The Beginning of Wisdom (Washington Wizards Take 2)

“The only real wisdom is knowing you know nothing” -Socrates

One of the fun things about having a blog is that I get to make strong statements about the opinions I form through analysis. A question or an idea comes into my mind and I go off to take a look at the data and see if there’s something there. Sometimes like Athena from the mind of Zeus the idea comes fully formed and doesn’t require much beyond making some tables and pulling some quips together (like the Half Baked Notion). Sometimes the process is not so easy. My last post belongs to the first category.

While looking at the data for all the teams in the NBA for the various post I’ve written on this site (like the draft analysis or the NBA playoff team review series) an idea struck me and stuck with me. The Washington Wizards in 2011 project to have a historically bad frontcourt. Given my feelings on the short supply of tall people this led me to the idea that the Wizards could have a historically bad season in 2011. To summarize from that post:

Big Men (F/C) are on average more productive than everyone else . They in fact account for 50% of all productivity.  This makes it harder for a center to be better than the average and thus accumulate wins in our model but this is not out of step with the reality of the situation.  Teams also have a lot more at risk with their big men. It’s also much easier for a team to accumulate negative value at center and power forward because there is much more at risk. So a bad point guard performance will hurt your team, but a bad center will ruin your team. This is why you need at least an average big man to be successful in the NBA. Much like with your typical fantasy football team and the running back position, you need at least to break even at that center or you’re in for a long season.

In table form it looks like this:

So as I typically do, I took off to the excel cave, looked at all the data and wrote a midnight post on the subject (Could the Wizards in 2011 be the Worst NBA Team of all time?). I then went to bed happy with my conclusions and figuring I could proof the article in the morning (after all not a lot of traffic overnight on my blog 🙂 )

So imagine my surprise when I woke up on sunday and I had a traffic spike. Imagine the scene:

While having some cream of wheat and listening to the football pre-game show, I see a web-address I don’t recognize redirecting traffic to my blog. So I go there. After I stopped choking on my breakfast,  I realize I’m reading the blog of the owner of the Wizards (Ted Leonsis).

“Wow”  I think while cleaning my cream of wheat from my laptop monitor, “People in the NBA are reading my stuff”

Now, in my own life I get paid to tell hard truths. I’m not a yes man and this has gotten me in trouble more than once. I’m the guy you call when your business or manufacturing site is in a bad way to tell you how bad it is and to turn it around quickly. It’s a seller’s market. I go in, I evaluate, I come up with a plan and I tell you what’s what no sugarcoating. It’s up to the client to decide what they want to do with the information. But as with anything, when bearing bad news there’s always the possibility of blowback. In those cases, keeping in mind the wise words of Socrates I’m willing to admit the possibility of being wrong , go off and rerun/expand my analysis to confirm or refute  my findings . This is such a case.

The first thing I looked at is who the Wizards were in 2011:

The Wizards had 24 players on the court in the 2009-2010 season who produced 29.16 Wins (and won 26 games).  Only six of these players remain on the roster and they produced 2.85 wins in 2009-2010 season. So they lost 26.31 wins to players departures, who did they replace them with? The following table show the veteran players on the Wizards roster and their production in the 2010 season:

The veteran players on the wizards roster produced 4.33 wins in the 2009-2010 season in 13280 minutes played (or about 67% of the 19680 minutes required for an 82 game season with no overtime). Prorated to a full season this is about 6.41 wins.  This does not look promising for the Wizards so far. Their fans have argued that:

  • 2010 is too limited of a sample size given some of the player injuries and ages
  • Arenas will be much better in 2010
  • The rookies  will carry the day
Let look at each one:
The numbers do not change dramatically for this group of players over the last three years. On average they produced 3.6 wins a season. If I look a little deeper and include my minute projections for next year:
Note that if every player on the roster matches their best year of the last three they’ll produce 16.4 wins, if they perform at their average rate they’ll produce 8.6 wins and at their worst they’ll produce -4.6 wins. The first and last scenario are not terribly likely (think spontaneous combustion or ball lightning).  So 8.6 wins is a reasonable prediction here.The caveat here is that Josh Howard had ankle and wrist surgery in 2009, ACL surgery in 2010 and is 30 and the bulk of his minutes could go to the volcanic black hole of basketball that are Yi Jianlian and Al Thorton. If these two get another 1000 minutes from Howard and Howard plays like 2010? the number of wins goes to around 2 wins produced by this group. So still, not a lot of cause for optimism there.
Let’s look at Arenas now. His numbers for the past five years look like:

