You Are Here > Basketball-Reference.com > BBR Blog > NBA and College Basketball Analysis

SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all Basketball-Reference content, so you can quickly and easily find the content you want.

Also, our existing Basketball-Reference blog rss feed will be redirected to the new site's feed.

Basketball-Reference.com // Sports Reference

For more from Neil, check out his new work at BasketballProspectus.com.

How Many Titles Should the Cleveland Cavaliers Have Won During the LeBron Era?

Posted by Neil Paine on July 15, 2010

One common media refrain when criticizing LeBron James' decision to "take his talents to South Beach" has been the idea that he left behind unfinished business in Cleveland. He and the Cavs posted consecutive 60+ win seasons in 2009 & 2010, each time securing the #1 record (and top playoff seed) in the Eastern Conference, but in both years Cleveland flamed out early. Many have used this as supposed "proof" of some character flaw on the part of James and his teammates, but what was the probability that this could have simply happened due to random chance alone?

To answer this question, I set up a very basic Monte Carlo simulation using the regular-season winning percentages of all playoff teams since the Cavs' first playoff appearance of the James era (2006). 10,000 times, I simulated the playoffs for each season, taking into account the postseason bracket & home-court advantage effects, and I recorded the team that won the Finals in each simulation. Here's how it broke down for each season:

Year Team W L Championship%
2010 Cleveland Cavaliers 61 21 36.8%
2010 Orlando Magic 59 23 21.1%
2010 Los Angeles Lakers 57 25 13.2%
2010 Dallas Mavericks 55 27 7.9%
2010 Phoenix Suns 54 28 5.3%
2010 Atlanta Hawks 53 29 3.7%
2010 Denver Nuggets 53 29 3.4%
2010 Utah Jazz 53 29 3.0%
2010 Boston Celtics 50 32 1.4%
2010 Oklahoma City Thunder 50 32 1.3%
2010 San Antonio Spurs 50 32 1.2%
2010 Portland Trail Blazers 50 32 0.9%
2010 Miami Heat 47 35 0.5%
2010 Milwaukee Bucks 46 36 0.3%
2010 Charlotte Bobcats 44 38 0.1%
2010 Chicago Bulls 41 41 0.1%
2009 Cleveland Cavaliers 66 16 44.7%
2009 Los Angeles Lakers 65 17 34.2%
2009 Boston Celtics 62 20 11.3%
2009 Orlando Magic 59 23 4.2%
2009 San Antonio Spurs 54 28 1.8%
2009 Denver Nuggets 54 28 1.6%
2009 Portland Trail Blazers 54 28 1.0%
2009 Houston Rockets 53 29 0.6%
2009 Dallas Mavericks 50 32 0.3%
2009 New Orleans Hornets 49 33 0.2%
2009 Atlanta Hawks 47 35 0.2%
2009 Utah Jazz 48 34 0.1%
2009 Miami Heat 43 39 0.0%
2009 Chicago Bulls 41 41 0.0%
2009 Detroit Pistons 39 43 0.0%
2009 Philadelphia 76ers 41 41 0.0%
2008 Boston Celtics 66 16 63.5%
2008 Detroit Pistons 59 23 12.2%
2008 Los Angeles Lakers 57 25 6.4%
2008 New Orleans Hornets 56 26 4.6%
2008 San Antonio Spurs 56 26 3.9%
2008 Houston Rockets 55 27 2.6%
2008 Phoenix Suns 55 27 2.4%
2008 Utah Jazz 54 28 1.8%
2008 Orlando Magic 52 30 1.5%
2008 Dallas Mavericks 51 31 0.5%
2008 Denver Nuggets 50 32 0.4%
2008 Cleveland Cavaliers 45 37 0.1%
2008 Washington Wizards 43 39 0.1%
2008 Philadelphia 76ers 40 42 0.0%
2008 Toronto Raptors 41 41 0.0%
2008 Atlanta Hawks 37 45 0.0%
2007 Dallas Mavericks 67 15 64.2%
2007 Phoenix Suns 61 21 15.5%
2007 San Antonio Spurs 58 24 6.1%
2007 Detroit Pistons 53 29 6.0%
2007 Cleveland Cavaliers 50 32 3.0%
2007 Chicago Bulls 49 33 1.7%
2007 Toronto Raptors 47 35 1.3%
2007 Houston Rockets 52 30 0.9%
2007 Utah Jazz 51 31 0.6%
2007 Miami Heat 44 38 0.3%
2007 New Jersey Nets 41 41 0.1%
2007 Orlando Magic 40 42 0.1%
2007 Washington Wizards 41 41 0.1%
2007 Golden State Warriors 42 40 0.0%
2007 Denver Nuggets 45 37 0.0%
2007 Los Angeles Lakers 42 40 0.0%
2006 Detroit Pistons 64 18 49.2%
2006 San Antonio Spurs 63 19 28.5%
2006 Dallas Mavericks 60 22 11.4%
2006 Phoenix Suns 54 28 4.1%
2006 Miami Heat 52 30 3.0%
2006 Cleveland Cavaliers 50 32 1.2%
2006 New Jersey Nets 49 33 1.1%
2006 Los Angeles Clippers 47 35 0.6%
2006 Memphis Grizzlies 49 33 0.3%
2006 Los Angeles Lakers 45 37 0.2%
2006 Denver Nuggets 44 38 0.1%
2006 Chicago Bulls 41 41 0.1%
2006 Washington Wizards 42 40 0.1%
2006 Indiana Pacers 41 41 0.1%
2006 Sacramento Kings 44 38 0.0%
2006 Milwaukee Bucks 40 42 0.0%

As you can see, despite having the league's best record over the past two seasons, Cleveland had less than a 45% chance of winning each year even if they had played at exactly the same level as they had during the regular season. However, calculating the expected odds of them being shut out twice in two years requires some additional math: since we know the probability of the Cavs not winning in 2010 was (1 - 0.368) = 0.632, and the probability of them not winning in 2009 was (1 - 0.447) = 0.553, the odds of them being ringless in consecutive seasons was (0.632 * 0.553) = 35%. Certainly that's still within the realm of reasonable probability, but the inverse is that they had a 65% probability of winning at least one ring in the last 2 seasons, and they failed to deliver.

If you expanded things to look at the entire LeBron era, you'd find that Cleveland won 0.8579 titles on average during the 5 years the Cavs made the postseason. Here's a frequency table detailing how often they won each # of titles during that span:

# of Titles Frequency
0 33.5%
1 48.0%
2 17.8%
3 0.7%
4 0.0%
5 0.0%

In this sense, the LeBron James era in Cleveland was absolutely a disappointment. The Cavaliers had a 66.5% probability of giving Cleveland at least 1 NBA championship during his 5-year run there, with the most common outcome being exactly 1 title, and they were unable to make it happen... That's nothing if not unfinished business.

That said, there's still roughly a one-third chance that the Cavs' recent postseason failings have been due to random chance alone. Unfortunately, this is not exactly solace to poor Cleveland fans, because if that's true it means they happened to end up in the unlucky 1⁄3 of alternate universes that featured James skipping town without any hardware. Man, and you know things are rough for your city when even the multiverse is against you.

ShareThis

117 Responses to “How Many Titles Should the Cleveland Cavaliers Have Won During the LeBron Era?”

  1. huevonkiller Says:

    2009 makes sense to me. The Cavs struggled against the Magic throughout the season, although LeBron played great.

    The Mo Williams effect is strong in both seasons. With a little help from Antawn.

  2. P Middy Says:

    Mo is a huge culprit. In 2009 his from FG% dropped 47% to 41%. In 2010 it dropped from 44% to 41%. In 2009 his 3FG% dropped from 44% to 37%. In 2010 his 3FG% dropped from 43% to 33%.

  3. Romain Says:

    Neil, it might be interesting to look at this from a historical perspective:

    - How many teams have posted back to back 60+ wins regular seasons and fail to win a championship? I can think of the Mavs (2006 & 2007) and the Jazz (1997 & 1998), is there any other exemple? How about failing to reach the Finals even once? I'm not sure this had ever happened.

    - Same questions with teams leading the regular season in back to back seasons?

  4. themojojedi Says:

    Neil, don't you often use/prefer an SRS based approach or Pythagorean Win % in these types of scenarios. That approach would probably add a few percentage points to 2009 and knock a few off 2010.

  5. ScottR. Says:

    James is the A-Rod of basketball, not original I know. He's a great talent but lacks something other great champions have--performance in the clutch and ability to carry a team on his back when it counts. He's more interested in the glamour of sports than the grit. If he put as much effort into the Boston series as he did into "The Decision" maybe his team advances. Seeing him and Wade and Bosh mugging it up in South Beach with all the other phonies made me realize he's exactly where he should be.

    I read somewhere that maybe it's our fault for expecting more of Lebron than he really had to give.

