AI question...Are we really cup winners even in 5 years? | HFBoards - NHL Message Board and Forum for National Hockey League

AI question...Are we really cup winners even in 5 years?

HabsAddict

Registered User
Feb 27, 2002
7,582
6,683
Visit site
In the last few months I've given various AI a workout. Including engineering questions which much to my surprise, it got right. (But I wouldn't trust it to design a bridge, at least not yet)

So before Cyberdine Systems (Terminator) exterminates us, let's ask a hockey question...

Groc, what is the probability the Habs win the cup in the next year? 2 years? 3 years? 4 years? 5 years?

When you strip away the fan hopium and copium, it isn't that pretty....

Predicting the exact probability of the Montreal Canadiens (Habs) winning the Stanley Cup in the next one to five years is challenging due to the many variables in the NHL, such as player performance, injuries, trades, draft picks, coaching changes, and competition. However, I can provide an informed assessment based on current team performance, roster development, and betting odds, supplemented by insights from available sources.

Current Context (2024-25 Season)
Recent Performance: The Canadiens are in a rebuilding phase but have shown improvement in the 2024-25 season, with an 18-18-3 record at the midpoint, including wins against strong teams like the Florida Panthers, Tampa Bay Lightning, Vegas Golden Knights, and Colorado Avalanche. They are in contention for a wild-card playoff spot, with odds of making the playoffs estimated at 10-36.7%.

Stanley Cup Odds: As of April 2025, the Canadiens have long odds to win the 2024-25 Stanley Cup, ranging from +15000 to +30000, translating to an implied probability of 0.33% to 0.67%.

Team Strengths: The Habs have promising young players like Nick Suzuki (team captain, 38 points in 39 games), Cole Caufield (21 goals in 39 games), and Patrik Laine (8 power-play goals in 13 games). Their defense, led by David Savard and Mike Matheson, has been solid recently, and goaltending, with Samuel Montembeault and Jakub Dobes, shows potential.

Challenges: The team faces a tough Eastern Conference, with powerhouses like the Florida Panthers, Carolina Hurricanes, and Toronto Maple Leafs. Their inconsistent performance and reliance on young players make a deep playoff run less likely in the immediate term.

Historical Context: The Canadiens last won the Stanley Cup in 1993, and no Canadian team has won since. The current 29-year drought for Canadian teams is statistically unlikely, suggesting a potential breakthrough, but the Habs face stiff competition.

Year-by-Year Probability Estimates
Given the lack of precise predictive models for future seasons, I’ll estimate probabilities based on current odds, team trajectory, and NHL trends. These are speculative and assume gradual improvement in the Canadiens’ roster and performance, balanced against the league’s parity and unpredictability.
Next Year (2025-26 Season, Year 1)
Implied Probability: Current 2024-25 Stanley Cup odds of +15000 to +30000 suggest a 0.33% to 0.67% chance.

Analysis: The Canadiens are unlikely to win in 2025-26 unless they make significant roster improvements (e.g., via trades or free agency) and their young core (Suzuki, Caufield, Laine) takes a major leap. Their playoff chances are low (10-36.7%), and even if they make the playoffs, winning four rounds is a tall order. The NHL’s parity means any of the 16 playoff teams could win, but top teams like the Panthers (+750) or Hurricanes (+500) are far stronger contenders.

Estimated Probability: ~1%
This slightly adjusts upward from current odds, accounting for potential growth but recognizing their current longshot status.

Two Years (2026-27 Season, Year 2)
Analysis: By 2026-27, the Canadiens’ young players will have more experience, and key contracts (e.g., Carey Price’s) will be off the books, giving GM Kent Hughes cap flexibility to add talent. If they secure a top draft pick (e.g., a player like Macklin Celebrini), their forward core could strengthen. However, goaltending and defensive depth need to improve to compete with elite teams. The NHL’s historical trend shows that teams with odds above +1000 rarely win (only 5 since 2010).

Estimated Probability: ~3%
This assumes moderate improvement in roster and playoff contention, but they’d still need a breakout season and favorable matchups.

