In the final 72 hours before opening night, names come off rosters in waves. John Poulakidas was one of them. A 1.98m guard, 23 years old, on a two-way contract, removed from the Dallas Mavericks system days before the first game. The transaction wire noted he had scored 16 points in a preseason game against the New Orleans Pelicans, that he had reportedly drawn interest from Olympiacos and Panathinaikos. Then he was cut.

I read that brief four times. Not because the content was complex, but because it was missing the thing that matters most. Eleven information points, not one carrying a traceable source. Every number tagged "Source: None." In my profession, that is reason enough to stop and check every word. The numbers in a contract do not lie, but the people who read them know how to conceal.

The real story is elsewhere. Not in the 45.5 percent three-point mark. Not in the 16 points against the Pelicans. It is in where the Dallas Mavericks sit against the salary cap, and what kind of player their three two-way slots are being allocated to. That is where the decision was actually made.

Context: The transfer market does not begin at the airport, but in the legal office filing cabinet

To understand this decision, you need to understand what kind of decision it is. It is not a trade. It is not a failed negotiation in the sense readers usually imagine — the phone call, the negotiator, the signatory.

It is a procedural act. A professional basketball league grants three two-way contract slots per team. Each slot is a scarce development asset. In the days before opening night, the entire league cuts down simultaneously to 15 standard contracts and 3 two-way deals. Hundreds of fringe players are released inside the same 72-hour window.

Poulakidas was on that list. The brief described him as having "attempted to secure a new contract, but this was not achieved." That is the reading of a wire reporter. The reading of an analyst is: his two-way slot is being reassigned to a different body.

One detail needs to be on the table before we proceed. Non-guaranteed deals — including two-way structures — typically guarantee in early-to-mid January. Cutting before opening night costs the team effectively nothing. This is the most important point in the entire operational equation: the financial cost of this decision is approximately zero. No penalty, no buyout, no money buried in the books. Just one slot freed.

And the Dallas Mavericks are not a team with room to waste slots. They are a taxpaying team constrained by apron mechanics. The exact position — first or second apron — requires further verification, but the consequences are clear: above the first apron, teams lose access to certain exceptions; above the second, further restrictions apply. Every salary dollar is scrutinized. Decisions become zero-margin decisions.

In that context, three two-way slots are not a place for experimentation. They are insurance.

Dallas' roster construction has one documented vulnerability: frontcourt availability. Its interior rotation carries a clear injury history. A rational front office facing slot pressure allocates two-way capacity to big-man insurance and defensive versatility — not to a third or fourth shooting wing in the queue.

That is the logic. But the logic only holds if Poulakidas truly is a replaceable archetype. And that is where the data starts to have problems.

Analysis: What the dataset says, and what it omits

The available numbers: 7.0 points, 1.3 rebounds, 14.0 minutes per game in preseason. Shooting splits: 50 percent field goal, 45.5 percent from three, 80 percent from the free-throw line.

I need to be direct: this is not an evaluation dataset. This is a highlight reel written as numbers.

The brief supplies exactly three figures that flatter a shooter — three-point percentage, free-throw percentage, field-goal percentage. And it omits the four that determine professional viability: assists, steals, turnovers, and any defensive measure. No minutes distribution, no matchup data, no information about the quality of opposition in the 16-point game.

That data selection is itself an analytical finding. Transaction briefs are typically built from box scores, and box scores surface whatever exists. If a player's best available summary line is 7 points and 1.3 rebounds, then the non-shooting contributions were likely near zero. Not because the writer was lazy. Because there was nothing to write.

1.3 rebounds in 14 minutes for a 1.98m wing is not a rounding error. It is a participation signal.

Normalized to per-36 minutes — an arithmetic illustration, not a measurement — he equates to 18.0 points and 3.3 rebounds per 36. The 18.0 points sounds fine. But 3.3 rebounds per 36 for a 1.98m wing sits below the league floor for a non-primary creator. League-wide, wings who survive on shooting alone still post roughly 4-5 rebounds per 36. Falling below that suggests limited physicality, low motor on the defensive glass, or both.

This is the number that likely drove the decision more than any shooting split.

And the 45.5 percent from three? This is where I need to be most careful, because it is the most abused number in transfer reporting.

Attempt volume is not reported. That is the critical gap. A shooter hitting 45.5 percent on 2 attempts per game is an entirely different story from 45.5 percent on 6 attempts, including movement threes off screens and dribble hand-offs. One is statistical noise. The other is a real signal.

I do not know how many he attempted. And the brief does not say. There are no junk rumors, only people who read rumors in a hurry.

There is a second layer of noise that any evaluator of fringe players must account for: preseason shooting percentages are systematically inflated. Defensive intensity is low, rotations are chaotic, coaching staffs are running auditions rather than game plans. Preseason efficiency for fringe players should be discounted toward league average, not taken at face value. For a specialist, a 45.5 percent preseason mark is closer to a neutral signal than a positive one.