  6. themojojedi Says:

    Romain, by my (late-night) count there have been 18 sets of back-to-back 60 win seasons in NBA history. I've listed them by highest level of achievement in the two year span:

    At least one title: 14 times (BOS 80+81, BOS 81+82, BOS 84+85, BOS 85+86, BOS 08+09, CHI 91+92, CHI 96 + 97, CHI 97+98, LAL 72+73, LAL 85+86, LAL 86+87, LAL 87+88, MIL 71+72, PHI 67+68)
    Lost Finals: 2 times (DAL 06+07, UTA 97+98)
    Lost to eventual Champion in Conference Finals: 1 time (MIL 72+73)
    Neither won Finals, made Finals or lost to eventual champion: 1 time (CLE 09+10)

  7. Neil Paine Says:

    I usually do use SRS for things like this, but nobody is saying "Cleveland had 2 back to back years with at least a 6.0 SRS and didn't win." What they are saying is, "Cleveland had back-to-back 60-win seasons and didn't win," so I framed the discussion in a way that more fans would understand. Eyeballing their pythagorean wins, I don't think using SRS would change things very much vs. using WPct.

  8. kaveh Says:

    An error in your analysis is calculating the odds of winning a championship, when a more relevant analysis would have been calculating the odds of the Cavs getting knocked out of the playoffs when they did. Just calculating the odds of not making it to the NBA finals for the last 2 years despite their circumstaces would most likely boost the figure from 65% to 90%! Forget about the fact that this last year they didn't even get out of the 2nd round of the playoffs!

    A number in the 90% range is far beyond any random variable. It shows just how badly Lebron and the Cavs failed.

  9. Neil Paine Says:

    Also, is it just me or did it seem like LeBron didn't really put that much effort into "The Decision" at all? He clearly didn't understand and/or give any prior thought to the implications of the show to his image, we learned later that it was Jim Gray & his agent's joint idea (he just stupidly went along with it), and the entire time he looked as rattled as I've ever seen him, like he wanted to be anywhere but on that stage with Gray. For someone who said he was "enjoying the process", he looked terrified.

  10. themojojedi Says:

    "I framed the discussion in a way that more fans would understand."

    Totally reasonable, and as you suggested its unlikely that minor changes in percentages would change your conclusions or your overall message.

    "Nobody is saying "Cleveland had 2 back to back years with at least a 6.0 SRS and didn't win." "

    No kidding you, but I actually was thinking EXACTLY this and went so far as to have a look into it :) Looking at it this way actually gives them at least a few "failure friends". Could the 93-98 Sonics please take a bow for underachieving!

    Adding to my previous post 6. if you want to count shorter seasons (and the BAA) where teams posted win-loss percentages equivalent to 60 win seasons you could also add:

    At least one title: BOS 64+65, MNL 49+50 taking the tally to 16 times
    Lost Finals: UTA 98+99 taking the tally to 3 times
    Lost to eventual Champion in Conference Finals: ROC 49+50 taking the tally to 2 times
    Neither won Finals, made Finals or lost to eventual champion: Sorry, still just Cleveland

  11. Romain Says:

    Thank you themojojedi.

    So the Cavs underperformed in the playoffs in a historical way... Not a surprise but it's still impressive to see them in a class of their own in that category.
    What makes it even worse is that they lost each time in 6 games, and each time there was no tough break or questionable call to use as an excuse... They were sumply outplayed.

  12. Neil Paine Says:

    Re #8 - OK, here are the probabilities that they'd reach each round:

    Round 2006 2007 2008 2009 2010
    Finals 5.1% 23.7% 0.7% 65.4% 51.9%
    ECF 10.4% 49.5% 3.6% 92.7% 78.2%
    ECS 71.8% 75.2% 57.5% 98.0% 92.9%
  13. Neil Paine Says:

    So Cleveland not making the Finals in both 2009 & 2010 had a 16.6% probability of happening due to chance alone.

  14. Neil Paine Says:

    Also, a frequency chart for the total # of times they should have reached each round in LBJ's six years:

    # of Times Finals App CF App CSF App
    0 11.9% 0.7% 0.0%
    1 40.4% 11.9% 0.3%
    2 37.4% 45.1% 4.7%
    3 9.8% 36.9% 23.2%
    4 0.6% 5.2% 43.2%
    5 0.0% 0.1% 28.7%
  15. Neil Paine Says:

    1 Finals appearance in 6 years is actually quite in line with expectations (it's the most likely outcome), it just didn't happen in the "right" year -- in 2007 they made the Finals with only a 24% chance of doing so, and in 09/10 they missed it both years with an 83% chance of getting there at least once.

  16. Spree Says:

    Neil, can you do these probabilities for this Heat team?

    If it is under 65%, then the Heat are in trouble considering LeBron's recent history.

  17. Bob Says:

    I'm not sure what losing in the playoffs despite being favored has to do with being Clutch. Lebron has been about as clutch as anyone I've seen in the playoffs, see Wizards Series, Game 5 vs Detroit, Game 7 vs Celtics, Game 2 and 5 vs Orlando, Game 2, and 3(despite the loss) vs Chicago. Game 3 vs Atlanta (probably his best performance of carrying a team on his back for an entire 48 minutes)

  18. Anon Says:

    "James is the A-Rod of basketball, not original I know. He's a great talent but lacks something other great champions have--performance in the clutch and ability to carry a team on his back when it counts."

    Someone needs to go back and read the past like, 3 entries on this blog that address this matter.

    By the way, I just love it how people take the likelihood of something as an absolute guarantee. Just because something is favored to happen, doesn't mean it WILL happen. It's called "chance". Perhaps going to Miami will helps LeBron get that ring he covets.

  19. ScottR. Says:

    Bob- Anyone who watched the Boston series knows Lebron was dogging it at best and quit at worst. Even if West was doing his mom that's no excuse.

    Anon-I agree that going to Miami will help Lebron get a ring. Do I think it's guaranteed? No.

    Lebron has Jordan-like talent, but also has a Pippen like mentality. And that's why he will never be considered one of the best ever.

  20. Anon Says:

    "Anyone who watched the Boston series knows Lebron was dogging it at best and quit at worst. Even if West was doing his mom that's no excuse."

    And anyone who watched that series also knew that it wouldn't matter either way when your 2nd best player couldn't find the basket with an NSA satellite.

    Not even gonna address the last statement in your post.

  21. Jason J Says:

    Neil - this is an interesting post. I'd like to see how the 1993 & 1994 Knicks stack up this way.

    To people who have decided that LeBron is a regular season stud who doesn't have the chops to win in the playoffs, I'd just like to point out that basketball is a team sport.

    LeBron's terrible playoff history:

    2006 - Out-dueled Gilbert Arenas and made clutch plays late to win crucial game 5 in Washington and help the Cavs advance to the second round. http://www.basketball-reference.com/boxscores/200605030CLE.html

    2007 - Took over crucial game 5 against 60+ win Detroit team to take control of the ECF and brought a limited Cavs team to the NBA Finals. http://www.basketball-reference.com/boxscores/200705310DET.html

    2008 - Had an enormous game 7 while his teammates collapsed around him against an all-time great Celtics defense. http://www.basketball-reference.com/boxscores/200805180BOS.html

    2009 - Set the PER record of a playoff season. Hit an incredible game winner to steal one against the Magic in the ECF. Had this ridiculous triple double to stave off elimination: http://www.basketball-reference.com/boxscores/200905280CLE.html

    2010 - Had this great game against a tough Celtics defense with an injured elbow after annihilating Chicago in games 2 and 4 of that series. http://www.basketball-reference.com/boxscores/201005070BOS.html

  22. Ahgo Says:

    Building an expectation mechanism from regular season statistics alone usually leads to dubious findings. The Cleveland Cavaliers did not have a 2/3rds chance of winning a championship ever because when it comes to playoff basketball, Lebron's supporting cast was not sufifciently talented, intelligent, or experienced to make the necessary overall adjustments to win a series against a 55+ win team.

    The incongruency between Cleveland's regular season and post season success the last two years is highlighted because Lebron James was in Cleveland. And that's fine. But the difference is fairly commonplace for teams that aren't built properly for championship runs. Phoenix and Dallas, it seems, suffered from even worse statistical fates over the decade than Cleveland did in the 2010. Of course, that won't receive any press coverage because neither team ever had as beastily an "alpha dog" as Lebron (or, arguably, an alpha dog to begin with). Still, both teams put up beastly regular reasons records for considerable stretches in the 00s, only to struggle regularly to close out series' in the playoffs and get eliminated earlier than expected.

  23. Jason J Says:

    How come I can't comment? Did I get banned?

  24. Jason J Says:

    Nope.

  25. Jason J Says:

    18 - Agreed. It's also not as though he didn't do his part the majority of the time. Do people forget the game winner last year? Game 5 against the Pistons in 2007? Game 7 against Boston in 2008? Game 5 against the Wizards in 2006? Game 4 against Boston this year? Every year since he started making the playoffs he's been friggin' good in the playoffs. Not sure what else people expect.

  26. Ryan Lasorsa Says:

    Nice analysis I've actually been waiting to see how statstics factored into the whole LeBron hype and I can't wait to see a Monte Carlo simulation for the 2010 season and see how the Heat are projected and how they actually do

  27. Neil Paine Says:

    Jason - I think Akismet blocked it because of the links... To our own site, even! Sometimes comments get filtered for no good reason, and I need to figure out what to do about it. (This is my apology/explanation for future blocked comments.)

  28. Jason J Says:

    Apoxplanation accepted.