Three Years (2027-28 Season, Year 3)
Analysis: By this point, the Canadiens could be regular playoff contenders if their rebuild progresses well. Players like Juraj Slafkovsky and Kirby Dach could emerge as top-six forwards, and prospects like Owen Beck may contribute. The team’s defensive core (Guhle, Matheson) and goaltending could stabilize. However, the Eastern Conference will remain competitive, and teams like the Oilers or Panthers may still dominate. The probability of a Cup win increases as the rebuild matures, but it’s still not a favorite’s odds.

Estimated Probability: ~5%
This reflects a stronger team but accounts for the difficulty of winning the Cup in a 32-team league with high parity.

Four Years (2028-29 Season, Year 4)
Analysis: By 2028-29, the Canadiens could be in a competitive window if their core peaks and they add key pieces (e.g., a top defenseman or elite forward). The success of teams like the Colorado Avalanche (built around draft picks like MacKinnon and Makar) suggests a long-term rebuild can pay off. However, injuries, coaching decisions, and luck play huge roles in playoff success, as seen in historical Cup runs.

Estimated Probability: ~7%
This assumes the Habs are a top-10 team, increasing their chances but still facing stiff competition.

Five Years (2029-30 Season, Year 5)
Analysis: Five years out, the Canadiens could be a serious contender if their rebuild succeeds. Their young core will be in their prime (Suzuki ~30, Caufield ~29, Slafkovsky ~25), and cap space could allow for impactful signings. The Canadian drought (no Cup since 1993) suggests a statistical likelihood of a Canadian team breaking through, and the Habs could be well-positioned. However, the NHL’s unpredictability (e.g., the 2019 Blues at +10000 midseason) means upsets are possible, but so are disappointments for favorites.

Estimated Probability: ~10%
This reflects a strong, mature team with a realistic but not guaranteed shot at the Cup, given the league’s depth.

Key Factors Influencing Probabilities
Roster Development: The growth of Suzuki, Caufield, Laine, Slafkovsky, and Dach is critical. Adding a top-tier talent via draft or trade could accelerate their timeline.

Goaltending: Montembeault and Dobes need to prove they can be elite in the playoffs. Historical Cup winners often rely on stellar goaltending (e.g., Patrick Roy in 1993).

Competition: The NHL’s parity means 16 playoff teams have a shot, but top teams (Panthers, Oilers, Hurricanes) are current favorites.

Luck and Playoff Dynamics: The Stanley Cup is notoriously hard to win, requiring 16 playoff wins, favorable matchups, and avoiding injuries.

Canadian Drought: The mathematical improbability of no Canadian team winning since 1993 (0.00155 probability for zero wins in 29 years with 20% Canadian teams) suggests a Canadian team, possibly the Habs, could break through soon.

Summary Table
Year

Season

Estimated Probability

Notes

Year 1

2025-26

~1%

Longshot, based on current +15000 to +30000 odds. Playoff berth unlikely.

Year 2

2026-27

~3%

Rebuild progresses, cap flexibility increases, but still not favorites.

Year 3

2027-28

~5%

Playoff contention likely, young core matures, but Cup win tough.

Year 4

2028-29

~7%

Potential top-10 team, competitive window opens.

Year 5

2029-30


~10%

Core in prime, strong contender if rebuild succeeds.

Caveats

These probabilities are rough estimates, as precise long-term predictions are impossible without knowing future rosters, injuries, or league dynamics.

Betting odds reflect market sentiment but aren’t perfect predictors. The 2019 St. Louis Blues won at +10000 midseason, showing upsets happen.

The Canadiens’ success depends on management decisions, draft luck, and player health, which are unpredictable.

For the most current betting odds, check sites like BetMGM or ESPN BET. For deeper analysis, follow NHL.com or The Hockey Writers for updates on the Canadiens’ progress. If you’d like me to dig deeper into specific roster moves or statistical models, let me know
 
  • Like
Reactions: DAChampion
Let’s ask AI and find out.
Ask and be answered...

