Small Sample Sizes Make the Best NBA Stats (2020-21 Edition!)

Last year, early in the season when most teams had only played two or three games, I took a look at the insane NBA records that would not only fall but be shattered if the players who had wild openings to the season kept up that pace for 82 games.

After Trae Young went for 37 points on just 12 shots and for one game had an absolutely bonkers PER of 59, it was clear that a sequel was in order.

So welcome to the Second Annual Small Sample Sizes Make the Best NBA Statistics column! Rather than get caught up in the pleasantries of an overlong intro, let’s just get right to the fun part, and to keep the stats in their proper historical context, let’s pretend we’ve got an 82-game season ahead of us rather than 72.

Last year, Young averaged 38.5 points per game in his first two contests; this year, he’s averaging 36.5, with just 73 points instead of 77.

That may not seem like much, but 36.5 points per game over an 82-game season works out to just 2,993 points, which means unlike last year, nobody in the NBA is on pace through two games to become the only player in NBA history not named Wilt Chamberlain or Michael Jordan to score 3,000 points in an 82-game season. (OK, Wilt was even more impressive; he topped the 3,000 point mark three times in seasons that were 79, 80, and 80 games respectively between 1961 and 1963, even going over 4,000 points in 1962. It’s a challenge to do that even in a video game today.)

Speaking of Trae, last year, Giannis Antetokounmpo broke Wilt’s record that had stood since 1963 for the highest PER for a single season at 31.86 (Wilt’s record was 31.82, showing just how hard it is to shatter a record like that.)

Trae is one of seven players who have a PER over 31.86 so far this season, clocking in at 42.67.

The other six? In order, Kyrie Irving, Khris Middleton, Ja Morant, Chris Boucher, DeMar DeRozan, and Rudy Gobert. Small sample sizes are weird.

Last year through two games, the most 3-pointers made without a miss was three. This year, the same is true; Deni Avidja, Raul Neto, and D.J. Augustin are all 3-for-3 from long range; nobody who has attempted four or more three-pointers has made all of them.

Kyrie Irving’s got a PER over 40, but that’s not his most bonkers stat so far.

Through two games, Irving has .589 Win Shares per 48 minutes. His teammate, Kevin Durant, has .419.

Between them, that’s 1.018 WS/48, and for a stat where there is almost a 1:1 correlation between Win Shares and actual team wins, that means that if they played all 48 minutes and the other three guys on the floor played like absolute garbage and put up no WS at all, the Nets would still be statistically expected to go undefeated.

Indeed, a team full of nothing but Kyrie Irvings in terms of WS/48 would be expected to win 241 games in an 82-game season.

Irving and Durant, since they have not in fact played all 96 minutes in their two games, have combined for just 1.2 WS, but that’s 60 percent of the wins for an undefeated team through the first week of the season. Irving and Young both have 0.7 WS so far; in an 82-game season, that would equate to 28.7 WS total, which would smash the record of 25.37 WS set by Kareem Abdul-Jabbar in 1971-72.

And oh by the way, Kareem played 44.2 minutes per game when he put up that total. Kyrie and Trae are averaging 29.5 and 30.5 minutes per game respectively.

Russell Westbrook is off to a weird statistical start in Washington. He’s currently on minus-0.1 WS through two games thanks mainly to his horrendous shooting (15-of-41 from the field including missing all six of his threes.)

And yet Westbrook is averaging 18.0 points, 13.0 rebounds, and 13.5 assists per game. With negative Win Shares.

Repeat after me: Counting stats start arguments. Advanced stats finish them.

Westbrook does, however, have a 55.4 assist percentage so far this season; this is the second-highest mark of his career (57.3 during that insane 2016-17 campaign) and would rank fourth all-time if it held for the entire season.

Westbrook’s 2017 campaign ranks third; of the top ten full seasons in this category, Westbrook and Chris Paul have one each, Steve Nash has two, and the other six, including the first, second, fourth, sixth, seventh, and tenth spot, all belong to John Stockton.

Oh, and just for fun, Trae Young also shows up on the bonkers assist percentage list as well; he’s at 50.84.

And one more wacky assist percentage stat while we’re here. Nikola Jokic has an AST% of 49.45 so far; that’s better than Magic Johnson‘s best season (49.30 in 1990-91.)

Myles Turner has been sending back shots like an angry customer at a restaurant sending back a blue rare steak for being overcooked. In NBA history, only three players have ever recorded a 10.00 block percentage or better for a full season; their names are Manute Bol, Manute Bol, and Manute Bol.

Turner’s 19.57 is nearly double the record of 10.81 set by Bol in 1988-89.

And yet it’s only 0.25 points better than…Chris Boucher.

How on earth did Boucher come out of nowhere for the Raptors on this young season? In 43 minutes off the bench over two games, he’s got 34 points and nine blocks with a TS% of .777.

Coming back to Young’s bonkers start, he leads the league in free throw makes (27) and attempts (31) and is on pace to shoot 1107-of-1271 over 82 games.

That would annihilate the record of 840 free throw makes set by Jerry West in 1966 and Young could do it in 72 games…if, y’know, he keeps getting sent to the line 15.5 times a game, which…well, don’t bet your life on it.

Meanwhile, in team stats, the Nets, led by Irving and Durant, have won their two games by a combined score of 248 to 194.

Surprisingly, that’s not even the best offense in the league per 100 possessions; the Nets’ 120.2 Offensive Rating is third behind Milwaukee (124.2) and the Lakers (122.0).

Ain’t it funny how the Nets and Bucks are in the top three after having had a chance to beat the snot out of the Warriors.

But as putrid as Golden State’s defense has been (a 120.4 Defensive Rating), it’s not as bad as Dallas (121.6) or Boston (124.2.)

That 124.2 Offensive Rating would be the best in NBA history by a mile for Milwaukee. Likewise, the Celtics are putting up the worst defensive season in the league’s annals.

Meanwhile, the Nets’ 94.0 Defensive Rating is compared against a league-wide 109 points per 100 possessions; that 86.24 Defensive Percentage would be, by a factor of over five points, the best defensive season in the history of the league, dwarfing the absurdly stingy defenses of the Pat Riley Knicks and the early-aughts peak of the San Antonio Spurs.

Eleven teams are still undefeated; another 11 are winless. Just eight teams have played two games so far and split them.

Of course, two games is one thing. Matching the 2015-16 Warriors’ 24-0 start is quite another.

But at the same time, the Nets’ 27-point average differential sets them up for 114 wins in an 82-game season according to the simplest “one point equals 2.7 wins over or under 41 for an even split” formula for correlating unadjusted point differential with total wins.

Oh, and one other thing. There’s one record out there that no matter how small the sample size, we’ll never see broken.

Wilt Chamberlain averaged 27.2 rebounds per game in 1960-61. Only three players—Rudy Gobert, Andre Drummond, and Nikola Vucevic—have more than 27.2 rebounds…in two games. Gobert leads the league with 17 per game. That record is not going anywhere.

Only two questions remain from this tiny sample.

One, just how good will these players who got off to hot starts do over the course of the season…

…and two, how many points will Trae Young score in his first two games next year? Because he sure seems to have a knack for coming out of the gate like a bat out of hell and scorching the nets to the sound of a classic Tim Kitzrow video game call.

He’s heating up. He’s on fire! The rest of the season is going to be fun…and a little disappointing when the cops show up to enforce the law of averages.