  29. P Middy Says:

    Always nice when the analysis backs up the empirical evidence. The Cavs blew it in 09 and 10.

  30. Jai Says:

    That's exactly what I'm sayin. I’m sayin that LeBron’s inexplicable choke in the playoffs wasn’t inexplicable at all. I think what happened was a light switch went on. He said to himself if we win the Eastern Conf Championship I’ll be stuck in Cleveland for life. I’ll have no justification for leaving. So what does he do? He tanks it!! Go look at every game LeBron played last year and you won’t find any remotely close to as bad as LeBron played in that Championship. A real champion would’ve acknowledged he played below his standards and vowed to return. A fake Champion jumps ship and leaves his shipmates to fend for themselves. Well LeBron, on paper you should finally win a championship. Just so you know, Championships taste sweet. If you win, yours will have a BITTER TASTE. Not so fast though, Kobe still has a sweet tooth

  31. Jason J Says:

    Thank the lord Kobe never tried to do that exact same thing after obviously throwing a game 7 against Phoenix in the summer of 2007 only to have LA refuse to trade him. Otherwise he'd be a fake champion too.

  32. AYC Says:

    Jason J (#21) is absolutely right; Lebron was in the same situation as Jordan in the late 80's; a dominant player with a mediocre team. In effect we are blaming LBJ for getting an average team to greatly overachieve in the reg season

  33. Daniel Song Says:

    Neil,

    Unfortunately, NBA Playoffs does not yield readily to statistical analysis. I used to bet NBA futures heavily before the online gambling ban, and found that regular statistics is complete, total, and utter garbage. It's such a bad predictor of playoff results that I was forced to scrap them almost immediately. It's a classic case of garbage in, garbage out.

    The best way to handicap teams' chances is to look at the Las Vegas lines before the first game, and try to project how the team will develop as it advances in the playoffs. Without applying subjective analysis, you won't get anywhere. I did this for years and had a winning season each and every year before the ban.

    Using subjective analysis, here are the numbers I got:

    Cleveland's % of winning Championship
    2006: 0.0
    2007: 4.6
    2008: 0.5
    2009: 36.7
    2010: 30.0

    Expected number of titles: 0.718

    # Titles Frequency
    0 42.1%
    1 44.7%
    2 12.5%
    3 0.7%
    4 0.0%
    5 0.0%

    It's tough to win a championship. Only one team can win, and Boston and LA have clearly been the class the last three seasons - with Orlando the only other team making a Finals appearance.

    Cleveland had their chances, but it's a difficult road. Dallas, Phoenix, Sacramento, New Jersey, etc. have all knocked on the door in the past and came up short. Heck, even a dominant Lakers squad came up short more often than not (though 5 of 11 is pretty good). There is no "should have" in this game - you either pull it off or you don't.

  34. James Katt Says:

    If you replace Lebron with Michael Jordan in his prime, Cleveland would have won 1-2 championships. Michael was totally focused on winning. He influenced his teams to achieve results over and above what they would have otherwise achieved. Lebron is unfocused on winning. He wanted to play and dance more than win. Lebron also frets and worries excessively when he has to focus on winning. He chokes. He is only saved on occasion by his physical talent. Michael was fearless. Lebron is fearful. Lebron simply does not have the mental fortitude to win.

  35. ScottR. Says:

    AYC--yeah Jordan was on some awful teams--Brad Sellers? Dave Corzine?--but he stayed the course and eventually got his rings. And even when MJ didn't win, like against the Celtics in 86 or Pistons in 88 or 89 he gave it everything he had, unlike James who I know think is the ultimate front runner.

    I agree that MJ had much better players than James did in Cleveland. Would Cleveland ever have gotten a good enough core around James to win? We will never know because so-called "King" didn't have the heart to stick it out. Had James left when he was 30 like KG no one would have cared, but to leave a 50 plus win team during your prime means you are a coward.

  36. Daniel Song Says:

    FYI, my subjective analysis projects a win total in the mid-to-high 50's for next year's Miami squad and about a 20% chance of winning it all. Depends on how things shake out with Orlando, Boston, and Chicago. Miami probably rates as co-favorites along with Orlando and Boston in the East and would be a considerable underdog vs. the Lakers in the finals, should both teams make it that far.

  37. Neil Paine Says:

    Here's an interesting thought...

    So everyone is saying LeBron took the coward's way out, that he was afraid of losing, that he had to team up with Wade and Bosh to win, etc. And it's true that he admitted fear -- he said he was afraid of being Minnesota-era Garnett, of being "31 with bad knees and no championships". He realizes that, fairly or unfairly, championships are what all-time great players are judged on, and right now he's lagging behind other greats (Bill Russell & Kobe had 3 by age 25, etc.).

    So here's where it gets savvy: he knows that championships are the measuring stick, but he also recognizes that the playoffs are a crapshoot. Yes, they're less of a crapshoot than in MLB or the NFL, where the best team wins 25% of the time (or more accurately, loses early 75% of the time due to random chance alone), but the typical best team in the NBA still only wins 50% of the time. And this isn't the team with the best record we're talking about, this is the literal best team with the most absolute talent -- they still lose prematurely half the time due to randomness.

    So maybe LeBron realized that he needed to load up, maximize his chances and create a superteam if he wants to be sure (or as close as possible) of a championship, since the raw number of championships won is going to be how people view his career historically. (Think people will care that he did it with D-Wade if James wins Finals MVP for one of the titles? People used to "discount" 3 of Kobe's rings because of Shaq; now they just say "Kobe has 5 rings, if he gets to 6 he's as good as MJ!")

    I'm not saying LeBron thought about it and ran the probabilities (yeah right), but I am saying Garnett imparted some wisdom about sticking around in a small market and trying to chase 1 ring without much help. They intuitively realized that even if you have the best team, you still lose half the time just because of luck, so you'd better get yours while you can. What better way to do that than to jump to a historically loaded team?

  38. Neil Paine Says:

    33. I don't think that it's necessarily as extreme as you're making it out to be. Justin won back-to-back stat geek smackdown championships using only regular-season SRS as a predictor.

  39. Daniel Song Says:

    James,

    It's difficult to win championships. Jordan had pretty similar talent on his roster in his first 5 seasons as Lebron did in his 7-year run. Pippen and Grant had already started to emerge in seasons 4 and 5, and Cartwright was pretty solid - definitely comparable, if not better than, the talent Lebron had in his last two seasons.

    With that talent, the Bulls won 47 and 50 games, and never made the NBA Finals. Before that, they never made it past the first round. So I think Jordan would've definitely had problems winning in Cleveland with the roster they had.

    Later the Bulls acquired more talent and won 6 championships, fielding some of the greatest teams in NBA history. In fact, they won 55 games and lost in the second round to would-be finalists Knicks in the year Jordan was gone.

    Bulls had some of the greatest teams in NBA history (including THE greatest in '96) and Jordan was a major, major part of the team. However, the team was stacked and it was the team effort that won all those championships.

  40. Neil Paine Says:

    I like how James Katt is suddenly a licensed psychologist.

  41. RDubya Says:

    Would it be a more accurate (but still frought with problems) projection would be to factor in the win/loss percentage of the regular season games vs. the teams the Cavs played against in the post season (Pistons, Hawks and Magic in 2009 / Bulls and Celtics in 2010)?

    Also, how about the coaching matchup factor? For example, there's no way the 2006/7 Warriors were better than the 2006/7 Mavericks, but Nellie knew the Mavs system (and Avery Johnson) enough to exploit every weakness... only to get trounced by Sloan's Jazz in the next round.

  42. dylan Says:

    Well said Jason J..LeBron definitely has some character issues, but you showed evidence against the "lebron can't step it up in the playoffs." People like playing revisionist history with Kobe's unsuccessful seasons (see seasons w/out shaq/gasol)

  43. Daniel Song Says:

    Neil,

    It IS that extreme and I have several years of experience in gambling markets to show for it. I played the futures in all the major sports (baseball, basketball, and football) and concentrated heavily on the postseason.

    It's a totally different game in the playoffs. Teams with the championship mindset step and and perform above and beyond their regular season numbers. Other teams find a way to fade. If you can isolate such teams, you can find huge weaknesses in the futures market.

    Justin deserves his props for winning the stat geek smackdown, but that's a whole different ball game. It's one thing to predict the winner of a series; it is quite another to compute accurate odds for each series, knowing how much to bet on each proposition, and making adjustments on a day-by-day basis. I've seen many a stat geek try to take their game to the gambling arena. Most of them don't make it. Those that do learn the ins-and-outs of subjective handicapping and the art of line shopping. Bottom line, regular season stats will not generate accurate playoff predictions.

    Based on what I saw from the ESPN stat geek smackdown (and John Hollinger), the vast majority don't have what it takes to make it. Their numbers simply aren't good enough. It's one thing to say that Cleveland has the best team, and quite another to be able to assess that they have about a 30% chance of winning it all. Without being able to do the latter, you won't be able to consistently make good predictions over the long run.

  44. Neil Paine Says:

    Okay, so what's your method if you don't use statistical analysis?

    EDIT: OK, I see, you outlined it above:

    The best way to handicap teams' chances is to look at the Las Vegas lines before the first game, and try to project how the team will develop as it advances in the playoffs. Without applying subjective analysis, you won't get anywhere.

