Why 84-0-0 Is So Rare: A-Squad’s Perfect Season

Start with what actual hockey has produced. The 1929-30 Boston Bruins went 38-5-1 in a 44-game season, a points percentage of .875 that nobody has matched in the near-century since. The 1976-77 Montreal Canadiens went 60-8-12 over 80 games, lost eight times all year and only once at the Forum, and remain the usual answer when people argue about the greatest team ever iced. The 2022-23 Boston Bruins went 65-12-5 and set the records for wins and points in a season, with 65 and 135.

Notice what even those teams did not do. The best of them lost roughly one game in nine. Hockey is a low-scoring sport played on ice, with a goaltender who can steal any single night, and a bounce off a skate or a post decides more games than in almost any other league. An undefeated season has never happened and nobody serious expects one. That is the sport A-Squad is modelled on, and the record you post is meant to feel like a season from that sport, scaled to 84 games.

We could have capped the projection so that no squad ever reached 84 wins. We chose not to, because the chase is part of the game. A ceiling nobody can touch is a rumour; a ceiling somebody reaches now and then is a story. So the record is projected by a statistical model built only from public box-score statistics, and if the six cards you drafted are strong enough, the model will say 84-0-0 and the grade will read PERFECT.

Strong enough is a very high bar. When we simulate well-drafted squads, they usually land in the high 50s in wins, which is a comfortable playoff team with a real shot at the Cup. A perfect season turns up in well under one in a hundred of those well-drafted runs. For random picks it effectively never happens. If you draft carefully every day you should expect to see one eventually, and you should expect it to take a while.

We also checked the model against more than 1,750 real NHL team seasons, so the records it produces sit in the same world as the ones in the record book. A squad that projects to 84-0-0 is not a glitch. It is six cards that, on paper, would have dominated any league they were dropped into.

Six cards carry the whole team. One player fills each slot, and the rest of the roster is treated as league-average, so a perfect season means your six added enough to an average team to win every night. Nothing about that is forgiving. A single weak slot is a third of your forwards or half of your defence, and the model notices.

Re-rolls help, and knowing when to spend them is most of the skill in the game. But a perfect season is usually a strong draft meeting a generous set of spins. You cannot force it. You can only be ready for it.

The daily has one more filter. Before a day’s spins go live, a solver plays them out and reports the best record it can find. If the obvious picks add up to 84-0-0, the seed is discarded and another is minted for the same day, until the solver comes up short. We added this after one early daily offered a perfect season on the plain reading of the board, several times over, and a board where everyone at the top is 84-0-0 is a tie, not a competition.

The solver is a strong drafter, not a flawless one. It can miss a line that a patient human finds, so a perfect daily is still possible. It just cannot be the default answer. On most days the best record available is short of perfect, and the board is about who gets closest.

Below PERFECT sit HISTORIC and DYNASTY, and we want to be clear that both are real achievements rather than participation prizes. A HISTORIC season is one of the best records in the history of the league, the kind that gets its own documentary. A DYNASTY season is a team that would be favoured in every series it played. In our simulations of well-drafted squads, DYNASTY is uncommon and HISTORIC is rare; most careful drafts land at CONTENDER or PLAYOFF, which is an honest result for six cards on an average roster.

If you post a HISTORIC season, you drafted a squad that would sit beside the 1977 Canadiens and the 2023 Bruins. That is not a consolation. Treat 84-0-0 as the thing you chase and HISTORIC as the thing you can be proud of on an ordinary Tuesday.

People ask which teams and eras produce perfect runs. We do not publish a list, partly because the list would become the strategy and partly because it shifts whenever the data pack is updated. We will say what any hockey fan would guess: Orr’s Bruins and the 1970s Canadiens turn up a lot, and that will surprise nobody who has opened a record book. The common thread is a handful of players whose seasons were so far ahead of their league that the adjustment for scoring rates still leaves them towering.

That is the whole secret, and it is not much of one. Draft the players who were the best in the world in their year, at the position they actually played, and put a great goalie behind them. Do that six times on one set of spins, and the model will do the rest.