10 Takeaways (Part 1): Fun and interesting Celtics stats as the season restarts

The NBA restart is only two weeks away. Just two more weeks before we have actual basketball to break down! During the hiatus, we’ve done some versions of 10 Takeaways that have often been only tangentially basketball related. Not this time.

We (OK...fine, it was me!) got a little excited with some of these stats. So, we’re breaking this version of 10 Takeaways into two parts. Here’s Part 1!

1. On the NBA Stats site (https://stats.nba.com/) you can get all sorts of data now. One of the stats that jumps out is the player tracking data that tells you how far and how fast each player runs each game. How do you use this? It can give you a sense of how a player is used on both ends of the floor. A traditional big man isn’t going to run as much, nor is he going to move as fast. A shooter who flies off screens is going to cover a lot of ground, and probably do it quickly.

For the Celtics, a few things stood out. On defense, Boston’s three wings (Jaylen Brown, Gordon Hayward and Jayson Tatum) all rank in the top-30 in distance covered per game. Brown leads the way at 1.19 miles run per game on defense (12^th in the NBA). Hayward is just behind at 1.17 miles and Tatum is 29^th in the NBA at 1.13 miles.

This data combined with what we see each game through the eye-test confirms Boston’s switch-heavy scheme. Rare is the possession that a Celtics defender stays on one player from start to finish.

None of that is probably surprising. What might be surprising to learn is that Gordon Hayward covers the most ground of any Celtic on offense. Hayward runs an average of 1.41 miles per game on offense (tied for 12^th in the NBA). What is definitely a surprise is that of all the players in the top-12, Hayward runs the fastest average speed at 4.79 miles per hour.

As you can see on this clip, Hayward tends to go hard when he gets the ball in his hands. This hard drive into the pull-up jumper in/around the paint has become Hayward’s signature shot in Boston:

2. We all know Jayson Tatum became an offensive machine this season, especially post-All-Star break. But when you look at his actual metrics, they become even more mind-blowing!

Tatum was roughly the same guy in terms of percentage and attempts on catch & shoot situations, as well as on post-ups. Where Tatum exploded in Year 3 was as a driver and as a pull-up shooter.

In Year 2, Tatum drove the ball on average 5.6 times per game. Those drives resulted in shot attempts 2.8 times per game. The scary part? Tatum turned the ball over 8.8% of the time when he drove. It’s pretty easy to remember how often Tatum would get stripped on a drive.

In Year 3, it looks a lot more like this:

Tatum is keeping the ball higher and tighter to his body. His drives are now up to 11 per games, with 6.5 resulting in shots. He still needs to finish better (Tatum still misses a lot of layups that maddeningly spin out on him), but he’s up to 47.3% when he goes to the rim. Most importantly? That turnover percentage is down to 5.1%.

The driving is nice, but the pull-up shooting from Tatum is crazy. In Year 2, he was already pretty good. Tatum took 5.4 pull-up jumpers per game and hit 35.7% of them. From behind the arc, he took 1.4 pull-ups a night and knocked down 32.4%. Those percentages are pretty good, as a pull-up is one of the more difficult looks in the game.

This season? Whew boy! Tatum took 8.3 pull-up jumpers a game and 4.5 of them from behind the arc. The crazy part? Tatum hit 39.3% overall on pull-ups and was even better from deep at 39.9%! For reference, of the high-volume pull-up shooters, Tatum’s percentage trails only that of pull-up maestro Damian Lillard.

Tatum can get his pull-up jumpers in two ways. He’s comfortable coming off a screen like this:

Tatum has also become a killer on the pull-up out of isolation like this:

That’s how you rapidly become one of the feared scorers in the league!

3. A popular narrative entering the season was that Boston’s defense would fall off without Al Horford. Many ran with Enes Kanter replacing Horford up front as being a disaster. But Brad Stevens went a different direction. He essentially replaced Horford with Daniel Theis.

Did the defense fall off? Not even a little. Boston ranks second in points allowed per game. That’s not a function of playing a middle-of-the-pack pace either, as the Celtics are fourth overall in defensive rating. Last season, with Horford, Boston was seventh in defensive rating, a full point behind this year’s team.