    The problem with subjective analysis is, um, it's subjective.

  45. Neil Paine Says:

    Also, remember that the point here isn't to predict (in this case, retrodict) who was going to win the championship. It is to create a naive set of expectations that serve as a proxy for the conventional wisdom on how often Cleveland "should" have won. In other words, it's putting a naive probability on statements like "Cleveland won 60+ games twice but they didn't win a title, what a choke!"

  46. Neil Paine Says:

    Furthermore, I'm not sure how much "different" the game is. Looking at the league average four factors in this year's regular season vs. playoffs...

    * The pace slows down (89.7 vs 92.7)
    * Shooting basically stays the same (.496 eFG% vs .501)
    * Turnovers go down (.128 per poss. vs .133)
    * Rebounding stays exactly the same (.263 ORb% in playoffs, .263 ORb% in RS)
    * They call more fouls (.262 FT/FGA vs .228)

    Different? Yes. But it's not like they're suddenly playing a totally different game. You just find the teams who are best poised to take advantage of the turnover and foul-calling changes.

  47. Daniel Song Says:

    Neil,

    I have tremendous respect for the betting market and that is always where I start. Places like Matchbook and Betfair are two of the best information sites. Tradesports was one of the giants but they folded due to the gambling ban. They're also great places to check your final numbers; if it's not a reasonable facsimile of market odds, chances are good that it's you, and not the market, that's wrong. You're working with small edges for the most part; my lifetime rate on return was something like 2-3%.

    Second, you need to watch numbers like a hawk and keep track of line movements. Look at previous lines, current lines, future lines, and line history. It's a great source of information and it helps you gauge public perception of the teams. Often teams will be overrated or underrated; with experience, you will be able to sniff out the disparities between perception and reality.

    Third, I do enjoy reading articles like yours, and articles from Truehoop, ESPN.com, and other popular media. They are a great source of entertainment. They also help shape public opinion and sometimes influence gambling odds.

    In all, my key to making good predictions was to bet often, keep track of betting markets, and locate biases in the betting public. It was an accumulation effect; knowledge and instincts gained over the course of years of experience and thousands of wagers.

  48. Jason J Says:

    40 - I like this revisionist notion that everything that happened in his career was by Jordan's choice. Mike's my boy and everything. I got a jumpman logo tattooed to me, so I'm biased as hell, but anyone who thinks Jordan sticking around through the hardtimes makes him a better person or player than LeBron probably doesn't remember that Jordan's first real taste of free agency came in 1996 after winning 72 games and a title.

    Who knows what he would have done if he could have left after losing to Detroit for the third time in a row in 1990? He might have said, "Screw it. Boston is one athletic wing away from being dominant, and I want to play with a bunch of veterans who know how to win and will make me look good. Mr. Falk set up a meeting with Red."

    He also had young, improving star talent on his team by the late '80s, which is something LeBron has not ever had. His position was improving while LeBron's could be seen as deteriorating. Maybe.

  49. Neil Paine Says:

    Daniel, I will say that I find the prediction markets very interesting, especially as a source of within-game probability:

    http://www.hardballtimes.com/main/article/prediction-markets-redux/

  50. Neil Paine Says:

    Here is an interesting APBRmetrics thread on this very topic:

    http://sonicscentral.com/apbrmetrics/viewtopic.php?p=21451

  51. Daniel Song Says:

    Neil,

    Thank you for your references.

    I freely admit that I used to be a stat geek back a while back. I also couldn't predict my way out of a paper bag and I had zero chance of beating the books.

    Once I started studying, experiencing, and betting prediction markets I improved quite a bit. I learned a lot of tricks and methods that I could have never learned by simply crunching numbers.

    I still do quite a bit of statistical analysis. However, my method is now more of a hybrid - one that requires subjective input for key variables. After I guesstimate these variables I use the computer to pump out the answers.

    From what I see, the most successful financial models work the same way. Those that don't often wreak havoc - as the sub-prime meltdown, "flash crash", and stock market bubbles attest.

  52. Anon x 2 Says:

    Neil, very interesting.

    Hey, I wonder what Kobe's (since becoming a starter), Duncan's, Nash's (in Phx), and Dirk's numbers are. Besides Lebron and Shaq, they have had the most chances on a single team to win a title.

    I'd love to see how they break down. In TD and Kobe's case, expected total titles would probably come into play.

  53. Anon x 2 Says:

    "the typical best team in the NBA still only wins 50% of the time. And this isn't the team with the best record we're talking about, this is the literal best team with the most absolute talent -- they still lose prematurely half the time due to randomness."

    What are you basing this on? I mean, what qualifies the "best team with the most absolute talent?" Point differential? Some other measure?

  54. P Middy Says:

    I know they're not the most smartest dudes ever, but Wilbon and Kornheiser constantly rag on betting lines because they're designed to induce gambling, rather than accurately reflect statistical probabilities.

  55. Daniel Song Says:

    P Middy,

    They're not the smartest dudes ever.

    I've done quite a bit of statistical work. The single best predictor of statistical probabilities is the betting line.

    Mistakes do happen but you're working with thin edges. The BEST pro gamblers I've known work with a 1-2% edge on the house and makes dozens of wagers daily.

  56. Ahlaker Says:

    @37

    "What better way to do that than to jump to a historically loaded team?"

    As I mentioned in another thread, wouldn't it be heaps better if he signed for the mid-level with Orlando, Boston or LA? Surely a team with Kobe, Pau, LBJ, Bynum and any scrub point guard is better than this Heat incarnation isn't it?

    The guy is already a multi-millionaire, if he really (I mean *really*) wants championships he goes to the best teams - right? Sure it's pie in the sky and he's too greedy to ever do it (not to mention the union might shoot him) but that's what he should've done if he really wanted rings.

  57. J.B Says:

    sorry, but this is just like a scientist reporting only the figures that support his case.

    You can't make ASSUMPTIONS - you're assuming that the results extrapolated indicate that the catalyst for the cavs (LeBron) was the main cause of failure. Their are way too many factors not accounted for that will influence this result. Individual performance (reg & post), title history (team, players & coach), Injuries etc...

    Nice use of math, but come on, you can do better!

  58. Spree Says:

    "So here's where it gets savvy: he knows that championships are the measuring stick, but he also recognizes that the playoffs are a crapshoot. Yes, they're less of a crapshoot than in MLB or the NFL, where the best team wins 25% of the time (or more accurately, loses early 75% of the time due to random chance alone), but the typical best team in the NBA still only wins 50% of the time. And this isn't the team with the best record we're talking about, this is the literal best team with the most absolute talent -- they still lose prematurely half the time due to randomness."

    Neil,

    With the Lakers and Celtics winning half the championships since Russell's time and the best team still only winning 50% of the time how many times were they not the best team?

  59. huevonkiller Says:

    The Heat are not in trouble, Spree. If anything I expect a Bulls-like Dynasty now (assuming no lock-out issues).

  60. Neil Paine Says:

    # 53,58:

    http://www.basketball-reference.com/blog/?p=3555

    It only applies to the NBA's current 30-team, 16-in-the-playoffs structure, but in the simulation we know which team was the most talented, and that team won only 48% of the time. Doug did a variation on this at PFR as well:

    http://www.pro-football-reference.com/blog/?p=57

    In all sports, the best team wins an uncomfortably infrequent % of the time.

  61. Ahgo Says:

    Isn't the biggest difference between the playoffs and regular seasons - and a major one at that - is you play the same team up to 7 times, thereby allowing the superior team to make the best adjustments and exploit matchups over a larger sample size (4 - 7 games in a row as opposed to 1).

  62. Anon x 2 Says:

    "# 53,58:

    http://www.basketball-reference.com/blog/?p=3555

    It only applies to the NBA's current 30-team, 16-in-the-playoffs structure, but in the simulation we know which team was the most talented, and that team won only 48% of the time. Doug did a variation on this at PFR as well:"

    Neil, I don't agree with the analysis, because although SRS is pretty cool and a pretty good indicator, I don't believe it represents the best or most talented team.

    For example, I believe the Lakers were clearly the best teams the past 2 seasons and Boston prior. In fact, this century, I'd argue only the Heat and Pistons won without being the best team (with the Pistons owing it to a Karl Malone injury somewhat and then as a result being the best team). Some of those years I'd argue Spurs/Lakers were so close to each other it was 50/50.

    Again, I'm not trying to indict SRS. I like it quite a bit and think it's pretty accurate...to an extent. But it can't account for injuries, mid-season trades, and other factors that play into while a team does poorer that their true quality of play (ala Boston taking it easy this year, as well as LA post all-star break).

    Now, I know you'll come back at me with the random allocation in that wonderful post. My argument would be it's based on not knowing who the best team truly is to begin with, so the resulting percentages won't be right.

    Unfortunately, I'm basing this on my own subjective observances and not an actual model. I get that's a problem, no need to remind me. I just don't think SRS can tell us who the best/most talented team is, only who played the best during the season. I think these are different.

    Did anyone really believe Cleveland was better than Los Angeles last season or this past one? I thought this year was closer and possibly arguable, but not the '08-09 season. LA was clearly better.