And the 16-point game against the Pelicans? That is a textbook variance event. The brief does not specify the quality of opposition — rotation players, end of bench, or summer-league level. Single-game preseason explosions are the most over-weighted data type by fans and the most correctly ignored by front offices.

Now age. Poulakidas is 23.

For a wing whose value is skill-based rather than athleticism-based, the age curve is friendlier — shooters typically hold value into their early 30s. But for a prospect — and that is exactly what a two-way slot buys — 23 is late. Development slots are priced against the remaining upside of an 18-to-21-year-old. A 23-year-old undrafted wing is being evaluated as a finished product. And finished products must win on production, not projection.

Put it together: 23 years old, undrafted, one skill, unknown attempt volume, sub-floor rebounding, no defensive data. Dallas was not choosing between a good player and a bad player. Dallas was choosing between a 23-year-old shooter with no other professional skill, and a two-way slot that could hold a younger, more versatile body.

Nothing in the data contradicts that decision.

Contrarian angle: What the official story is concealing

Now I need to reverse the narrative frame the brief constructed.

The telling of "attempted to secure a contract but did not succeed" implies a competitive failure. A player who tried and was not enough. But the mechanics show this is a decision about scheduling and prioritization, made inside the league-wide cutdown window. The same decision. Two entirely different readings.

The first reading places the question on the player. The second places the question on roster construction.

And if you place the question on construction, there is a notable blind spot: Dallas sits in apron constraint, its frontcourt carries injury history, and its future draft stock was materially reduced after the February 2026 blockbuster trade. Their primary competitive advantage right now is rookie-scale surplus — cheap, controllable production from recent high picks. That is the only thing allowing a top-heavy payroll to function.

In that structure, each two-way slot is an insurance card. And if the frontcourt is the known weakness, allocating slots to big men is reasonable. But reasonable does not mean optimal. There is a scenario the brief does not consider: Dallas just released a 1.98m shooter with an 80 percent free-throw mark and a 45.5 percent preseason three-point line — a legitimate professional shooter profile — to take a slot they will fill with a lower-ceiling candidate. If that happens, this is not a zero-cost decision. This is a zero-cost decision quietly destroying a small asset.

That is the real risk. Not the risk of overpaying — but the opposite: a decision that costs nothing, and because it costs nothing, no one checks it.

But I cannot push the story toward "Dallas was wrong" either. To say that, I need attempt volume. I need defensive data. I need to know who the Pelicans game was played against. The brief supplies none of it. Three sources are never too many when a number decides someone's career. I have one source, and that source has no source.

There is one more point about the European context. The brief says he drew interest from Olympiacos and Panathinaikos. This is the least reliable part of the whole text — interpretive framing, not a transactional fact. But if true, it changes the entire reading. A shooter not good enough for an NBA rotation can still be a leading offensive factor in Europe. There, three-point shooting is scarcer, and the value of a shooting specialist rises. The most likely truth is not "he cannot play." The most likely truth is he can play, just not at NBA rotation level, and his value is highest in a league where shooting is scarcer.

That is the most important reframing. And it does not make Dallas wrong. It only makes the story more accurate.

Key takeaway

This archetype is the most replaceable in the league. Every offseason produces a fresh cohort of 1.96-2.01m catch-and-shoot wings. The marginal value of any single one is near zero. Dallas' decision is therefore not a talent verdict — it is a slot-allocation verdict.

But there is a question the brief did not ask, and it is the question I want to leave behind: if a 23-year-old, 1.98m wing shooting 45.5 percent from three in preseason was cut inside a 72-hour window at zero cost — why did no one in the system keep him as an affiliate, zero-cost reserve? One possible answer: the system is already crowded at that position, and a two-way slot is worth more than a shooter who cannot defend. The other, more plausible answer: no one checked the attempt volume.

I once moved faster than a phone call and paid for it in reputation. I know that feeling. And I know that in this market, the only way not to repeat it is to read the number one beat slower than everyone else, then ask: what could be distorting this number?

With Poulakidas, the answer is: the fact that it says nothing at all.

The next domino is not in Dallas. The next domino is in Athens — or in any league that recognizes an 80 percent free-throw mark and a stable three-point stroke is still a profession, just not one that pays in an NBA two-way deal.

The lesson lies elsewhere: when a team is apron-constrained and reallocating slots, every cut looks like a verdict. Most are not. But to tell which is which, you need attempt volume. And this brief did not give it to you.

The airport, the contract, and one phone call from a small club — that is how I surface the truth. This time, I only had the contract. No airport. No phone call. And that is why I will not sign my name to a rushed conclusion.

---

This article is based on the transaction brief regarding the Dallas Mavericks waiving guard John Poulakidas. All figures are quoted verbatim from secondary reporting; per-36 conversions and efficiency rates are arithmetic illustrations, not measurements. Dallas' exact apron position and the 2026-26 contract guarantee date require further verification.