Last season Houston won 52 games and Dallas won 26. On Friday, Oct. 9, the two teams open their preseasons against each other at The Venetian Arena in Macao (8 a.m. ET, 8 p.m. local) as part of the NBA China Games. Neither team has played a game this season. The pick is Dallas Mavericks +2.5. Before the case for it, we need to state what this pick does not have.
The main storyline is Kyrie Irving. He has been cleared for the China Games with no restrictions after 19 months out, and this should be the first time he shares the floor with Cooper Flagg, the 2025 No. 1 pick and reigning Rookie of the Year. The two teams play again in Macao on Oct. 11, then open the regular season against each other in Houston on Oct. 21.
The number
- Selection: Dallas Mavericks +2.5 (spread) - Model probability: n/a - Market-implied probability: n/a - Edge: none published - Model confidence: none published
Our engine returned no data for this game. We have no model probability, no market-implied comparison and no quantified edge. This is an unusual entry for our ledger, and it should be read that way. The pick is logged and will be graded on the public record like every other. It is not supported by the model output that normally sits behind our selections. The reasoning below rests only on researched context about rosters and availability. Weight it accordingly.
Why the pick leans this way
This is a basketball game, so the usual framework of pitching, park factors and run totals does not apply. The engine supplied no totals, no streak data and no projected score, and we will not reconstruct any of them. The argument for +2.5 is structural, and it has three parts.
1. Houston is managing its returning veterans. Fred VanVleet missed all of last season after tearing his right ACL in September 2025. The stated plan is to manage his workload early. Steven Adams is expected to play limited minutes in his first game since January 18. When a 52-win team controls minutes for two key returners in an exhibition opener, its rotation can look less like its regular-season version than its record suggests.
2. The number is small. A 2.5-point line means Dallas covers by winning outright or by losing by 1 or 2. Preseason openers often feature deep benches and lineup experiments in the second half. In those conditions, a gap built over an 82-game season does not reliably show up in a single 48-minute game.
3. Dallas has a reason to play its headliners. Irving is cleared with no restrictions, and the Irving-Flagg pairing is the reason this trip draws attention. The 26-win season came with Irving out for the full year and Flagg in his rookie year. If both play meaningful minutes, Dallas fields a top end it never had last season.
That is the full case. It depends on context, not on a modeled edge.
The case against
The case against is substantial.
Irving may not play this game. He is set to play in only one of the two China Games, and we do not have confirmation of which one. If he sits Friday, the main reason to expect a different Dallas team is gone for this matchup.
Dallas is thin up front. Dereck Lively II is recovering from foot surgery. The availability of Morez Johnson Jr. and Aldama was uncertain heading into the trip. First-year head coach Dusty May has said he has "No Idea" what his starting lineup will be. A team unsure of its frontcourt and its starting five carries real downside, especially against Alperen Sengun, a two-time All-Star.
Houston's depth is real even with minutes limits. The core of Kevin Durant, Sengun and Amen Thompson is intact. Marcus Smart and Bogdan Bogdanovic were added specifically for depth. If VanVleet and Adams play reduced minutes, Houston can still fill them with experienced players. Dallas, with its frontcourt in question, may not be able to do the same.
The baseline gap is 26 wins. Last season's records were 52 and 26. Even after discounting for Irving's absence and Flagg's rookie year, that is a wide gap to close with 2.5 points.
Preseason variance cuts both ways. The same randomness that helps a short underdog can produce a double-digit margin if one coaching staff empties its bench earlier than the other. We cannot predict substitution patterns, and late-game minutes in exhibitions are largely beyond analysis.
How often does this pick lose? Without a model probability, we will not offer a figure. Any percentage we stated here would be invented. A reasonable reader should treat this as close to a coin flip with significant availability risk on the Dallas side, and possibly worse if Irving does not suit up.
---
21+ only. Predictions carry real risk of loss. Set limits in advance, and if it stops being fun, call 1-800-GAMBLER.