)

Take a simpler version of North’s ledger: $145,000,000 active, $8,000,000 dead, and no other charges. Its total is $153,000,000. Assume North waives a player whose already-counted $3,000,000 charge remains fully owed after he clears waivers, with no separate adjustment.

The roster loses a player and the active column falls. The total does neither. Anyone using only the active column would mistake a change in classification for new spending capacity.

Now assume a permitted replacement signing adds a $2,000,000 charge without removing another charge. North reaches $144,000,000 active plus $11,000,000 dead, or $155,000,000 overall. It has more active salary than immediately after the waiver, less than before the waiver, and a higher total than at either earlier stage.

That sequence explains why “they cut a $3,000,000 player and signed a $2,000,000 player” is insufficient to establish savings. The decisive detail is what happened to the original obligation.

A transaction analysis should therefore include a before-and-after ledger. The outgoing player’s name disappearing from the roster tells you something about the basketball plan. The related accounting entries tell you whether that decision also changed the spending position.

Separate a smaller obligation from a longer timetable

Buyout and stretch headlines invite a second kind of confusion: treating a reduced amount and a redistributed amount as the same benefit. They answer different questions even before any detailed eligibility rule enters the discussion.

Consider an abstract obligation of $9,000,000. If an agreed reduction leaves $7,500,000 owed, the amount has fallen by $1,500,000. If the original $9,000,000 is instead spread across additional periods, the timing has changed while the total remains $9,000,000.

Those examples illustrate arithmetic, not a proposed NBA stretch schedule. To evaluate an actual transaction, the verified cap allocation must replace the abstract timetable.

For roster planning, the distinction matters because a smaller present charge can have two very different explanations. One leaves less to account for overall. The other can leave charges on future sheets that otherwise would have been clear.

The useful follow-up is to compare every affected season. A single attractive number in the first column cannot establish the full cost of a decision whose consequences extend beyond that season.

A hold can explain apparent room without being a new signing

Suppose South’s illustrative ledger contains $126,000,000 in active charges, $5,000,000 in dead charges, $18,000,000 in holds, and $3,000,000 in other documented charges. Its total is $152,000,000. Against our invented cap line of $160,000,000, the displayed difference is $8,000,000.

A reader who subtracts only active payroll from that cap gets $34,000,000 instead. That is an overstatement of $26,000,000 relative to the full ledger. Nothing about the subtraction is mathematically wrong; the input answers a different question.

Now examine a deliberately isolated accounting scenario. Assume a documented event removes one $12,000,000 hold and, after all other adjustments, the remaining entries stay unchanged. South’s total becomes $140,000,000 and the difference from the invented cap becomes $20,000,000.

That calculation describes the specified scenario. It does not establish that removing the hold is available, sensible, or consequence-free in a real case. A responsible projection names the action and checks everything else it changes before presenting the result as usable room.

The same discipline applies to a new contract replacing a placeholder entry. In an abstract ledger, replacing a $12,000,000 entry with a $15,000,000 entry increases the total by $3,000,000. Adding the full $15,000,000 while retaining an entry that should have been replaced would overstate the result by $12,000,000.

The essential question is whether a transaction adds, removes, or replaces a line. That small distinction can explain a larger mismatch than any error in adding the players’ salaries.

Cap space is a comparison, not a bank balance

North’s simplified $153,000,000 total sits $4,000,000 above an invented $149,000,000 cap. Writing the result as negative $4,000,000 is just one way of displaying that relationship. It does not measure the franchise’s cash or the owner’s willingness to spend.

It also cannot, by itself, approve or reject a proposed contract. A useful basketball argument needs to identify the mechanism for that specific move, then check the relevant constraints. The negative number is one input in that investigation.

There is an equally common mistake on the positive side. A page showing $20,000,000 of theoretical room can become a headline saying the team will sign a $20,000,000 player. Between those statements sit assumptions about the source, pending decisions, remaining roster work, and the proposed transaction.

Express those assumptions openly. “This scenario shows a $20,000,000 difference after the specified entries change” is reproducible. “The team has $20,000,000 to spend” may hide every decision needed to reach that number.

Even a correct spending calculation leaves the basketball question unresolved. If an imagined plan fills one position while leaving three rotation roles unaddressed, showing that the first move fits is only the beginning of evaluating the plan.

The cap, tax, and aprons are separate reference lines

The NBA’s 2025-26 announcement set the cap at $154,647,000, the tax level at $187,895,000, the first apron at $195,945,000, and the second apron at $207,824,000. These are historical figures for that season, not current thresholds. Official 2025-26 announcement

Their separate labels matter. A distance from one line is not a distance from every line, and a single generic total may not be the appropriate input for all four comparisons.