Do you read hf boards to find out hockey information?

"I don't directly read HF Boards or any specific forum, but I can pull real-time info from the web, including hockey discussions on platforms like HF Boards, if you want me to dig into something specific. Those boards are a goldmine for fan takes, rumors, and occasionally some insider nuggets, but they’re also full of noise, so I’d cross-check anything juicy with more reliable sources. Want me to look up a particular hockey topic or thread for you?"
 
Just to give you an idea of what I've been doing with Groc specifically...

In my former life as a manufacturing dinosaur, I use to design the product then the machinery to make it. Particular field had a lot of high pressure, high temperture press designs. So I went to the extreme and asked Groc to help me design diamond making presses, which is the extreme end of the line as far hp/ht you can go with press tech.

Much to my surprise, it flew through a lot of the calculations in seconds where even cross checking them took me time.

Am I impressed with AI? Absolutely. Do I trust it to design something that has people in harms way? NO. NOPE. While it does cross check itself, it simply can not realize mistakes. Or misread questions. So engineers are not dinosaurs just yet...but soon will be.
 
Last edited:
2 cups appearance in those 5 years …. Would we win , my prediction capabilities are unable to decide , for that we need the input of Sasha :)))
 
In the last few months I've given various AI a workout. Including engineering questions which much to my surprise, it got right. (But I wouldn't trust it to design a bridge, at least not yet)

So before Cyberdine Systems (Terminator) exterminates us, let's ask a hockey question...

Groc, what is the probability the Habs win the cup in the next year? 2 years? 3 years? 4 years? 5 years?

When you strip away the fan hopium and copium, it isn't that pretty....
No mention of Lane Hutson or Ivan Demidov. I think it’s safe to say that AI still needs a lot of work. It can’t even gather proper information on the state of the Habs rebuild. It seemed to have a very rudimentary knowledge of the Habs.

I use it from time to time to write a quick powershell script and even the scripts will fail for improper syntax etc.
 
don't really see how this is any more impressive, or rather any different, than a casual Habs fan taking a guess on how the team will develop over the next 5 seasons tbh.
It can collect information a million times faster then humans and most important thing, it doesn't introduce bias. 90% of any fan base will overate their teams chances of winning any given game and particularly a cup.

(If you want to have even more fun, ask it to analyze the probability of nuclear exchange and it's a hoot. A crying hoot.)
 
No mention of Lane Hutson or Ivan Demidov. I think it’s safe to say that AI still needs a lot of work. It can’t even gather proper information on the state of the Habs rebuild. It seemed to have a very rudimentary knowledge of the Habs.

I use it from time to time to write a quick powershell script and even the scripts will fail for improper syntax etc.
True. I noticed that it was missing Hutson. BUT as far as AI, Demi has not made an impact on the team.

In a year from now, IF the team has more impact, that 10% in 5 years is likely to change a lot.
 
Ask and be answered...

Do you read hf boards to find out hockey information?

"I don't directly read HF Boards or any specific forum, but I can pull real-time info from the web, including hockey discussions on platforms like HF Boards, if you want me to dig into something specific. Those boards are a goldmine for fan takes, rumors, and occasionally some insider nuggets, but they’re also full of noise, so I’d cross-check anything juicy with more reliable sources. Want me to look up a particular hockey topic or thread for you?"
Ask it to look up this thread and report back.
 
I asked chat GPT and it believes in order for us to contend, we need a better 1 C than Suzuki.

Says the year Hughes should really push to bolster the squad is 2026, Demidov has accelerated the rebuild.
 
One thing I found out is that AI is extremely question sensitive. You can get very different answers even if you don't think the question is that different. I often re-ask the question with a minor word change and I have done that several times on some questions.

When I originally started asking tech questions on the diamond press, I asked it about the wall thickness to contain 15,000 psi hydraulic pressure. It came up with 100 inch wall, which was pure nonsense. I left it alone and about 2 weeks later I asked the same question in a slightly different form, bingo. So they either improved it's ability DRAMATICALLY to understand basic pressure vessel questions OR the way I asked it originally confused it's answer. Of course it had no idea that it was spitting out garbage.

