Salt City Seven: Bogey, Close Wins & How Much to Trust Projections

Every week here at SCH begins with the Salt City Seven: seven regular features that let us relive the biggest moments, key performances and hot issues in Jazzland from various angles. Check in every week for the quotes, stats, plays and performances that tell the stories from the last 168 hours in the world of the Jazz.

“It’s great to have Bojan, I’m not gonna lie to you. He’s made my life ten times easier. He’s made everyone’s life ten times easier.”

-Jazz star Donovan Mitchell, via KSL.com, on Bojan Bogdanovic

The last time the Jazz finished the season with two scorers touting averages north of 20 points per game was the 1991-92 season, when Karl Malone posted 28.0 and Jeff Malone added 20.2. Current Jazz star Rudy Gobert would be born a few months later, and Mitchell wasn’t yet a twinkle in his parents’ eyes.

It’s really, really early, but the Jazz are currently on pace to enjoy that same luxury for the first time in 27 years. For that matter, only twice since the Malones played together have the Jazz gotten this deep into a season — nine games in — with multiples 20-point scorers. Deron Williams and Paul Millsap both averaged 20 over the team’s first nine games in 2010-11, and Gordon Hayward and George Hill did the same in 2016-17, although part of that is due to Hayward only appearing in three of the team’s first nine games that year due to injury.

So what the Jazz have right now with Mitchell dropping 24.8 per outing and Bogdanovic adding 21.8 is pretty unique.

“Bogey” has been turning heads in the Beehive State and beyond, even before scoring the eventual game-winner versus Philadelphia on Wednesday and then adding a buzzer-beating three on Friday night to beat the Bucks. The Croatian forward has had momentum-shifting runs and big shots in each of Utah’s six wins so far. He has carried the torch for the Jazz for entire quarters, including picking up for Mitchell or Mike Conley when they have struggled.

He has also been playing great individual and team defense. It shouldn’t come as a shock that Bogey is a plus defender¹, but it’s certainly not at the top of his résumé. And at his size and position, he often has to guard very different players from game to game or even within the same game. Per matchup tracking data on NBA.com, Bogdanovic was primarily responsible for checking 6-foot-10 All-Star Al Horford on Wednesday night and held him to 1-of-6 shooting. In the very next game, he had a very different assignment with All-Star wing Khris Middleton visiting, and held him to 3-for-8.

He’s defending. He’s knocking down open shots. He’s attacking the rim. He’s fist-pumping his way into the hearts of Jazz fans.

But to say that he makes his teammates’ job 10 times easier only tells part of the story; they’re making his job easier, too.

Brooklyn fans got to see some of the raw scoring ability that Bogey brought from Europe, and then as a Pacer he got to further develop as a creator. But what the basketball world is seeing from Bogdanovic right now is without a doubt his best basketball to date. He has never scored this efficiently, partially because he has cut some mid-range stuff out of his diet, but mostly because he’s just in an incredible rhythm.

He may not stay over 20 all season, both as his shooting regresses toward the mean and as other factors around him change. Eventually, Conley will get going and the Jazz will want more shots coming off his hands. And Rudy Gobert has rightly requested to be more involved after literally leading the NBA in per-shot efficiency last season. So eventually the offense may get a little more spread out — in a healthy way — even if that means a little bit less scoring from Bogey.

But at least for now, the Jazz have a second weapon of historically unique proportions, and a big-moment scorer who is flat out winning them ball games.

3-1

With three of their last four wins coming by a one-possession margin, the Jazz now have more wins by three points or less than they did all last season. Not that last year’s mark is a hard bar to clear: they were 0-7 in such games last season.

+27.9

File this in the “a bit too early to be conclusive” category, but this is still interesting to monitor. The five-man lineup of Conley, Mitchell, Bogey, Gobert and Joe Ingles has the second-best Net Rating of any quintet in the league with at least 35 minutes together. All five of those guys also have positive Net Ratings in the clutch — led by Ingles at +36.4.

22%

One thing I still wish the Jazz would address offensively: their propensity to be “Team Floater.” A whopping 22% of their non-garbage time shots come in the category CleaningtheGlass.com calls “short mid,” from 4 to 14 feet from the basket. That’s the fourth highest chunk of those shots in the league, even thought they convert the fourth lowest percentage of them.

97.8

Utah will head into this week’s games in the top spot for overall defensive efficiency. They also have the best eFG% defense and hold opponents to the second fewest corner threes and fourth fewTest rim attempts, also per CTG. They own the league’s second-best halfcourt defense (Lakers are first), and the best transition D after a live miss. Their only weak spot right now is defending after a live turnover.

Mountain Memories as Jazz Run Familiar Flex Pindown

This season, the Jazz are paying tribute to their ’90s brethren by wearing the purple mountain jerseys of that era’s franchise legends in certain games. But when I saw this play action against Philadelphia, it felt like an even more fitting homage to those Jazz teams, because this is something they ran All. The. Time.

