FanPost

Titles Expected vs Titles Won


I thought some people might enjoy some analytics I was working on this morning.

In the question of who is the "greatest of all time" analysts, fans and former players always compare candidates based on the following (among others like awards and career stats):

  • # of Titles Won
  • Record in NBA Finals
This has always irked me because it doesn't take into consideration the strength of the teams that player played on. It's obviously not as impressive if you win a title with the #1 ranked team right? Does it really mean anything if a player played their entire career on losing teams and never even had a legitimate chance to win a title?

I decided I would crunch some numbers and see how a few of the "greatest of all time" performed with respect to the expectations their teams had to start the season.

I took preseason title odds data from Basketball Reference (see https://www.basketball-reference.com/leagues/NBA_2022_preseason_odds.html). These odds go back to the 1984-85 NBA season. I then used those odds to compute the likelihood a team would win the title based on preseason rank. I found all of the following:

  • 1st accounted for nearly 50% of all wins (18/37)
  • 1st and 2nd rank were most prevalent accounting for 70% of wins
  • 2nd accounted for more wins than 3rd to 5th combined
  • Ranks from 3rd to 8th didn't show any logical distribution
  • 9th or lower resulted in no titles
Given that I assigned a win probability based on rank which can be found below

  • 1st (18) - 0.49 (49%)
  • 2nd (8) - 0.22 (22%)
  • 3rd to 5th (7) - 0.06 (6%)
  • 6th to 8th (4) - 0.04 (4%)
  • 9th and lower (0) - 0.00 (0%)
These buckets could be improved by:
  • Considering conference rank. I think a team picked 3rd or 4th but 1st in their conference should probably get a higher probability of winning
  • These are gambling odds. Public perception impacts them. Some minor adjustments could be made to account for this (e.g. lowering likelihood for defending champions)
  • 9th or lower ranks should at least have some minimal probability (I could use the gambling odds to assign this)

I then tabulated the expected titles for nearly all of the NBA players on the Top 75 list whose career started during the 1984-85 season or later. There were a few special circumstances that should be noted:

  • Injuries - If a player played a single game for their team that season it was included. Probabilities could be adjusted due to this to improve accuracy
  • Trades - The team used for the player was the one they started the season playing for. I don't think there was a scenario for any of the players below where they were traded to a title winning team. But I do know of a few mid-season trades involving these players that greatly increased the destination teams chance of winning
Here are the players ranked by greatest difference between titles won and titles expected:

  1. Jordan +3.44
  2. Duncan +1.86
  3. Curry +1.22
  4. Hakeem +1.11
  5. Robinson +0.95
  6. Kawhi +0.86
  7. Giannis +0.65
  8. Wade +0.52
  9. Shaq +0.05
  10. Durant -0.03
  11. Kobe -0.04
  12. Iverson -0.06
  13. Nowitzski -0.34
  14. Harden -0.42
  15. CP3 -0.49
  16. Lebron -0.50
  17. Barkley -1.14
  18. Malone -1.26

This FanPost is a submission from a member of the mighty Golden State of Mind community. While we're all here to throw up that W, these words do not necessarily reflect the views of the GSoM Crew. Still, chances are the preceding post is Unstoppable Baby!