Thoughts: 1/14/2020 Let’s reevaluate where the Jazz are and where they’re going.

3rd in the Power Rankings, 10 game win streak, 15-1 in their last 16, league high +11 NetRtg since December 10th, the number 1 offense and number 7 defense in that time. The Jazz have been tearing up the schedule of late, clearly, but how strong is the correlation of the current win streak to actual playoff success? Here’s a cool graphic.

The Jazz current streak started in game 24 and has continued through game 40 so far. It’s kinda funny that they happened to get hot during the least meaningful part of the season. I guess this just tells us what we already know: the team is good enough to beat the bad teams, but when they start to play good teams, we’ll see where they’re really at.

I saw this thought from Spence earlier today, which got me wondering, how often does the actual best team advance? Let’s take to a Twitter poll first to take the objective temperature on this one.

I think it’s fair to assert that the better team doesn’t win 100% of the time, whether due to injuries, poor matchups, or just plain luck. Take a quick look at Raptors topping the Warriors just last year, the 1st seed Mavericks losing to the Warriors in 2007, the 8th seed Grizzlies beating the top ranking Spurs in 2011, and more. Could it be as high as, say, 90% of the time?

Now let’s look at this in a more analytical, definitive light. It’s hard to determine who was truly the better team within a playoff series, since some teams aren’t exactly gunning for wins in the regular season. Thomas Bassine researched this a while back, and just from who had the better regular season NetRtg from 2003-2018, he found that the team with the higher NetRtg won roughly 75% of the time. That’s pretty high! And again, that’s not a definitive definition of who’s a better team. We see better teams lay off the gas all the time in the regular season – I’m looking at you 2019 Raptors and Kawhi’s Load Management. Still, we’ve got a more researched, concrete number.

How can we do better? For a few years now, 538 has been evaluating teams based on their player impact metrics rather than their team season stats. We can run a comparable weighted roster in order to filter out injuries, load management, and just not going 100% in the regular season (LeBron!). Let’s try out giving each player a team rotation number, based on MPG and % of games started. From there, we can weight how many meaningful minutes each rotation player is likely to play in the playoffs against opponents. Below is a look at how I’d weight the roster.

So using these weights, we can see a true strength of a roster. Let’s simply run it with a player impact metric to get team ratings and see just how often the higher rated team wins. As for impact metrics, RPM, RAPM, PIPM, BPM, RAPTOR would all work. I’m planning on using RAPM, but in the future I may run it again with one of the other stats to see if I get different results. All around, though, only a matchup or two out of the 15 per year will swap favorites when changing these stats.

…I’ll continue these thoughts in a future post.

The Jazz just hit 10 wins in a row, which leads to me pulling this graphic from an absolute classic, Basketball on Paper by Dean Oliver. The average team hitting a 10 win streak typically wins 55 games. If the Jazz can pull it up to 15, there’s a decent likelihood they could hit 60. As of right now, 538 has got the Jazz at a predicted 54 wins, BPI has got them at 53, BBref has them at 52, inpredictable has them at an implied (wins + (2.7 * rating +41) / 82* games remaining) 54 wins, and their NetRtg implies 56 wins. We’ll see where they land.

Okay. Enough of me talking about dumb stat and prediction stuff for the night. I’ll possibly write tomorrow on the huge impact Rudy and Joe are having on the Jazz right now. It’s pretty incredible.