The screen Ingles sets across the lane is called a flex screen, and it’s a pick that John Stockton set multiple times a night. The basic idea behind a flex screen is to try to score an open layup for (in this case) Jeff Green, but most teams are too smart to allow that. So what the ’90s Jazz ran instead — and what their contemporary counterparts run here — is a stacked action where a guard sets a flex screen, then immediately takes a pindown screen to pop out to the perimeter.

Stockton ran it all the time, either for a shot or to flow right into pick-and-roll with the big. Jeff Hornacek ran it. A little later on, Matt Harpring would run it but then curl around the pindown screen for a free-throw line jumper. This play is basically a little slice of Jazz history.

The idea behind this little flex action, screen-the-screener play is that it makes the defender have to respect the initial action (or risk giving up the basket to Jeff Green), which should make him late to fight over the pindown. Josh Richardson does a good job protecting the basket without losing Ingles, but then Joe saves the play by making a snap decision based on a smart read. He sees that Richardson is trying to fight his way under the pindown, so he changes directions, and instead of coming to the top of the key, he veers to angle left. Richardson has no chance now, even tripping over his own teammate as he tries to react to Joe’s smart adjustment. 

Close wins and wins over elite teams almost always mean multiple game ball candidates — it simply takes multiple great performances to beat the best teams. But here’s where we landed after two straight photo finishes against the Eastern Conference’s best squads.

Jazz 106, Sixers 104: Rudy Gobert Rudy was the majority pick and also the right pick. His 14-and-16 line was awesome, but as is often the case, he dominated this game defensively. Sixers players shot 6-for-21 (28.6%) when guarded by Gobert, which is even more impressive when you figure that he was primarily responsible for one of Philly’s two All-Star bigs every minute he was on the court. But that doesn’t mean it was an easy call. Bogdanovic hit the Jazz’s two final field goals, had a game-best +21 and was also superb defensively. Mitchell got the Jazz going with a personal 8-0 run when they were struggling, then finished with 24 points and eight assists. That’s not all: Joe Ingles came alive and scored 16, Mike Conley had a superb floor game, and Royce O’Neale guarded extremely well.

Jazz 103, Bucks 100: Bojan Bogdanovic How can you NOT give it to Bogey after that? It wasn’t just the game-winner, although shots like that are extremely rare. Bogey’s shot was just the fourth 3-pointer this season made in the final two seconds of an NBA game to take the lead, and only a dozen such shots were made last season. It wasn’t the 33 points, either. It was the fact that he bailed the Jazz out with big buckets or hot stretches every time Milwaukee got close. When the Bucks cut Utah’s 7-0 lead down to one, Bogdanovic responded with eight of the Jazz’s next 14 points. Then when the 21-point lead got trimmed all the way to seven, he scored the next 13 Utah points. O’Neale did solid work on Giannis Antetokounmpo, and Gobert was everywhere on defense — challenging shots out in space and locking down the paint. His contest against Middleton on the Bucks’ final play forced the turnover that set up Bogey’s heroics, and at one point in the second quarter, he and O’Neale teamed up to force four straight misses by Giannis in the paint. Giannis, umm, doesn’t miss in the paint.

Rather than regurgitate what the predictive models say at this way-too-early juncture in the season, let’s pause and use this space this week to talk about what those models do, and how much stock a fan base should put in them.

The projection systems referenced most often in this space are FiveThirtyEight’s model, the SRS-driven version at B-Ref, and ESPN’s playoff odds fueled by Basketball Power Index. (Note, those last two will debut later in the season, but we can already figure out who those models are going to like, because BPI and SRS are already live. SRS is always tracked in the standing’s on B-Ref’s season summary page, and you can find BPI here.)

The model at 538 is a bit more stable right now, because instead of using game outcomes — which is a limited set of data points with most teams having played 7-9 games — it relies more on a macro measure of the quality of a team’s talent. And they have built their model to trust a player’s preseason projection more than the current season at first, and then gradually put more weight on the current season as the sample gets bigger.

I won’t get into the minutia of exactly how they weight it, but let’s use James Harden for example. Because the Beard has averaged nearly 2,800 minutes in the past three seasons, by the time he plays roughly 1,800 minutes this season, their model will start to trust his 2019-20 offensive performance more than it trusts his projection going in. For someone like Dante Exum who has played far less, that will happen around the 700-minute mark. The reasoning behind this is simple: Harden has pretty well established who he is, so if suddenly he’s performing at a very different level, there’s a strong likelihood that something has changed contextually in the short term.

But there’s a downside to this stability, and it’s that 538 tends to trust a team’s pedigree, for better or worse, a little longer than an outcome-only model would. As such, it might be slow to recognize a player’s or team’s improvement as legitimate.

Personally, though, I find value in the more stable models, especially early on. It’s easy to get duped by a team’s hot or cold start, which is why I tend to viscerally trust what’s in a team’s DNA. And because 538’s model does adjust for who’s playing and who’s injured or resting, if someone is performing way above or below expecations for flukey personnel reasons, the model can help suss that out, too. For example, the Clippers’ RAPTOR rating jumped about 100 points from Wednesday night to Thursday night. Why? Because Kawhi Leonard rested on Wednesday but played the next night.

