Calgary Flames Coach Consults ChatGPT After 3-11 Start

Disclaimer: This article is based on actual news from the real world – honestly! However, it has been sprinkled with a healthy dose of satire.

CALGARY — After fourteen NHL games of what team officials are calling “a developmental phase” and what statisticians are calling “mathematically humiliating,” the Calgary Flames have achieved a 3-11 record and pioneered a new frontier in sports analytics: asking a chatbot why they suck. The franchise, valued at $1.6 billion and currently performing like a $47 beer league team, sits at 3-11 after consulting ChatGPT, the language model that cannot solve 2+2 without a disclaimer. Initial results suggest the AI understands hockey approximately as well as the Flames play it.

Head coach Ryan Huska spent Monday evening in what he described as a “rabbit hole” with ChatGPT, the AI chatbot notorious for generating fake legal citations, declaring that Alaska is larger than Europe, and that eating a small rock daily is “probably fine.”

According to footage from the team’s behind-the-scenes show “The Chase,” a title that implies pursuit of something the Flames have not demonstrated they even want, Huska fed the language model five games worth of data, career shooting percentages, and what he called “a bunch of different things,” a phrase that in professional athletics inspires tremendous confidence when applied to billion-dollar coaching decisions and also typically precedes disaster. Pressed for detail, Huska did not elaborate, possibly because ChatGPT’s terms of service prohibit sharing conversations where the user begs a machine to explain why they’re bad at their job. 

ChatGPT’s analysis, generated by an AI that thinks basketball is played with a puck, was presented by Huska to the team during a locker room meeting typically reserved for actual hockey strategy.  ChatGPT projected the Flames would average 2.36 goals per game this season and also suggested that Elmer’s Glue would make for a great pizza topping. 

The Flames are now effectively being coached by a chatbot that cannot reliably count to ten, solve basic algebra, or accurately determine whether or not the Chicago Fire could refer to either a soccer team or “that thing that happened with the cow”. OpenAI’s GPT models have been documented to hallucinate statistics, invent nonexistent Supreme Court cases, and once told a user that submarines can fly if they go fast enough.

The technology is not designed for mathematical computation. It is a language prediction engine that learned everything it knows about overconfidence, racism, and misogyny from training exclusively on downvoted Reddit comments. 

This is what the model does when asked questions outside its training: it makes things up and delivers them with the confidence of a tenured professor. The Calgary Flames are now trusting this technology to fix their offense.

ChatGPT’s apparent conclusion, as interpreted by Huska, is that the Flames “need to score more goals.” This revelation required artificial intelligence in the same way that breathing requires a manual. 

Huska stood before his team, armed with ChatGPT’s findings and the look of a man who had asked the universe for answers and received a loading screen. “For our team in here, love you all, genuinely hope you know that, but we don’t have a 60 fing goal guy in here,” Huska told the players, an observation anyone could have gleaned from payroll data, scouting reports, or simply watching the team play. 

He continued with what behavioral psychologists might call ‘tough love’ and what employment lawyers might call ‘creating a hostile work environment’. “Don’t accept it, I guess, is my message to you. People say the Calgary Flames can’t score. F that. You can’t score because you’re fing okay with it!” 

Huska’s entire strategy apparently distilled to “shoot more,” an approach that has guided peewee hockey since its inception. Huska did not explain how one becomes less okay with not scoring.

The Flames’ decision to consult ChatGPT comes amid speculation that the team may be intentionally losing to secure a better draft position. However, sources close to the organization insist the losing is organic, requiring no additional effort or AI assistance. “We’re not tanking. Tanking implies a plan,” said one anonymous player. “We’re just historically bad. At this rate, we’ll finish last and somehow still pick seventh.” 

The distinction between tactical failure and authentic incompetence matters to exactly two groups: statisticians and philosophers. Both agree the Flames are historically terrible. Only one group cares why.

Huska insists the team can improve, though his confidence now appears to rest on the judgment of a language model that recently told a user that penguins are capable of flight and that the moon is larger than the sun, “depending on perspective.” The coach did not clarify whether ChatGPT had been asked for its opinion on line changes, power play formations, or whether the Zamboni driver might be a better option in the net.

ChatGPT, when asked for comment, generated a response explaining that hockey involves scoring goals and that the Flames should “consider implementing strategies that maximize offensive production while minimizing defensive vulnerabilities.” It then offered to provide motivational quotes, recalculate the statistics using different parameters, or write a haiku about perseverance. 

ChatGPT has projected three possible outcomes for the next Flames game: victory, defeat, or the game ending in a tie, which has been impossible in the NHL since 2005. When informed of this, the AI apologized and offered to explain overtime rules, suggesting its hockey knowledge may be comparable to the Flames’ playoff chances.

The NHL has not formally addressed AI coaching, though a spokesperson confirmed the league is “aware of the situation and monitoring developments.” Translation: they’re deciding whether this violates rules that don’t exist yet. 

This story is based on fully factual news, but if we got it wrong, blame these guys, we’re just here to make it funny.