  63. Dave Says:

    Neil, so we are comparing apples with apples (ish), the RS numbers for ONLY the teams in the playoffs is:

    Pace 92.0 89.7 92.7
    Offence Defence Playoffs RS
    eFG% 0.508 0.488 0.496 0.501
    TOV% 0.131 0.134 0.128 0.133
    ORB% 0.263 0.254(1-DRB%)0.263 0.263
    FT/FGA 0.237 0.225 0.262 0.228

    Whilst this breakdown still supports your conclusions: Pace goes down, FT/FGA goes up, TO % goes down - but none as much as first indicated. ORB% is still constant, and now we get good insight into eFG% (playoffs seems to be an average of offensive and defensive abilities of teams present...) and the teams in the playoffs had a strength at getting more FT/FGA, whilst not much better (than league average) at preventing this.
    Interestingly, it seems ORB% is controlled by the offensive team, not by defensive team...

  64. Ray Says:

    Are there any numbers that indicate how predictive the results (W-L, point differential, etc.) of a regular season series are for the result of a postseason series? Because as we all know, the Cavs needed a squeaker at home (where they went 39-1 when LeBron was playing) to take 1 of 3 against the Magic in '08-'09, getting manhandled in the other two games. The matchup nightmares that were the cause of Cleveland's woes against that team were not resolved come playoff time, so the "upset" wasn't a shock to many who put more emphasis on factors other than W-L.

    This year, the Celtics were a radically different team in the postseason, as I know Neil has already written about before (weren't they second most improved historically?), and while the season series wound up tied at 2-2, we all remember the last game, which came in April and was almost an exhibition for how the Cs were going to consistently play in the playoffs. I don't think ANY metric could have accurately captured their true odds of getting to the finals and winning it all.

    Hopefully, the history that is written about all this takes into account the above, because the Cavs had one serious matchup problem too many the first year and lacked that extra championship gear the next (with several culprits worse than LeBron) to be realistic title favorites either time. I know hindsight is 20/20, but these are things that were being said.

    Or, to put it another way, Cleveland may have had a 66.5% chance of winning at least one championship these last few years - but it involved avoiding either Orlando in 2009 or Boston in 2010. Ouch.

  65. Neil Paine Says:

    #62. I respect your feelings about SRS, but you should understand that the post I linked wasn't specifically about SRS -- I just chose SRS because it was easy to work with and the scale was already intuitive. I could have just as easily assigned teams random "true winning percentages" with an average of .5 and a standard deviation of .15, which is the NBA's distribution since 1999-2000. I could have even come up with some bizarre metric nobody has ever seen before, as long as I could create win probabilities from it. The point isn't the number, it's the distribution and how it relates back to real-life win probabilities.

    Now, in real life I agree, it's totally impossible to know for sure who the best team is/was. We can't do it. But the beauty of the simulation is that we can know which team is randomly assigned the highest rating, and since that single number describes their entire talent level in the simulation (the only other input is luck, the random number generator that determines wins & losses), we can say with absolute certainty that the team with the best rating in the simulation is in fact the league's best team. Then we track that team all the way to the end of the season, and we find that they win the championship 48% of the time. This is what I mean when I say the best team only wins half the time in the modern setup -- in the simulation, we know who was the best team (remember, we're totally omniscient about team talent levels), so if the talent ratings are set up with a realistic spread, it will produce a good representation of the actual game if only we knew which team was truly the best in real life.

    Are there assumptions in this method? Sure. We're assuming a team's talent level is unchanging from November to June, which may not be true even in the absence of injuries or player movement (although it's impossible to say with any certainty whether or not it is, because apparent observed changes may be due to random variation as well). We're assuming teams don't adapt as a series goes on (as #61 points out, this is probably a faulty assumption). There are many other assumptions that could also be challenged, but none have anything to do with SRS, because the method isn't about SRS. It's about randomly assigning strength ratings to 30 teams according to our best estimate of the NBA's real-life distribution, noting which team has the highest strength rating, and tracking how often that team is the NBA champion when the season is completed. And according to the model, that happens only half the time.

  66. Anon x 2 Says:

    I think what the post above is hinting at is how reliable is SRS in terms of predicting a champion. If the answer is 50%, it leads us to two possible conclusions.

    A. The best team in the NBA doesn't win because their so-called dominance is slight.

    B. The formula is good at putting teams into tiers, but not good at telling us who is the prohibitive favorite and/or best team when the playoffs roll around (and we can discuss why).

    Am I missing anything?

  67. Anon x 2 Says:

    Neil, scratch my last post. You seemed to already address it in part, specifically that SRS wasn't a focal point. My response came before your latest post.

    "so if the talent ratings are set up with a realistic spread, it will produce a good representation of the actual game if only we knew which team was truly the best in real life."

    This is the part I would say I'm unconvinced of. I think, in general, about 2-4 teams are well above all the rest of the league to the point that i don't believe it's possible for another team to win without a fortune of good luck related to things like injuries.

    So it's not SRS at issue (my bad), it's the actual strength ratings. This is why I had issues with stuff like Hollinger's playoff odds (based on his rankings). It told me that Miami, Milwaukee, Charlotte, Chicago, Portland, OKC, Atlanta, and Utah had a combined chance of winning it all at like 25%+, but I personally saw it at 0% (or at least approaching zero). I didn't agree with his strength ratings. I really only thought LA, Boston, orlando, Cleveland could win the title, with Phoenix being a very improbable dark horse. What I'm saying is that I would give all those percentage points from the little guys after the top 4 (or maybe 5 depending on year) and give it to the top dog and make his strength rating a lot higher.

    As I believe I mentioned, this decade I can only recall 2 titles to go to a team I didn't consider the best and one of those were helped by injury and I'm sure Mavs fans have another excuse (lol). Bulls were always the best team when they won and Rockets the first season. At least, from my subjective perspective.

  68. Gil Meriken Says:

    I ran my calculations and found that the Lakers losing to Boston in the 2008 Finals was just random chance, but winning in 2009 and 2010 was due to their strong performance as a team.

  69. Anon x 2 Says:

    Another way to put it is I see a "threshold" rating. Think of a carnival ride line. You must be a certain height to ride! In the NBA, you must be at least X good to have a chance at winning a title. If you're not that good, then you can't really compete. Only teams that cross this threshold may be given a strength rating, everyone else revert to 0 (though I don't know if that last part is accurate because I do believe a strength team can be beaten by a non-one in a single series, just that non-strength teams' probability of winning 4 series is approaching 0.

  70. Anon x 2 Says:

    Gil, by random chance you are surely meaning to say "Bynum's collapsible knee.": :(

  71. huevonkiller Says:

    #68

    The Celtics winning in 2008 was just "random chance"? I'm not certain I follow. The Lakers won in 2010 by a much closer margin as well, could that not also be "random chance"?

  72. themojojedi Says:

    Anon X 2, I think your thoughts on thresholding, minimum "contender" requirements and shrinking the non-contenders championship probabilities toward zero have strong merits.

    Even just looking at SRS only, which doesn't account for any of the nuances of playoff matchups, home court advantage, injuries, mid-season trades, team composition, heart of a champion, lack of killer instinct, insert talking-head cliche here, etc, etc, in the modern/3-point line era there is enough evidence to suggest that further investigation of the empirical threshold for a champion is warranted. Here are the the SRS ranks for the last 31 champions:

    1st: 14 times
    2nd: 3 times
    3rd: 4 times
    4th: 5 times
    5th: 1 time
    6th: 3 times
    11th: 1 time (Hou 95 "Heart of a Champion")

    So a team with an SRS outside the Top 6 has won only once in 31 years (3.22% of the time) and 83.87% of champions have come from the Top 4 SRS teams. So it really makes the combined 25%+ chance at the championship for those lesser teams that you mentioned seem way overstated when taking a frequentist perspective.

  73. nimble Says:

    Krake-Chimera vs Kobe!Noble shall prevail...

  74. themojojedi Says:

    Hadn't looked at Neil's Ten Thousand 2010s when I made my post, which is obviously way more comprehensive and demonstrates the shrinkage of probabilities for those non-contenders much closer to zero than the Hollinger playoff-odds model.

  75. Anon x 2 Says:

    Themojojedi: And that Houston team had a major mid-season trade, to boot. In fact, I bet a handful of those champions below 1st and 2nd in SRS can be explained by a major injury or mid-season trade ('95 Rockets, '01 Lakers, '05 Pistons, '06 Heat, '09 Lakers).

    Perhaps if I knew of a listing.

  76. Jerry Says:

    Your exponent in your WinPyth formula, or whatever you use to forecast games/series, is way too high. 9.5 is best for forecasting when you use point differential or SRS and WinPyth. Never ever did they have a 44% chance of succeeding in '09

  77. AYC Says:

    How about the random chance of KG being hurt in 09, and Perkins getting hurt before game 7 this year. Seems like random chance is the only reason the Celts didn't win 3 in a row

  78. Neil Paine Says:

    # 76. I didn't use the pythagorean formula, I just used regular-season winning percentages, the log5 formula, and a home-court advantage of 60%. If Team A is at home and Team B is on the road, this yields:

    p(Team A win) = ((Team A win%) * (1 - Team B Win%) * .6) / ((Team A win%) * (1 - Team B Win%) * .6 + (1 - Team A win%) * (Team B Win%) * (1 - .6))

  79. Jason J Says:

    #75 - "Anon x 2 Says:
    July 16th, 2010 at 3:12 am

    Themojojedi: And that Houston team had a major mid-season trade, to boot. In fact, I bet a handful of those champions below 1st and 2nd in SRS can be explained by a major injury or mid-season trade ('95 Rockets, '01 Lakers, '05 Pistons, '06 Heat, '09 Lakers)."

