Believe it or not, we too like to have some little fun around here. We’re not just doom and gloom fueled by sad Ryan Adams reviews all the time. So recently we spotted a new sus AI feature in the official NBA app, one that has hitherto been kept surprisingly quiet. At the time of this writing, the ‘Ask NBA (Beta)’ module can be located in the ‘Around the League’ section under the ‘Discover’ tab as part of the native iOS interface. Currently, it’s being described as a “chatbot application” by the NBA’s fine print lodged at the bottom of the start screen:

Critically, there is no explicit mention of artificial intelligence anywhere on the product’s real estate. Nonetheless, the initial chat-like layout, as well as the ‘questions to get started’ prompts, look a lot like your everyday large language model navigation interface. While one could assume those pictured are relatively low-stake Q&A items hardcoded as ‘if-then’ statements baked into the source of this chatbot, they do set this whole experience up as a quasi-‘ask me anything’ deal. At any rate, if it walks like a duck, it quacks like a duck to us. Yet, it does kind of baffle us how little information or fanfare there is out there about this, and how surprisingly low-key this has been kept by the league so far. Needless to say, we gave this thing a go.

It didn’t start off… great. Below is the chatbot tripping over a somewhat banal data retrieval question about what team had been doing best in games against opponents outside of their league conference, just a few days ago. Alas, the NBA chatbot returned a plain answer listing the two teams at the top of the Easter and Western conferences at the time, citing their overall winning record instead of the opposite-conference one. (For the record, we did ask Google’s Gemini AI the same question—you know, the cannibalizing top search result that is swallowing the SEO market whole—and it did fall short of reporting such a trivial data point too…).

In order to get over the jilted disappointment of our first query, we decided to get cute with some Sixth Man of the Year predictions about halfway through the season. Thankfully, our AI bot friend provided us with sounder insights, bundling some of the well-known usual suspects, Miami’s Jaime Jaquez Jr, Minnesota’s Naz Reid, and Atlanta’s Nickeil Alexander-Walker into its top 5 prediction. We also didn’t mind the neat written summary at the end, partly justifying the table analysis—the holistic approach addressing both media and betting markets feels like a faithful depiction of current day NBA zeitgeist chatter. However, its inability to report a precise points-per-game number for Orlando’s Anthony Black (on top of a completely absent field goal percentage…), continued to leave us perplexed.

We then decided we wanted to test the NBA chatbot’s reasoning oomph by posing a more nuanced—if fascinating—question revolving around potentially unsung ‘diamonds in the rough’ players. We asked for its analysis over what players have been boasting above-average efficiency impact ratings but low playing time. In other words, we were hoping to alley oop NBA head coaches who read this site with allegedly untapped potential on their bench. To drive its insights this time, the chatbot opted for the NBA’s Player Impact Estimate (PIE), a league metric purported to show a player’s percentage contribution to all game boxscore events in totality (points, rebounds, assists, etc.), calculated using a simple and straightforward custom formula:

PIE = (PTS + FGM + FTM – FGA – FTA + Deff.REB + Off.REB/2 + AST + STL + BLK/2 – PF – TO) / (Game.PTS + Game.FGM + Game.FTM – Game.FGA – Game.FTA + Game.Deff.REB + Game.Off.REB/2 + Game.AST + Game.STL + Game.BLK/2 – Game.PF – Game.TO)

This one definitely got intriguing. According to our friend the bot, recently waived San Antonio forward Riley Minix would have counted as an incredible secret ace up Mitch Johnson’s sleeve, had they kept him on the team. With a stunning 25.7 PIE in just three games played this season, the Vero Beach, Florida-native left an impact like no other in the two minutes and change he played on average in each of the contests. For comparison, league heavyweights SGA, Wembanyama, Antetokounmpo, and Jokić’s PIEs all currently fare between 21-23. Similarly, the number two secret weapon in the league appeared to be former Detroit Pistons shooting guard Colby Jones, until you realize he got waived by the franchise back in November last year, after just one league appearance.

Matter of fact, of the ten players listed by the chatbot as part of this query, only Indiana Pacers veteran TJ McConnell has played more than ten games this season (24). Aside from him, Marcus Sasser (Pistons) and Pete Nance (Milwaukee Bucks) are the only other players that have reportedly been featured in more than just a handful games hitherto. Frankly, we could have used a tad more discernment by our NBA stats partner, but hey it’s not like that bar was set awfully high from that first prompt… If anything, this answer confirmed TJ McConnell to be Indiana’s own version of a bench mob leader Tasmanian devil.

This brings us to our pièce de résistance. Just in case the coaching staff hadn’t thought about it, we took it upon ourselves to try to boost the Brooklyn Nets’ mid-season performance review. Just because. Naturally, we began by daring the NBA AI to predict the 2025-2026 regular season record for the New York franchise, to which it annoyingly clarified it’s only set up to “answer questions about NBA games, players, teams, stats, rules, schedules, league history, and on-court action”. Too bad none of these attributes contribute toward a team’s regular season record.

Next, we turned to a rose-colored glass half full, and in spite of the Nets’ 11-23 losing record as we’re typing this (matching a .324 win percentage) we wondered if there were any statistical categories the young Jordi Fernandez-coached team led the NBA in. Well, with a bolded emphasis on not, the ballers’ GPT regretted to inform us that the Brooklyn Nets do not lead the league in any major team statistical category. Ouch. And yet, Michael Porter Jr is hooping like his life depends on it. Thusly, we rolled up our sleeves and fairly flipped the query script, this time inquiring about potential boxscore stats the Nets might be worst in the league in.

Unfortunately for Barclays Center patrons, according to our chatbot the team does rank last in the NBA in both rebounds and points per game—not exactly two negligible impact metrics if you’re them. To add insult to injury, the AI agent assumed Knicks-esque features and doubled down on the analysis, by adding that the squad is “also near the bottom in field goal percentage (44.9%)”, although it clarified as consolation that “Indiana is slightly lower (44.1%)”. Similarly troublingly, NBA intelligence is quick to point out how “[t]he Nets also average 15.7 turnovers per game, one of the higher marks in the league but not last overall” (emphasis in original…). Pheeeew.

To round things off for the rebuilding team East of the East River in New York, we threw them a tactical lifeline by sculpting a more moderate and forcefully balanced query as a last-ditched effort. Pressing the NBA chatbot for the team’s strengths and weaknesses so far resulted in a more uplifting outlook as Brooklyn enters the All-Star break and starts to focus on the final part of the season. According to the NBA itself, the Nets are good to continue to sharpen their shotmaking from key high-scorers (the aforementioned MPJ, as well as a healthy-again Cam Thomas). They should also continue to rely on the ascension of younger bucks such as Day’Ron Sharpe and Russian rookie Egor Dämin. The former, jointly with longtime Net Nic Claxton, also helps lead the team’s rebounding improvements, “a marked step forward from last season’s struggles inside” per the app.

Simultaneously, the answer did not hold back and strongly reiterated the team’s inconsistent defense, high turnover rate, as well as early-season injury setbacks as grave issues still plaguing them. In addition to needing to whip up more offensive creation beyond Thomas (on top of MPJ, might we offer), the chatbot concluded by addressing the Nets’ lack of experience: “With five first-round rookies and several new veterans, [their] biggest limitation remains time—they’re still learning how to play together at both ends”. Incidentally, the sentence might well double as mad libs description for the state the NBA chatbot finds itself in. Much like the Brooklyn Nets, it’s keeping its development on the DL.

We’d like to thank you sincerely for taking the time to read this and we hope to feel your interest again next time.

AV