Playback speed
×
Share post
Share post at current time
0:00
/
0:00
Transcript

The Analytics One... Feat. Neil Paine

Take a deep dive into Neil's current NBA Forecasting model and estimated RAPTOR—Neil's years at the iconic 538 and his time in the Atlanta Hawks have shaped his modeling systems.

It was great having Neil Paine on the pod, and I can’t wait to continue checking in with him and his forecast model throughout the season.

Neil’s most recent forecast and estimated RAPTOR player ratings are here. For more information about how it works, see below. Also, check out the MLB, NFL, NHL,

Neil’s Substack
🏀 2024-25 NBA Forecast and Estimated RAPTOR Ratings 📈
The following page contains my 2024-25 season predictions for each NBA team, and Estimated RAPTOR player ratings. For more info about how it works, see below. Also check out the MLB, NFL, NHL and WNBA versions…
Read more

As always, The Low Man Help Podcast would not be possible without our outstanding producer, Jacob Sutton

- Check out his work here!

I’ve attached my notes from the Neil Paine podcast, along with a timestamped guide to the episode and a few edits that may or may not be included. Enjoy!

  • 0:01 - 20:13 Neil's Background and the Changing Landscape of Analytics.

  • 20:13 - 27:10 Why The Stats Forecast Hates the Grizzlies (and Marc's Ja Skepticism).

  • 27:10 - 36:14 RAPTOR's Origins & Jokic Breaking The Model.

  • 36:14 - 46:40 Can SGA be a Number One Option on a Championship Team?

  • 46:40 - 1:01:01 Estimated RAPTOR's DeMar DeRozan Love and Quantifying Fear.

Biggest Variance Between Models To Make The Playoffs:

Neil currently has two different versions of the forecast model. One is stats-only and the other is a composite of stats + future odds from Vegas.

Right now there are a few teams that have a wide variance between the models.

Stats-only Model likes these teams more than the composite:

  • MIN (11%)

  • DEN (10%)

  • GSW (10%)

  • IND (8%)

Composite likes these teams more than the stats-only model:

  • PHI (19%)

  • MEM (34%)

    • MEM: Stats-only gives them a 4.8% chance of making the POs vs. the composite model gives them a 30% chance.

Biggest Variance Between Models To Win Championship:

Stats-only Model likes these teams more than the composite:

  • BOS (13%)

  • OKC (10%)

Composite likes these teams more than the stats-only model:

  • NYK: (5%)

Thanks for reading Low Man Help! This post is public so feel free to share it.

Share

Estimated RAPTOR & SGA:

SGA is #1 in estimated RAPTOR - Neil and I met because he shot me a message after reading my SGA piece last year, here is the piece that is referenced in the podcast.

It’s pretty easy to see that SGA has put the time in when no one is watching to develop his game. He has consistently addressed weaknesses and tried to make them strengths.

I have immense respect for how difficult development is once you get to his level in the league.

Three point shooting is a weakness that SGA is trying to develop. He’s currently at 6.2 3PA per game this season, up from 3.6 last year.

To be the best player on a championship team (at his size) SGA needs to be able to stretch the defensive shell in the PO past its comfort zone to create easier 4v3, 3v2, and 2v1 opportunities for teammates more frequently.

Shooting off the PnR action is the easiest way to creates these opportunities in the POs. I believe that SGA and Oklahoma City saw this exact same thing and they are trying their best to address it.

SGA 23/24 Playoff PnR Blitz data via Second Spectrum:

  • During the PO last season, SGA was Blitzed in the PnR 2.75% of the time.

For reference:

(Player - % Blitzed - PnR Per 100 in 23/24 PO)

  1. Curry: 39.1% (PER100 - 30)

  2. Dame: 25% (PER100 - 36)

  3. Ant: 17.9% (PER100 - 48)

  4. Luka: 13.5% (PER100 - 50)

  5. Kyrie: 11.6% (PER100 - 30)

  • In ISO situations, SGA was doubled 4% of the time (6th most):

    • SGA had the most ISO’s PER 100 in 23/24 PO - 28.3.

    • Next closes is Tatum (18.3), then Luka (17.6).

Teams are willing to let SGA get his and keep the rest in check.

Thanks for reading Low Man Help! Subscribe for free to receive new posts and support my work.

DeMar DeRozan and Quantifying FEAR:

REGULAR SEASON VS. PO WIN%

RS: .514

PO: .397

NET: - .117

  • DeRozan 24/25 PnR data via Second Spectrum: Filtered for players above 200 PnR (59 players qualified)

    • 2nd in PPD at 1.211 - DD is incredibly efficient at HIS game.

    • 47th in Blitz% (1.89%) - Go ahead, play your game

    • 58th (2nd to LAST) in qSQ (45.57) - PLEASE take the shots YOU want.

    • 57th (3rd to LAST) in AST% (11.3%) - His action is ONLY about him, it does not generate OPTIONALITY for the team.

The DeRozan PnR usually results in one thing and one thing only, a midrange jumper. It goes in sometimes, it doesn’t go in other times. The point is that teams will let him take as many as he wants, it crates zero FEAR.

  • ISO Per SS: With players above 100 ISO (25 total)

    • 1st in PPD at 1.279 - Again… DD is incredibly efficient at HIS game.

    • 14th in Double % (3.05%) - Middle of the pack here.

    • 25th (Last) in qSQ (41.87) - PLEASE PLEASE PLEASE take the shots YOU want.

Estimated RAPTOR and other models love DeRozan. He is efficient and one of the best in the world at HIS game, but HIS game does not create a lot of opportunities for teammates and has a pretty low ceiling in the playoffs.

Congratulations on Surviving Thanksgiving From Marzie!

Low Man Help is free. But if you enjoyed this post, you can support Low Man Help by buying a coffee. This money will be reinvested into LMH to create better content for you!

Buy Me A Coffee

Intelligentsia House Blend fuels LMH and Marzie, and I can’t get enough of it!

Discussion about this podcast