    They didn't all have additions by trade, but most had good reasons that they did not excel as much in the reg season as the playoffs.

    '95 Rockets - Picked up Drexler late, and he gelled just as the post-season got underway.
    '01 Lakers - Shaq missed a lot of time in the reg season.
    '05 Pistons - Don't recall anything odd during the season... team boredom?
    '06 Heat - Lots of minor injuries throughout the reg season esp. Posey, Zo & Shaq.
    '09 Lakers - Bynum missed a lot of in the reg season (though they had a great record).

  80. Daniel Song Says:

    I've done a bit of reverse engineering and found that the betting favorite to win the NBA Championship 14 out of the last 20 seasons. Not bad.

    Quite often, the betting favorite didn't have the best regular season numbers, but stepped it up come playoff time.

    Betting markets usually provide the best information in terms of predicting winners. Handicapping and number crunching can be helpful, but any edges against the public are incremental at best.

  81. Neil Paine Says:

    Where did you find the data on past betting favorites?

  82. Jonathan Says:

    This seems like one of those situations where statistics alone are sorely lacking.

    1) Certain teams are far easier than others to prepare for in the playoffs. The Cavs had no players other than Lebron who could be counted on offensively, several very weak points defensively, and a very predictable offense. They are a perfect example of a team that's easier to beat in a 7-game series than on any given night, because you can focus your defense on Lebron and your offence on their defensive weaknesses. Those deficiencies are masked during the faster-paced regular season.

    2) The playoffs are about matchups, not overall team strength. In the 2008 Celtic series, 2009 Orlando series, and 2010 Celtic series, the team the Cavs lost to had 3 of the 4 biggest matchup advantages on the court.

  83. Anon x 2 Says:

    Themojojedi:
    "They didn't all have additions by trade, but most had good reasons that they did not excel as much in the reg season as the playoffs.

    '95 Rockets - Picked up Drexler late, and he gelled just as the post-season got underway.
    '01 Lakers - Shaq missed a lot of time in the reg season.
    '05 Pistons - Don't recall anything odd during the season... team boredom?
    '06 Heat - Lots of minor injuries throughout the reg season esp. Posey, Zo & Shaq.
    '09 Lakers - Bynum missed a lot of in the reg season (though they had a great record)."

    '01 Laker also missed Kobe with significant time (when he came back is when that massive run started)

    '05 should have been '04 Pistons. My bad. Sheed trade mis-season.
    '09 Lakers should have been '10 Lakers. Kobe, Drew, Pau, missed a lot of time.

  84. Daniel Song Says:

    Neil,

    I've dug up much of the information from goldsheet.com, where they have a point spread log from ~1995 to 2009.

    I've had to reverse-engineer much of the results by looking at point spread logs from the past. I was really active only from about 2000 to 2006.

    For years 1991-1995, I've retro-examined the factors that go into creating the odds and guesstimated the betting favorite. I believe Portland would have been a slight favorite over Chicago entering the '91 playoffs (though I'm not certain) and the Bulls were obvious favorites in '92 and '93. Houston were the favorites in '94, while '95 was wide open (but definitely not Houston).

    List of favorites since 1996: (At the start of the playoffs)
    1996 Chicago
    1997 Chicago
    1998 Chicago
    1999 San Antonio
    2000 LA Lakers
    2001 LA Lakers
    2002 LA Lakers
    2003 San Antonio
    2004 LA Lakers
    2005 San Antonio
    2006 Detroit
    2007 Dallas
    2008 Boston
    2009 LA Lakers
    2010 Cleveland

  85. Daniel Song Says:

    We can continue this conversation via e-mail if you wish. I'll send you an e-mail straight away.

  86. Daniel Song Says:

    Let's see if Dan Gilbert can deliver on his promise to deliver a title to Cleveland before Lebron wins one.

    It's not as impossible as it sounds. It's been proven over the course of NBA history that the best way to build a team is through (1) draft, (2) trades, and (3) free agency. It's always been in that order and I don't anticipate it changing any time soon. The Lakers' signing of Shaq has been the only time that a big-ticket free agent delivered championships - but even then, they had to draft Kobe and shape the rest of the roster through trades and incremental signings.

    The cheapest way to obtain good players is to draft well. And if Gilbert can find a way to shed bad contracts and get out of salary cap jail, they will be in a position to (1) pay their draft picks who pan out, and (2) sign role players at reasonable prices. They won't have to pay the luxury tax and they'll have more trade possibilities as well. If done well, Cleveland can potentially become a good team within 3-4 years. Or even sooner, if they get lucky with another superstar through the draft.

    As for Miami, I predict they have about a three-year window. If they don't win a championship within that time, the situation will implode. You will see either Bosh or Lebron opt out and/or traded, and it will be a full-scale meltdown from there. I can easily envision Miami being stuck with a 31-year-old, broken-down Dwyane Wade, a bunch of bad contracts, and Bosh and Lebron leaving for greener pastures. We shall see...

  87. Neil Paine Says:

    Some would say they have a 1-year window, since the lockout could wipe out all of 2011-12 and a resulting hard cap might force teams to release high-paid players. Miami may very well end up divesting themselves of Bosh by the summer of 2012, retaining only James & Wade, if that.

  88. Jason J Says:

    Daniel - That's an interesting point about the most successful means of acquiring top level talent to build contenders. It should be noted though that the scenario of LeBron joining Miami may actually parallel Shaq joining LA moreso than it does the average free agent signing. He certainly is the most talented free agent to change teams since Shaq.

  89. Anon x 2 Says:

    "Some would say they have a 1-year window, since the lockout could wipe out all of 2011-12 and a resulting hard cap might force teams to release high-paid players. Miami may very well end up divesting themselves of Bosh by the summer of 2012, retaining only James & Wade, if that."

    I definitely don't see a hard come coming without grandfathering current teams. It's not just Miami that would be screwed, by L.A., Dallas, and other teams. Punishing good teams seems awkward.

    Lockout very real, but a hard cap instituted in 2012 seems doubtful to me.

    Jason, I disagree because you also have Bosh there with Mike Miller. Jerry West built the majority of that team through draft and trades. Shaq was the only real pickup, as the team was built with Nick the Quick, Eddie Jones, Elden Campbell, Fisher, and Kobe. Players eventually were moved, but the team was mostly home-grown on Shaq's arrival.

    Same with the Spurs, Pistons, Celtics, and recent Lakers. Ron-Ron and VladRad were the only major FA (with the space cadet traded for Shannon Brown). Farmar, Kobe, Bynum, Sasha and Fisher all drafted (though Fisher returned as a FA 3 years ago via unusual circumstance), Pau and Lamar were traded for.

    Boston's only impact FAs were PJ brown and Posey, rest through trades and draft. Spurs core was drafted and the rest again mostly trades and drafts, with Finley being the only major FA that comes to mind. Bulls and Rockets are also the same for the most part. Pistons in '04 as well.

    Miami is truly trying this with a method we've never seen before. Essentially 1 to 2 (depending on Chalmers) players home grown and the rest through free agency (sorry, the Lebron/Bosh trades do not count).

  90. Anon x 2 Says:

    hard come = hard cap. sheesh. lol

  91. Jason J Says:

    Anon X - i didn't mean to say it was an absolute parallel or anything. i was just mentioning that this particular high profile FA signing is for a talent capable of bringing in titles, like shaq, where most players capable of bringing titles - garnett, duncan, kobe, etc. have chosen to stay with their current team whether for comfort, loyalty, the extra money, or continued success. it's just been almost unheard of for a perennial MVP candidate to change teams as a FA in his prime, and the last time it happened, as far as i can recall, was with Shaq in the summer of 2006, and it led to titles (eventually).

  92. Jerry Says:

    In response to #78. I highly doubt that winning % and log5 outperform SRS/PD and WinPyth but I do not have prove of this (yet). But since it gave Cleveland a 44% of winning it all, I would say there's a pretty good chance that it always underrates the underdogs

  93. Jason J Says:

    #91 - I should have said "summer of 1996" not 2006.

  94. Neil Paine Says:

    #92 - Actually, the goal of this method isn't to have the best predictive accuracy... Allow me to quote myself from #45:

    "...the point here isn't to predict (in this case, retrodict) who was going to win the championship. It is to create a naive set of expectations that serve as a proxy for the conventional wisdom on how often Cleveland 'should' have won. In other words, it's putting a naive probability on statements like 'Cleveland won 60+ games twice but they didn't win a title, what a choke!'"

  95. Jerry Says:

    If you start out like this, with those "naive set of expectations", then I feel it's definitely wrong to write things like

    "However, calculating the odds of them being shut out twice in two years requires some additional math: since we know the probability of the Cavs not winning in 2010 was (1 - 0.368) = 0.632, and the probability of them not winning in 2009 was (1 - 0.447) = 0.553, the odds of them being ringless in consecutive seasons was (0.632 * 0.553) = 35%."