To see the arithmetic without confusing historical and current data, return to a wholly fictional threshold card. Assume an example has already supplied a distinct, correctly defined salary input for each calculation.

These figures describe four distances. They do not produce a tax bill, authorize a trade, or establish exception eligibility. Their purpose is to show why “over the cap” cannot be casually substituted for “over the tax” or “over an apron.”

Notice that the hypothetical inputs differ as well as the thresholds. Reusing $170,000,000 in every row would yield tidy arithmetic while ignoring the premise that each test has its own defined input.

This is a good test of any confident salary claim: which line is being discussed, and which number was compared with it? If either is missing, the conclusion is difficult to evaluate.

Compare seasons without mistaking a new column for savings

North records $153,000,000 this season and $149,000,000 in the next season’s preliminary ledger. The headline difference is a $4,000,000 decline. That sounds like a simple cut until the categories are separated.

In this example, the active group costs more while an older obligation shrinks. Saying North reduced its active spending would be false. The total fell because a $6,000,000 reduction elsewhere more than offset the $2,000,000 increase.

The next question is whether the future column is complete enough for comparison. If one column includes a completed roster and the other includes only already-recorded commitments, their difference should be described accordingly.

For an illustrative sensitivity check, assume North later adds $10,000,000 of net charges to that future ledger. The projected $149,000,000 becomes $159,000,000. What looked like a $4,000,000 decline becomes a $6,000,000 increase relative to the current $153,000,000.

Neither result is mysterious. One describes the preliminary entries; the other includes an additional assumption. Good analysis keeps the recorded amount and the planning assumption visible instead of blending them into a falsely precise forecast.

Diagnose data problems before drawing roster conclusions

A perfectly sensible explanation for some mismatches is that two parts of a website have different update times. Suppose North’s summary includes a new $4,000,000 charge while its player table has not yet added the related row. The visible discrepancy may disappear when both views reach the same snapshot.

That is a possibility to verify, not an excuse to apply automatically. Record the season, filters, and update time, then identify the entry that accounts for the difference. If it cannot be found, keep the discrepancy open.

A useful review proceeds in a fixed order. First align the season and date. Next remove display filters and confirm what the table includes. Then add the documented categories and compare that sum with the reported total.

Only after those checks should you investigate a remaining discrepancy as a likely record problem. Starting with an elaborate transaction theory before checking the season selector wastes effort and can produce an impressive explanation of the wrong dataset.

Rounding deserves its own check. If fifteen displayed rows are each rounded to the nearest $100,000, each row can differ from its underlying figure by up to $50,000. The sum of the displayed rows could therefore be as much as $750,000 away from the sum of the exact values.

That is a theoretical bound, not an assumption that every row rounded in the same direction. It explains why a small disagreement between abbreviated totals deserves an exact-value check before an accusation of missing salary.

Missing values require different treatment. A blank dead-money field gives no evidence that dead money is $0. Substituting zero makes a calculation possible while concealing the fact that one of its inputs was unavailable.

Likewise, a missing tax figure cannot establish that a team owes nothing. The honest output is an incomplete comparison until the relevant value and its meaning are available.

Turn the salary sheet into a better basketball argument

Imagine two fictional teams each show a $171,000,000 total. North’s ledger includes $145,000,000 active and $26,000,000 elsewhere. South’s includes $165,000,000 active and $6,000,000 elsewhere. The headline totals match, but the composition raises different questions.

For North, investigate what sits outside the active group and how long it remains. For South, inspect which roster contracts account for the larger active commitment. Neither pattern proves that one front office is better or that one team has more useful options.

Performance, contract duration, and the specific entries still matter. An expensive active player could be central to a contender or difficult to fit into its rotation. A charge outside the active group could have a near-term end date or persist across several planning cycles.

The cap sheet gives those basketball judgments a structure. It helps separate what the team has committed, what the displayed roster explains, what assumptions drive a projection, and which unanswered question could change the conclusion.

Use the team cap tracker to identify the comparison you want to investigate and the contract book to examine individual entries. Save the season and snapshot alongside any calculation. A number detached from its definition and date is easy to share and difficult to defend.

The most useful conclusion is not that a roster looks cheap or expensive. It is an explanation of which entries create the constraint, which verified action could change them, and what the resulting team would still need on the court.

Sources and methodology

Rules were checked against the 2023 NBA-NBPA Collective Bargaining Agreement, linked by the NBPA’s CBA page. Relevant provisions are Article VII, Sections 2, 4, 6 and 7(d). Historical thresholds come from the NBA’s June 30, 2025 announcement.

All North/South ledgers, hypothetical thresholds, transaction amounts, and calculations were created for this guide. They demonstrate reconciliation and scenario analysis, not actual transaction eligibility. This guide makes no calculation of a real team’s current room, tax liability, or available exceptions.