Some of that is definitely explained by the wings improving. Some is also explained by Kemba Walker giving a consistent effort, as opposed to last year’s come-and-go effort at the point guard position. But a chunk goes to Theis. He’s been just as good this year as Horford was last season.

The first worry is that opponents would attack Theis at the rim more often. That hasn’t happened. Theis defends 6.5 shots per game at the basket versus 6.4 shots per game for Horford last season. Theis has been slightly better defending at allowing opponents to shoot 55.4% at the rim versus Horford’s 56.5% allowed in 2018-19.

This play looks a lot like the ones Boston regularly got from Horford. Theis reads Victor Oladipo’s drive, know his guards have the perimeter covered and is able to stay with Oladipo on the crossover step for the block:

A big part of Horford’s value as a defender is tied to his ability to hold his own on switches on the perimeter. Theis is just as good though, and in some ways a touch better. Theis defended 4.1 shots per game that came from greater than 15 feet. In 2018-19, that mark was 4.7 attempts per game for Horford. Being able to take some shots away is a huge part of defending on the switch. And their percentage allowed was within one percentage point of each other.

Overall, switching from Horford to Theis hasn’t hurt Boston’s defense in any noticeable way.

4. Another popular story coming into the season was how Kemba Walker was more comfortable playing off the ball than Kyrie Irving. And how Boston would probably use Walker coming off screens more often. The numbers tell us that hasn’t really developed as expected.

Walker and Irving are similar shooters coming off screens, but the Celtics actually used Irving off screens more in 2018-19 than they have Walker this season. In addition, Irving took 3.9 field goals a game off catch and shoot attempts, while Walker has only taken 2.6 such attempts per contest. So, Walker off ball didn’t materialize as expected.

What Brad Stevens did do is trade some of Kyrie Irving’s ISO possessions for Walker running pick and roll. Walker ran pick and roll for Boston on 10.1 plays per game, or on 48.8% of the actions he was involved in. Those plays resulted in 1.08 points per possession, which is in the 94^th percentile of all pick and roll ballhandlers in the NBA.

Irving wasn’t a bad pick and roll player for the Celtics by any means, he just didn’t like to operate that way. Irving ran pick and roll on just 6.6 possessions or 29.4% of his actions. That resulted in 0.99 points per possession, which ranks in the 85^th percentile.

Trading Irving’s less efficient shots off screens and out of ISO for Walker running pick is part of what helped Boston jump from 12^th in offensive rating in 2018-19 to fifth this season.

The other big difference? Walker is a killer on pull-up jumpers. Walker took 6.2 three-point attempts per game on pull-ups (fifth-most in the NBA). He knocked down 36.4% of those attempts, which puts him up in the Damian Lillard, Jayson Tatum area as a pull-up shooter. Irving took very few pull-up three-pointers last season, at just 3.0 attempts per game.

By using Walker more in pick and roll situations, combined with his terrific pull-up shooting ability, the Celtics were able to find more efficient offense this season. Because Walker will pull up from three regularly, it helps widen and space the floor. This can be a really efficient weapon if Boston faces Milwaukee, and their drop defense coverage, in the playoffs.

5. Grant Williams is still a bit of a mystery box offensively, but there are signs he can be a plus player. Right now, Williams’ shot profile is that of a stretch big. 44% of Williams’ shots came from behind the arc. Considering he shot just under 25% from deep, that’s…not great!

But let’s look a little deeper. Williams, almost comically, couldn’t hit a three to save his life to start his career. He missed his first 25 attempts from long range. After that? Williams hit 21-of-60 three-pointers, which is a more than respectable 35%. Considering Williams has always been a good free throw shooter and was decent from three in college, we can believe in that 35% more than the 25 straight misses.

But what will set Williams apart is his ability to hurt teams when they put a smaller defender on him. This is either going to happen directly or when teams switch. It’s super small sample size theater here, but Williams was wildly effective on post-ups this year. On 18 total post-up possessions. Williams shot the ball 17 times and scored on 10 of those, for 58.8% shooting. That puts Williams in the 99^th percentile on post-ups.

If Williams can make a couple of plays like this when he’s got defenders he can bully inside, it will go a long way towards complementing what should be a solid stretch game.

Part 2 to come on Friday!