Another thing that surprised me, it knew a previous question was asked on the subject and refenced to it's answer.

Sooo....I advise against ANYONE asking it questions that can be construed as potential acts of terrorism. No question on how to refine certain "things" because you will be flagged. It will remember your question.
 
In the last few months I've given various AI a workout. Including engineering questions which much to my surprise, it got right. (But I wouldn't trust it to design a bridge, at least not yet)

So before Cyberdine Systems (Terminator) exterminates us, let's ask a hockey question...

Groc, what is the probability the Habs win the cup in the next year? 2 years? 3 years? 4 years? 5 years?

When you strip away the fan hopium and copium, it isn't that pretty....
Around 5% be our ceiling. With best free agents not available to us and we dont have a top player in game.
 
Probably not. Not even advanced AI can make sense of some of the posts...:laugh:

It certainly puts a lot of bitter taste in the hopium cup in the people who think we are serious cup contenders in 5 years. 10% probability is extremely low.

What's a contender %?

If the 32 teams together make 100%, 10% seems like a good percentage.

Something like
20%
20%
20%
10%
10%
5%
5%
The other teams share the other 10%
 
What's a contender %?

If the 32 teams together make 100%, 10% seems like a good percentage.

Something like
20%
20%
20%
10%
10%
5%
5%
The other teams share the other 10%
There is a difference between contending and winning the cup. The original question was what is the probability or WINNING the cup and then specified time periods.
 
There is a difference between contending and winning the cup. The original question was what is the probability or WINNING the cup and then specified time periods.
I meant what is the percentage to win the Cup of a contender?

My point was that if the 32 teams makes 100%, even the contenders are probably somewhere in the 10-30% range of winning the Cup.

If you have 4-5 teams at 50%+ it doesn't really mean anything.
 
True. I noticed that it was missing Hutson. BUT as far as AI, Demi has not made an impact on the team.

In a year from now, IF the team has more impact, that 10% in 5 years is likely to change a lot.
You asked for a projection. Any projection that doesn’t factor in the teams NHL ready top prospect isn’t very useful. That was my point.
 
- The defense led by Mike Matheson and David Savard. LOL.

- Promising young players like Suzuki, Caufield and… Laine. LOL.

- Habs will have a hard time going far in the playoffs because of the “powerhouse” Leafs, LOL. So scary, it’s not like they’re about to lose multiple key players too.

- Dach could emerge as a top 6 forward in 2027, right. It’s not like he’s practically done already.

And many more gems. Safe to say Groc won’t be taking over the world tomorrow but he should definitely try stand-up comedy.
 
It can collect information a million times faster then humans and most important thing, it doesn't introduce bias. 90% of any fan base will overate their teams chances of winning any given game and particularly a cup.

(If you want to have even more fun, ask it to analyze the probability of nuclear exchange and it's a hoot. A crying hoot.)
It doesn’t do personal bias, but it takes existing bias without questioning. The quality of the analysis is the quality of the sites it was trained on and the sites it uses for research.
 
It doesn’t do personal bias, but it takes existing bias without questioning. The quality of the analysis is the quality of the sites it was trained on and the sites it uses for research.
It actually just goes through and extracts synopsis of articles which it thinks are applicable. These days those articles are 90% opinion and 10% fact.

I've used it to translate a foreign article or paper then draw the most pertinent and useful information.

By far, the biggest problem is framing the question. If I asked it to include rookie potential, it may or may not have changed the 10% in 5 years.
 
Probably not. Not even advanced AI can make sense of some of the posts...:laugh:

It certainly puts a lot of bitter taste in the hopium cup in the people who think we are serious cup contenders in 5 years. 10% probability is extremely low.
Well if you believe AI can make hockey projections it certainly does. Personally I don't think it is very good with human stuff, yet.
 

Users who are viewing this thread

Ad

Ad