    "If you expanded things to look at the entire LeBron era, you'd find that Cleveland won 0.8579 titles on average during the 5 years the Cavs made the postseason"

    .. without putting your statement from #94 on top of the article or at least starting what I quoted here with "going by this method" or similar.

    You start with expectations that are definitely wrong (you knew this), then reach (wrong) conclusions but actually phrase it like they're real facts

  96. James Says:

    I just don't understand why people have such a hardtime understanding that a team must have at least two legit star level players and a strong team defense in order to win an NBA championship. LeBron James has never really had so much as one and the only time he had a player produce like a somewhat viable second star in the playoffs (Daniel "Boobie" Gibson) he took an extremely flawed Cavs team to the NBA Finals.

    That the Cavs just continued to pick up aging big names that other teams had deamed expendable gave me the same impression that it likely gave to LeBron James. The Cavs were interested in winning enough to spend money to take on contracts many other teams were no longer willing to pay but they weren't concerned enough to go out and make sure they got a second real impact player in their prime or in any case had shown no signs that they were capable of making such a move.

    LeBron James can take a team with no glaring holes and a solid supporting cast to the leagues best record because he has proven to be the single most dominant force in the sport these past few years. If the 09 playoffs proved anything its that no matter how dominant a single player is that they still can't take on the leagues best teams in a best of seven setting playing 1 vs 5.

    I've got no issue with LeBron James feeling that he couldn't trust the Cavs to get him the help that every great player needs to win titles. Just as i've got no issue with Bosh coming to the conclusion that Colangelo quite possibly never intended to build a proper team around Bosh and that he was just waiting for Bosh to leave so he could build around his guy (Andrea Bargnani) to cover himself for taking the guy #1 overall in 06.

    I fully understand why both Bosh and LeBron felt compelled to put themselves in the best position to win as both have been saddled by poorly run franchises in my opinion. I also fully understand why Dwyane Wade took less money and trusted Pat Riley and because he has seen both sides of being a champion vs going it alone its easy to see why being rich and winning trumps being more rich and not winning.

    I think its fairly easy to see that a guy like say Amar'e Stoudemire who has played around a bunch of talent and likely takes it for granted was the most concerned with getting his money and the brightest personal spotlight by going to the Knicks. I'd expect in the not so distant future he'll be wondering exactly why being the man vs being a contender was so important to him.

    I just think in the spotlight of the summer free agent period that the people who are talking about how LeBron James just gave away his chance at a legacy on Jordan's level by taking the easy road will be singing a different tune should this squad get into Jordan's teams level of greatness.

    He gave up a legacy of potentially being the best player of all time to never win a title for a very good chance to be one of those guys with 5X titles and multiple MVP awards.

    Perhaps when people get a bit of context like say the Cavs losing 50+ games next year that they'll remember that without LeBron James, having a Mo Williams and Antawn Jamison as two of your best players is a surefire trip to the lottery and finishing in the bottom third of the league in team defense.

  97. Anon x 2 Says:

    "I just don't understand why people have such a hardtime understanding that a team must have at least two legit star level players and a strong team defense in order to win an NBA championship."

    Who was Houston's 2nd star in 1994? Who was Duncan's 2nd star in 2003 (and don't even mention D-Rob who played under 24 mpg in the playoffs).

    Do you consider anyone on the '05 Pistons stars? Which 2?

  98. Anon x 2 Says:

    correction: '04 Pistons. Done that twice now!

  99. Daniel Song Says:

    Anon X, some interesting questions, but I'll give it my best shot.

    1994: Olajuwon was Houston's only established star. However, all the role players stepped up and played extremely well in the playoffs. Guys like Maxwell, Horry, Thorpe, Smith, and Cassell may not be household names (OK, maybe except for Cassell, but he was a rookie getting limited minutes back then), but they performed when it counted. They also had an extremely good defense and you can make the argument that they had three stars on defense.

    2003: San Antonio had a deep and experienced team. Parker was starting to emerge, while Bruce Bowen had emerged as a defensive force. They were truly exceptional on defense and they were actually the favorites to win that season!

    2004: Pistons had several excellent players on their roster. I still can't believe Billups doesn't get his due. He has twice the accomplishments of Steve Nash yet gets half the credit. He's been one of the elite point guards in the league for close to a decade now. Ben Wallace was a defensive monster back then. Rasheed Wallace was still useful and Hamilton and Prince were also heavy contributors. They almost went back-to-back in 2005 but couldn't quite pull it off.

    I think the key to winning championships is to have a good offense AND defense. People focus so much on the offense, yet it's only 50% of the game. I'm not about to hand the Miami Heat the 2011 title until they show they can actually play defense once in a while. You can have Bosh, Lebron, and Wade and add in Amare Stoudemire and Steve Nash while you're at it. If you play zero defense you'll still lose.

  100. Anon x 2 Says:

    ^^^ I understand all that, but i was looking at the claim that you MUST have 2 stars. This is not true. It's just very hard without them.

    As for both offense and defense, The Detroit Pistons won with the 18th rated offense. Lakers this year had a mediocre offense (though it was biased down a bit by all their injuries).

    I believe something like no one outside the top 10 in defense has ever won the NBA title. Miami was 9th, but that was biased due to Shaq missing so much time during the season and in the playoffs they were around top 3, IIRC.

    What Detroit did (awesome D, very meh O) was an outlier, but championship teams are almost always in the top 5 defensively. Lakers this year were 6, but were #1 heading into around March before Kobe and Drew went down with injuries.

    I think defense is more important than offense, but you can't have the Bobcats offense at the same time!

  101. Daniel Song Says:

    In my experience the most important factors to winning a championship are (1) officiating, (2) talent, (3) experience, and (4) coaching. Regular season numbers aren't quite as important. I've seen too many championship-caliber teams dog it in the regular season, only to turn it on come playoff time. It's the Randy Moss effect - they play when they feel like it.

    This is where a lot of numerical analysis goes wrong. Talent doesn't always show up in the regular season numbers. Experience and coaching are hard to quantify. And you can have all the factors in your favor and outplay the opposition - but a few whistles in some key spots will completely change the game. Miami in '06 is a perfect example. There are many others.

    Officiating has been one of the points that has driven handicappers nuts for years. Even a couple of whistles changes winners into losers. For an extreme example, Dwayne Wade shot close to 20 free throws per game in '06 - which prompted me to look at all the games in '06 when Kobe shot 20 or more free throws. In those games, Lakers were undefeated and outscored their opposition by more than 15 a game, while Kobe averaged over 50 points per game. Refereeing makes a huge difference!

  102. Daniel Song Says:

    P.S. About Wade, I'm referencing the '06 NBA Finals.

  103. themojojedi Says:

    101. Really? You believe that the number 1 factor in winning a championship is officiating? That will come as quite a blow to the legacy of Bill Russell ;)

    The 2006 Finals is an extreme example, but surely it is the exception rather than the rule. Even with the outrageous number of calls for Wade, the Heat only won by 2 in Game 3 (18FTA), by 1 in Game 5 (25FTA) and by 3 in Game 6 (21FTA), with HCA in the first 2 of those games. Let's say you put the exaggerated free throw numbers solely down to officiating. That indicates to me that even in this extreme case it takes a significant amount of referee interference to barely overcome the basketball-related factors (e.g. Dallas with a regular season efficiency differential larger by 2.6 as one measure of this). That seems to indicate a fairly low sensitivity to officiating rather than high.

    There's also the issue of separately identifying the components causing the FTA. Part of it is officiating, while other aspects being captured are the quality and aggressiveness of the player getting the calls. Apart from that, the FTA numbers here might just be an indicator that flaws in the Mavericks defense are being systematically attacked and exploited, which is typical in a 7 game series. These 2005-06 Mavs did get outscored 62-61 through 3 quarters by Kobe, another attacking perimeter player, who averaged 43 ppg and 16.7 FTA/g in their regular season series.
    Similarly, LeBron averaged 41ppg and 10.5 FTA/g in two regular season meetings. Exploiting the weaknesses of opponents speaks more to talent, experience and coaching than officiating.

  104. Daniel Song Says:

    Themojojedi,

    I've been a Lakers fan and I will readily admit that officiating was perhaps the biggest factor in several of their championships. Lakers got the calls down the stretch in Game 7 vs. Portland in 2000. Sixers were definitely robbed in Game 3 in the 2001 Finals. There's absolutely no way the Lakers should have beaten the Kings in 2002. In 2009, the Lakers were too dominant for the referees to make a difference but the Lakers lived at the free throw line down the stretch vs. Boston in Game 7.

    Officiating was one of the factors that just killed my earlier attempts at analyzing playoff teams. After being taken to the shed a few times, I finally decided to incorporate them into my power ratings and started adjusting them based over the course of the playoffs, based on the calls they were (or weren't) getting. My results started to improve once I got better at predicting potential refereeing bias.

