The Golden State Warriors won 73 of their 82 regular season games this year, and it was the most absurd thing I've ever watched. They outplayed other teams, stayed healthy, and got lucky in spots, but there was a heavy, heavy emphasis on the part about outplaying teams. If they were to play a 162-game season with that same winning percentage, they would go 144-18. A baseball team with that record probably wouldn't need the wild card.
It makes sense today, then, to look at what it would take for a modern baseball team to set the record for most wins in a season. As usual, there are disclaimers.
Disclaimer #1: Basketball is actually a different sport.
Pretend there are pictures here with me, a measuring tape, a basketball, and a baseball, labeled "Exhibit A" and "Exhibit B." I promise, it's true. There aren't even any bats in basketball.
More specifically, basketball has fewer players on an active roster, which makes it easier to accumulate a plurality of great players. Building a starting lineup in basketball is more akin to filling out a pitching staff (five starters + seven reserves of diminishing importance) than a full 25-man roster. The 2015 Indians would have dominated, but for that pesky hitting.
Also, Steph Curry taking 30 shots in a game is (very) roughly analogous to Mike Trout getting an extra dozen at-bats or so, and Trout would take those at-bats directly from Johnny Giavotella. That helps eliminate some of the sample-size nonsense that can happen in any given game. It's why people freaked out when the Lakers beat the Warriors this year, but no one will blink when the Phillies beat the Dodgers in a single game.
Disclaimer #2: There's no golden "best record" to chase.
The best record of all time? Why, that's the 1869 Red Stockings, who were 67-0. Sure, they played with a ball made out of armadillo shells and nighttime hadn't been invented yet, but that shouldn't make much of a difference, right?
Okay, if the 1869 Red Stockings don't count, the record probably belongs to the 1906 Cubs, who were 116-36. They played just 152 games, though, so maybe it's just the 2001 Mariners, who were 116-46. Except, why should the Mariners get credit for playing an extra 10 games and essentially losing all of them? They should get credit for playing an integrated game, though. It's all a mess.
I'm not sure if you can sell the public on a chase for winning percentage (which would require a modern team to win 124 games), so just go for straight wins, then. It'll satisfy almost everyone.
What would it take for a baseball team to win 117 games?
Be one win better than the 2001 Mariners
Oh, don't be a smartass.
This is a fair point, and I can assure you the Warriors had their moments of good fortune. The Warriors' record needed James Harden to dribble off his own leg in the closing seconds of one game, Will Barton to miss two shots in the final seven seconds in another, and Curry to hit a 40-footer in the closing seconds of a completely daffy game.
All of those things are reasonable, explainable occurrences, but for all of those events (and dozens of others) to happen in the same season? It's a little like a good run of cards at the blackjack table for the best card-counter in the world. Doesn't need the help, but the help doesn't hurt.
The Warriors were 7-2 in games decided by three points or fewer. The '01 Mariners outplayed their Pythagorean record by seven games. It takes a little push of luck to get from a stunning, unbelievable record to a record that's historically significant, and there's no shame in it.
Have better players than the other team
Come on, I already asked you not to be a ...
No, way better players
Okay, fine. At least this allows me to remind you that the '01 Mariners won 116 games, even though they had jettisoned Randy Johnson, Alex Rodriguez, and Ken Griffey, Jr. within the previous three seasons. The return for Johnson and Griffey had a part in 116, sure, but it's still impressive to let three of the inneriest of inner-circle Hall of Famers* go and come out better on the other side.
*based on performance only, you lawyers
Let's loosely define "way better players." The 1906 Cubs won 116 games with a lineup that combined for only a 103 OPS+. That's good, but rather ordinary. Their real trick was a pitching staff that combined for a 151 ERA+, which is like David Price and Sonny Gray combining to throw every inning in a team's season.
Take the 2015 Cardinals. They had five excellent starting pitchers, and they combined for 20 wins above replacement, which was the biggest reason the team won 100 games. But if they had the top five pitchers in baseball last season -- Zack Greinke, Jake Arrieta, Clayton Kershaw, Dallas Keuchel, and Max Scherzer -- they would have combined for 40 WAR. Take those wins, slap them onto the '16 Cardinals, and bingo bongo, you have a new record.
So all a team would need to do is accumulate the five best pitchers and have them all stay healthy and effective. Sabermetrics is such a joke, anyone can do this stuff.
But, yeah, you're seeing the problem with chasing 116 wins. It takes scenarios that are incredibly, impossibly unlikely with 30 teams chasing the same players to fill out a 25-man roster. Even if Matt Harvey, Noah Syndergaard, and Jacob deGrom all pitch as well as they're capable of in a best-case scenario, there are still two spots to fill. That might not mean much on a march to a division title, but it means there's way too much variance for a march to an all-time record.
Even if you don't just limit your search to a pitching staff, you still need all-time, historically significant performances. Bret Boone had one of the greatest performances by a second baseman ever, an outlier of a season from a 32-year-old that PECOTA wouldn't have predicted after a keg of fermented code. Ichiro was in his prime, and was also something of an unknown since he was a rookie in MLB. The defense wasn't just good, it was otherworldly.
They were better than everyone that year, and it wasn't just because they had players who were necessarily going to be better than everyone in the following year.
The Mariners got 700 excellent combined at-bats from 36-year-old and 37-year-old utility players. They had relievers behind other relievers, and they had a hotshot young starter to fill in when their fifth starter pootered out.
Have excellent health
The Mariners used just eight starting pitchers that year, and two of those made just a couple starts. They had three 200-inning starters, and seven of their starting position players appeared in 130 games or more, with the catcher making 123 starts.
Do all of that, but more of it
To recap, in order to break the MLB wins total of 116, a team would have to accumulate a collection of a) the best players in baseball having b) the best seasons of their careers while c) getting more than a little lucky and d) surrounding their best players with remarkable depth and e) avoiding the injury krakens that suck the other teams down. They would have to do that, all of that, better than any team before them has done it.
Which is all to say it might not happen in your lifetime. It's why there was nearly a century between 116-win teams. This is basically a love letter to what the 2001 Mariners accomplished, and a warning that you shouldn't expect it again.
It isn't impossible. When Sam Miller examined the fringes of PECOTA's million simulated seasons, he found the Cubs and Dodgers winning 121 games in an alternate universe, with the Mets winning 123. Note that none of them eclipsed the winning percentage of the 1906 Cubs in a million freaking simulations, which means that record might not be reachable, but the raw wins total isn't totally out of reach.
It would take just about the freakiest season we've ever seen. Basketball lends itself to regular season domination and expected outcomes in a way that baseball can't, so we have to rely on freaky outliers for anything in baseball that's similar to what the Warriors just accomplished.
In a way, that's a shame. In another way, that's kind of why baseball fascinates a lot of us in the first place.