Note: this has been updated on 6/22/2016, following the NHL’s release of 2017 Expansion Draft procedures.

The four most recent expansion teams, which entered the NHL in 1998, 1999, and 2000 respectively, had an enormously complicated task of assembling the best possible team out of an unprotected list of roughly 500 players.

The number of possible team combinations in any of these cases was so large that it almost defies logic.  The number of team combinations that Atlanta could have in 1999 was 16,448,918,502,604,299,999,498,731,520,000,000; this does not include trade options or other things that would have an impact as well.  When the unprotected list was pared down further by eliminating players who weren’t worth being considered (in my opinion), the number drops down to a mere 21,499,084,800,000.  Nashville, Columbus, and Minnesota all dealt with similarly large numbers.

Put another way, the odds of selecting the winning Powerball jackpot numbers are 1 in 292,201,338.  The number of team combinations that Atlanta could have put together would be like selecting the winning Powerball numbers 56,293,097,818,067,828,284,546,498 times.  Paring down the draft board to the players who should be considered, thus eliminating guys who aren’t worth considering, would be more like selecting the Powerball numbers 73,576 times.

However, things aren’t just a little different this time around; they’re a lot different.  The four most recent teams only had to select players from 26 teams; a new one will have to select players from 30 teams.  This starts taking the team combination numbers into astronomical numbers.  But there’s a bigger issue to consider.

In the 1998, 1999, and 2000 expansion drafts, eliminating large numbers of unprotected players was fairly easy.  Unsigned European players had their rights held by their NHL team indefinitely, so an expansion team could simply look at a 34-year-old Esa Keskinen, who’d never played a minute of pro hockey in North America, and remove him from the list.  Enforcers were still plentiful, although in a smaller number than they had been a decade prior, and weren’t really worth considering; this also eliminated huge numbers of possible selections.  And pending unrestricted free agents were another story; a Group III free agent was at least 31 years old and on the downswing of his career, and a Group V was younger, in his prime, and much more likely to test the market.  In the two CBAs since, Group V has been eliminated completely, and Group III players today can be unrestricted beginning around age 26.  In addition, the salary cap and salary floor have made the job infinitely more difficult to assemble a team out of nothing.

I said in a couple of other essays that I firmly believe that an expansion team needs to have someone like a co-GM simply for expansion draft duties; he (or she) would handle nothing but putting together the initial team, obviously while in close contact with the regular GM who would already have plenty to worry about with picking up unsigned free agents and preparing for the entry draft and the regular free agency period (starting July 1).  I said this while looking at the last four expansion teams; the job has gotten infinitely more difficult since then.

The next expansion team will be picking from 30 teams, with almost no enforcers, no unsigned European players, no Group V free agents, and few Group III free agents to strike from their draft board right off the bat.  It’s very possible that instead of being able to choose from only five or six realistic players per team in the next expansion draft, the new team will have to choose from ten, fifteen, or even twenty viable options.

How many possible combinations exist then?

The NHL’s official procedures for the 2017 Expansion Draft dictate that Las Vegas will have to select one player from each of 30 teams.  Of these, they will have to select three goalies, nine defensemen, and fourteen forwards, with four additional players from any positions.

For the sake of round math, let’s just say that each existing team will have exactly two goalies, three defensemen, and five forwards to choose from (which I believe understates the probable reality dramatically).  The equation for possible combinations would be 10x10x10x8x8x8x8x8x8x8x8x8x5x5x5x5x5x5x5x5x5x5x5x5x5x5x10x10x10x10.

This leaves us with 8,192,000,000,000,000,000,000,000 combinations, which would be like selecting the Powerball winning numbers 28,035,463,684,290,179 times.  And this is with a specific set of parameters that artificially restrict what the probable reality is.  Remember, the number of possible combinations with the Atlanta draft board I assembled was only 21,499,084,800,000, which would be like selecting the Powerball numbers 73,576 times.  73,576 in 1999, which I believe was a task too large for a GM to handle, compared to 28,035,463,684,290,179 in the next expansion draft.  That number, by the way, is 28 quadrillion, 35 trillion, 463 billion, 684 million, 290 thousands, 179.  And again, this is most likely understating the size of the draft boards by an astronomical amount.

And remember the other caveat, which is that these numbers don’t even include trade possibilities or other conditional moves that could be made.  The next expansion team will have to create scouting reports from scratch on every single player under contract to an existing team (roughly 1,400), every player whose rights are held by an existing team but not under contract (roughly 600 more), in addition to an enormous number of prospects and players who aren’t tied to an existing team in any way (which is in the tens of thousands).  And considering how much more high-level talent is out there than 16, 17, and 18 years ago, the complexity of the job becomes crushing.

My advice to a new team is very simple.  Hire your permanent GM, and hire a temporary co-GM on a one-year contract who only handles matters related to the expansion draft.  When that contract is up, he steps away into a reduced role or away from the team entirely; this eliminates the chance of a power struggle that could be devastating to an expansion team.  The co-GM should not be either a regular scout or scouting director, who will all have plenty enough to do as it is to create a database from scratch.  Nor should it be an assistant GM, who will have an enormous number of duties and headaches as it is as well.  Ideally the co-GM would be someone like the Roman general Cincinnatus, who was granted enormous (dictatorial) powers to deal with a short-term emergency in the form of a besieged army; upon taking charge, leading the army to victory, and eliminating the threat, he resigned his position as dictator and went back to his farm.  He went from farmer to dictator back to farmer in the span of fifteen days, his duty done.

An expansion team will have to find a modern-day Cincinnatus, whose job will be to put together the best first-year team in history and then go away completely.