    As for the 2006 NBA Finals, the fouls often came in key spots that stopped Dallas' runs or fueled Miami's. Wade was fouled consistently, but there were many, MANY instances in which he just jumped into the player, threw up an airball, fell down and got two free throws. Plus, it seemed that Dirk Nowitzki was being fouled every time down as well, yet he did not get near the number of free throws. It was really an infamous Final and the general consensus was that Dallas was getting hosed - not that Dallas were a bunch of thugs who were justly punished.

    I'm not sure how much Russell and the Celtics were affected by the officiating. What I do know is that the "stars" system and the officiating bias targeted toward certain teams has been around at least since the early 80's. People have been joking about this for 30 years now and virtually no one was surprised when it was revealed that an NBA referee was found to have fixed games. But even if you don't buy the anecdotal evidence, what I know for certain is that my prediction accuracy increased once I started to incorporate possible refereeing bias in my handicapping.

  105. themojojedi Says:

    Daniel Song,

    I agree that officiating can impact games and series, particularly those that are more evenly matched, and it makes sense that you could increase your prediction accuracy when incorporating referee-related variables. I also agree that the timing of crucial calls can have a huge impact, but I'm sure that you could find 4 or 5 crucial basketball plays for every seemingly unethical (and not just incompetent) call. But I don't think conceding these points establishes officiating as the most important factor in winning a championship. Do your officiating variables add MORE predictive power than the basketball-related variables?

    Given that you are putting your money on the line based on your predictions I'm guessing you have a pretty sophisticated model. If you don't mind me asking, what seem to be some of the stronger officiating "biases" that you've found? Big city bias, star treatment bias, referee allocation bias, individual referee bias, home court officiating bias, bias toward certain styles of play?

  106. Daniel Song Says:

    themojojedi,

    There are several types of officiating bias and each one of them must be considered.

    The most well-known factor is widely known throughout the betting circles - the infamous "zig-zag" effect. It's a simple system in which you simply play the straight-up loser of the previous game of the series. It's so powerful that the Las Vegas line eventually moved 1-2 points to account for this effect. People cite "motivation" as the reason for this effect, but I think officiating has just as powerful an effect. This system is especially powerful when an underdog plays its first home game down 0-2 in the series.

    Just two cite two examples in the playoffs, Lakers were up 2-0 vs. both Oklahoma City and Phoenix. In game 3, Lakers were beaten both times as their opponents racked up considerable free throw advantages. This was especially curious since Lakers are much more geared towards an inside game than either of their opponents. It was interesting to see how the games transpired. Lakers took the ball to the basket, got mugged, and was left searching for in vain for whistles. After a while they started to jack up contested threes since that seemed to be the only way they could score. Obviously it was a losing proposition and they went down in defeat.

    There was a somewhat related example in both of the conference Finals. The Lakers and Celtics both seemed to be headed toward sweeps. Not surprisingly the Suns and Magic received the benefit of the calls in the next couple of games as the series tightened up. When the officiating became more even, the Lakers and Celtics were able to close out the series.

    Yet another example is the frequency of fouls that are called. Some games are foulfests while others are wrestling matches. Obviously, the foulfests will result in higher-scoring games than wrestling matches for the most part. And foulfests tend to favor finesse teams while wrestling matches tend to favor physical teams. In addition, sometimes star players will get into foul trouble, testing a team's depth. This type of bias will generally influence the total more than the spread, though you can seriously change a game by putting a team's best player in severe foul trouble.

    The most infamous, and hated, example of officiating bias is the "star" treatment - both toward established stars and their teams. Even Magic Johnson and Larry Bird were reviled back in the day for being able to run into the lane, throw up a wild shot while falling to the floor, and get two free throws. Among today's players, Lebron James is an obvious example - he often runs right to the basket and bowls over a defender in the lane, and gets two free throws. And he never seems to be whistled for a foul no matter what he does. Dwayne Wade is an even more egregious example as he just jumps into the defender on the 15-20 foot jump shots and still gets two free throws. Kobe, not as much - though he got several gifts down the stretch in Game 7 vs. the Celtics. The opposite of that is Dirk Nowitzki, who gets mugged on every play but almost never gets to shoot free throws.

    Teams also get calls. 2006 Miami was a great example. The 2007 Spurs-Suns series, another one (though Suns received a Christmas present in Game 4). Celtics got the calls against the Lakers in Games 1 and 2 and it left the Lakers in too big a hole to dig out of. I can go back to the previous Lakers three-peat, the Bulls' two three-peats, and the Lakers and Celtics in the 80's. It's the same old story.

    This affects the future betting in a significant way. All but 2-3 teams become unbettable since they (1) don't have the star player who will rack up bogus free throws, or (2) don't have the "pedigree" that will allow them to get the calls in a closely-contested series. Yes, these factors can be overcome, but not often enough to justify betting on these teams. This is why the "favorites" are usually the best ways to bet the futures. The vast majority of my wins were from backing the top team(s) at short odds or shorting good regular season teams that didn't have the above two factors.

    I still find NBA entertaining and feel it has the same appeal as watching pro wrestling or your favorite TV show. It's somewhat scripted but you still don't know what's going to happen, and it always manages to produce its share of excitement. Let's see what happens in 2011!

    P.S. Unless you are the LA Lakers, Miami Heat, Boston Celtics or the Orlando Magic, you have zero chance of winning the championship. Don't say I didn't warn you.

  107. Jason J Says:

    Daniel - I think you exaggerate the gifts given to Bird and Magic. Neither of them averaged very many foul shots per game in a much more grabby era of defense.

    Also Dirk gets to the line a ton for a guy who is primarily a high post jump shooter. With the exception of 2010, his free throws have been right in line with Michael Jordan's when the Bulls switched to the triangle offense and Michael began most possessions in the same spots that Dirk does. Maybe that's the wrong comparison due to the era difference in what constitutes a foul (frankly I think Bryant and Pierce both get a lot more benefit of the doubt whistles than MJ or Dirk in those jumpshooting / high post situations - but they also sell it better).

  108. Daniel Song Says:

    You could be right, but my goal was to find a way to cash more tickets. I'd rather have a biased and flawed analysis lead to the right answer than to have a completely numerical analysis lead to the wrong answer. I've seen several attempts to predict NBA outcomes simply by using regular season statistics. All of them failed miserably, including my own.

    However you do make a good point in that Pierce is yet another one of those floppers who get bogus free throws. I think Kevin Durant is yet another player that has shown potential in that arena. Unfortunately we've gotten to the point where flopping is the most important skill to have on offense and I would love to see that change. If I wanted to watch flopping all day I can always watch soccer. I would like to see something different in basketball.

  109. Jason J Says:

    No argument regarding the annoying overabundance of flopping in today's game. Fisher was driving me insane in this last finals, but like JVG said, he's been doing it to great effect since he came into the league, no reason to expect him to stop now.

  110. themojojedi Says:

    Thanks for the info Daniel, interesting to get a betting perspective on officiating.

  111. Ryan. Says:

    Daniel, why is it that you offer rhetoric as opposed to method? This seems - and forgive me for being judgmental here - like one of the all-too-common charades constantly haunting the hallway of stat analysis. I guess what I'm saying is, until you provide true basis and not just poetics, I'll call BS.

  112. Gil Meriken Says:

    111.

    Ask yourself, what is "true basis"?

  113. Sean Says:

    Hard to say. An equal question might be: 'just how many wins during the regular season should teams like the 2009 & 2010 Cavs have had?'

    The post season in some sports seems to be a different animal than the regular season and relying on regular season success alone to predict post season success can be foolhardy.

    Sports where you play 'best of' series tend to be like this. The preparation is different. Playing the same opponent 7 games in a row or even 4 games in a row, home and away is different than different day/ different city. You are dissecting an opponent and countering counter-moves from the same foes over a week/ 2 week span.

    With hockey it gets whacky with the 'hot goalie' effect, baseball has it's 'how's your rotation set up' effect and nobody forgets your moves night after night in basketball if you keep playing the same guy(s)...

    LeBron can be figured out. A strong team will take away what you do best, then it's your move. LeBron maybe doesn't have enough ways to beat you in a long series if you're a good team. He does during the regular season, perhaps because he's flying into your city, playing one game, then leaving. You can't make adjustments. The travel is different in the playoffs... and older team like the Celtics may have benefited from staying in a city for extra days during a series. There's a lot that makes the post season different.

  114. King Kong Says:

    There are lots of great players that haven't won a ring, or only won one ring. The problem is that LeBron was (and is) endlessly hyped by everyone, and his continual failure is constantly excused.

    I do think he and Kobe are the two best players in the league, but wait until he proves himself to suck him off or you're going to end up with even more excuses

  115. Eric Says:

    Looking at the results compared to the probabilities is troublesome for this method. The favorites to win the championship only did so 1 out of 5 - the 2008 Celtics. The favorites from each conference only reached the Finals 4 out of 10 chances. The actual champion had probabilities of 3%, 6.1%, 63.5%, 34.2%, and 13.2% against favorite probabilities of 49.2%, 64.2%, 63.5%, 44.7%, and 36.8%.

  116. Neil Paine Says:

    Again, these are not intended to represent the teams' "true" probabilities of winning the championship; it's what we would expect their probabilities to be based on regular-season winning % alone.

  117. cheap Sennheiser Says:

    I think certification is good in some cases, but often not worth it.