For those reasons, 538 is my jam early in the season. Eventually the model will start to put more stock in new data points, but it will bring them into the mix gradually so that people who trust their model don’t get too carried away with a hot stretch.

SRS, on the other hand, is all about outcomes. It attempts to measure a team’s quality by looking at their point differential — a fairly reliable indicator of future success — but with adjustments to account for schedule strength. It doesn’t account for personnel in any specific way, and it doesn’t care at all what we thought about a team before the season started.

This might make it sound a tad more objective at measuring the present performance, but sometimes that objectivity is a vice. Once SRS starts fresh at the start of a new season, it doesn’t care that Harden has played at an MVP level for the past three seasons. To be honest, it doesn’t even really care that Harden exists, except as a part of a composite that won this game by X and lost that game by Y.

That’s why SRS gets a little jumpy in response to small samples, and why I don’t watch it as much this early in the season. Right now, SRS thinks that Phoenix is the third best team in the Western Conference, Houston the tenth. Both of those statements are probably going to prove very false. While 538 understands who Harden is and assumes that eventually a team with an MVP candidate will start to perform better, SRS just cares what the scoreboard said the last eight times the Rockets left the court. After teams have played 15-20 games, SRS will be a little more steady and can start to be a good indicator of team quality. In the meantime, it’s interesting from a directional standpoint.

BPI’s starting point is similar to SRS: it uses the score of each game and adjusts for schedule strength and game location. It also makes further adjustments based on miles traveled and rest. BPI’s creator claims that it factors in “preseason expectations” — defined as a blend of the team’s Vegas over/under and the previous season’s performance — but is pretty opaque about describing how exactly that enters into the rating Still, you can tell just by looking at the rankings how different the SRS and BPI systems are; while SRS believes in the hot starts by the Suns, Mavs and Thunder, BPI’s top five Western Conference teams are exactly the five who drew the most fanfare going in. So the really quick version of all three:

• 538: Player-driven, with a lot of weight on past performance, making it more stable but also slower to recognize a real change in performance.
• SRS: A just-the-facts-ma’am ranking based on game outcomes, that might actually be too objective to be terribly useful when the sample size is so small.
• BPI: Similar to SRS in that it’s drawn from game outcomes, but with some mysterious adjustment made to account for how good we thought the team was coming in.

Now the good news for Jazz fans: Utah ranks extremely well in all three models. FiveThirtyEight gives them the highest current rating and forecasts a second-place finish, while they rank second among Western Conference teams in SRS and tops in BPI.

No matter where you look for validation, the Jazz are elite, folks.

After just two games last week, Utah has another weirdly light week with just three games, and both Saturday and Sunday off for a second straight weekend.

Monday 11/11, Jazz at Warriors: Five months ago today, the Warriors woke up two wins away from another NBA championship. Since then, just about nothing has gone their way. Kevin Durant and Klay Thompson got hurt. Kawhi Leonard beat them on their court. Durant left. An abysmal start. Then Steph Curry and Draymond Green also suffered injuries. Now, the once-mighty Warriors are 2-8 and coming off three straight losses as they prepare to face the Jazz. The bright side: expect to see more of former Jazz guard Alec Burks than you otherwise would have.

Tuesday 11/12, Jazz vs. Nets: A home game against this version of the Nets — Durant is out for the year — would normally be pretty manageable, but the schedulemakers made it a tough one. Utah will be on a back-to-back, coming from West to East after a late tip-off. The Nets meanwhile, will be rested; they’re already in Salt Lake after losing at Phoenix on Sunday.

Friday 11/15, Jazz at Grizzlies: This will be Conley’s first trip back to FedEx Forum since the Grizzlies sent him to Utah. That also means it will be a reunion with former Jazz guys Jae Crowder and Grayson Allen. Crowder leads the rebuilding Grizz in minutes played and has already endeared himself to the locals with a game-winning three. But overall, the Grizz are expected to struggle, and have started the season 2-7.

To close this week, let’s return to the Bucks game, just seconds before Bogey’s big shot.

Bucks fans were upset that Middleton committed a turnover that stopped the clock and set up Utah’s last chance, as opposed to just taking the block or throwing the ball away to let time expire.

Some people understood, though, and humorously pointed out the giant human hurling toward Middleton through space and time. First there was this tweet by a longtime Jazz tweeter and clip-maker.

If you wondered why Middleton traveled on that play, shoot over that 😂 pic.twitter.com/5I3aiftNG3

— Adam (@prodigyJF) November 9, 2019

Then there was this reaction by the Bucks’ TV crew who understood Middleton’s panic.

From the Bucks’ broadcast. Commentary on the forced travel made me chuckle. pic.twitter.com/OjgU5zjUWU

— Ken Clayton (@k_clayt) November 9, 2019

Sacre Bleu!

---

Another week in the books. Seven more things coming next Monday!