I love sports. In fact, I’m obsessed with them. Whether it’s football, darts, UFC, NFL, if it’s on TV, I watch it. For a long time I have struggled with fast-paced American sports as they air late at night and the speed of the game makes me sleep trance.
A few months ago, a friend of mine invited me to a local ice hockey game and saw a puck drop in Scottish semi-professional hockey, I was absolutely hooked. Since then, I have seen Prime Video showdowns introducing professional North American hockey players. More importantly, I was hooked on the NHL.
Over the weekend I’ve seen all of the NHL games featured in British Social Times. I’ve been to most Edinburgh Capitals games to get both sides of the hockey spectrum. I was in such a love for the NHL, so I bought a ticket to go to Stockholm, Sweden in November and saw the Nashville predators face the Pittsburgh Penguins.
Anyway, you’re probably thinking to yourself, this isn’t about AI, this is just about ice hockey. But that’s where things get interesting. Over the past few weeks I have been trying to select NHL teams to show loyalty, but considering that I have covered AI for my work, I have seen whether it helps to cut down on hockey support options rather than just checking whether it helps me test AI research tools.
Last week or so, I used ChatGpt and Gemini to create research reports, fun quizzes, and even AI-generated podcasts to make decisions a little easier. Here’s how the process went, and how AI is best for such a fun project, but it’s not entirely reliable.
Personality check
Let’s draw a picture. I’ve been watching the NHL for nearly two months. During that time, I learned a lot about players, teams, and hockey rules in general. So I thought my first port should be a fun quiz for teams to decide what suits my sporting preferences.
I went to ChatGpt 4o and asked, “Can you ask me a question to determine which NHL teams to support? Make it a fun quiz.” ChatGpt has compiled a 10-question multi-choice quiz that aims to find out what kind of PlayStyle you want to see, what kind of fan you want to be, and how successful you want to be in the short term.
After finishing the quiz, ChatGpt decided, “You’re probably the perfect fit for the Toronto Maple Leafs!” For the team’s “rich history, huge fanbase, fast, exciting play style.” Not bad, but a random quiz that matches your personality to your team is probably the worst way to determine this.
That being said, it’s a starting point and I’d be lying if AI said I wasn’t happy linking me to a team that I really enjoy watching on live TV.
Deep research has failed me
Okay, so I try out the stupid quiz and have a starting point, but try to be real, you can’t decide which team to support based on the personality tests your AI has built. Next Step: AI Research Tools.
Research tools like ChatGpt Deep Research and Gemini Deep Research are currently trending topics in the AI world, and what is a better way to see what they can do than seek in-depth NHL research?
Ultimately, anyone who I choose to support is determined by who I can see, living in the UK makes midweek games almost impossible, and weekend games are incredibly sporadic.
I decided to truly understand my actual options. I had to regularly investigate how the team plays at reasonable times in the time zone.
To do this, I asked both the ChatGpt Deep Research and Gemini versions for the following prompts:
Both programs then asked follow-up questions about my preferences regarding research, such as whether they wanted to take into account regular season games and playoff games.
I give some additional information and then let the AI do that. Gemini Deep Research was much faster than ChatGpt, but in fact it took several minutes for Openai research analysts compared to over 10 minutes.
Gemini’s report provides a comprehensive percentage table per season, and on average, the Tampa Bay Lightning led a total of 28.35% of the games in UK normal times over the past five NHL seasons.
Meanwhile, ChatGpt discovered that the New York Rangers were teams with most matches at a reasonable time for the UK audience over the past five seasons, with 14% of matches being categorized into the time frame I was looking for.
Have you noticed the difference? Both deep search tools have had completely different results for me. Not ideal…
The results were so different, I decided to run through Gemini again to see what would happen. Fast forward about 5 minutes and it was said that 45.5% of the Montreal Canadiens games are in my desired time frame, but I know that’s not true.
So after using deep research for an hour, there was a dilemma. First of all, it takes too long to collect information and carry out the actual research necessary to fact-check my fun AI experiments. Secondly, the results of ChatGpt look more realistic than Gemini’s two attempts.
What was my decision? And what have I learned?
After testing AI research tools to help select and support NHL teams, there were several ideas and we came to a conclusion.
First of all, AI research tools are unreliable, but that doesn’t mean they can’t be used for fun projects like I tried today. I don’t have a really accurate answer as to which teams play the most matches between 4:30pm and 11pm UK time, but I have a rough idea of what the East Coast teams have played before.
If you plan to use deep research for academic research or bring together important information, seriously rethink it. Relying on AI to help you select NHL teams is one thing. Another thing to rely on AI for real work.
I enjoyed using the research tools in this article and even managed to feed one of the reports directly into Gemini’s audio overview to create realistic podcast clips that I think are great.
So who did I choose to support? Well, the AI tools I ultimately used didn’t help me much other than confirming that I already knew about the East Coast team I played previously.
I’ve noticed that since I started watching the NHL, I’ve been rooting for the Montreal Canadiens. As a French speaker, I have always loved the city. So, I’m not 100% sure yet, but now I look like a Hubs fan…
You might like it too