  • 2006: 3384 MP     Age 24 WP48 .158 ADJP48 .295 Wins 11.14
  • 2007: 2942 MP     Age 25 WP48 .170 ADJP48 .313 Wins 10.42
  • 2008: 425 MP        Age 26 WP48 .066 ADJP48 .232 Wins .58
  • 2009: 63 MP           Age 27 WP48 .198 ADJP48 .366 Wins .26
  • 2010: 1169 MP       Age 28 WP48 .078 ADJP48 .237 Wins 1.9
The argument here is that he’s back and healthy and has something to prove and the 2011 Wizards will get the 2007 Arenas and not the 2010 version. Now as a fellow cuban-american, I have a soft spot in my heart for Arenas and would love this to be true but reality is a rough mistress. Gilbert will be a 29 years old coming off multiple knee surgeries  in the 2010-2011 season. At best, I think Gilbert might be good for 2000 minutes, a .120 WP48 and about 3.2 extra wins. But I’m already assuming Howard’s spontaneous rejuvenation in my analysis and while one might happen, to is unlikely (unless Marty McFly and a Delorean are prominently involved). The initial analysis  holds up so far but surely the rookies will get it done.
Rookies are notoriously hard to predict but luckily I have some data lying around.

So the majority of rookies fall between -.100 WP48 and .200. For PG, only 3 20 or under players have been better than .150 WP48 as rookies (Magic, Chris Paul and Rajon Rondo) and based on the previous analysis of  PAWS40 and the fact that he’s not 6’9″  I feel comfortable saying Wall will not be as good as a rookie as any of those three players. So a handy dandy chart for the performance of the wizards rookies looks as follows:
So if Wall has an exceptional season and both rookie PFs are good the wizards might see 8.6 to 12.9 wins from their rookies.  Typical performance only gives them 4.3 additional wins.
So if we put it all the scenarios together:
So if absolutely everything breaks the Wizard way they’ll win more than 20 games and if everything goes poorly they won’t win a one. But these are extremely unlikely scenarios.  Based on everything I’ve seen here  I’d forecast this team from between 9 and 21 victories. So at the end of the day my initial analysis holds, the Washington Wizards in 2011 as currently built should expect to get worse and not better and this  may make them one of the worse teams records wise in NBA history (and with some bad luck quite possible the worst). As always, time will prove whether I am right or I am wrong.
Arturo Galletti Written by:


  1. Zach
    September 14

    Your analysis is quickly surpassing Nate Silver’s as my favorite daily read!

    It looks to me like their biggest hope of avoiding a disastrously bad season is a combination of either Arenas or Howard returning to 4 years ago plus JaVale McGee taking a big step forward in production.

  2. September 14

    Thanks! Nate Silver is a rockstar, I’m blushing like a schoolgirl.
    The big problem for Arenas and Howard is age and injury history. Howard in particular is a worry. In two years, He’s had left ankle and left knee surgery. If he’s out or hobbled then the worst record ever is absolutely in play.
    McGee hasn’t shown it in two years and he’s not getting the ball with this roster. I think he could develop into a decent bench player but that’s kinda it for him.

  3. September 14

    Wow! Talk about thorough! Two posts from you, and then one from Berri. Wizards fans should consider themselves lucky.

    • Chicago Tim
      September 14

      I think the three posts prove that Wizards fans should *not* be feeling lucky, unless they like thorough statistical analysis of their misery. I guess the one ray of hope is that the owner is on notice, and if they do win 22 games or less this year, perhaps he will make some changes.

      • September 14


        Yeah and we were working independently and came to the same conclusion. If this team wins anything north of 25 games we are probably witnessing Wall morphing into Paul 2.0. This is not a likely scenario.

  4. Alvy
    September 14

    Arturo, the worst team in Nba history managed to win 11 games right (Dallas Mavs)? I figure that a range between 11 to 19 wins is in store for these Wizards

    • September 14

      Nuggs in 98, Dallas in 93 won 11. I’ll be very curious to see the over/under on this team.

  5. Tom Mandel
    September 16

    As always, Arturo, your analysis is imaginative and compelling. And, as always (or *almost* always in any case), the Wizards aren’t going to be very good this year! But, we knew that. This ain’t like the truth-telling you do in your job, AG. Anyone who ate breakfast this morning knew this. Hey, use those gifts to prove something *counterintuitive*, why dontcha! 🙂

    Still, about the Wizards we also can see some predictors for future success:

    1. A new owner; without doubt he is a gifted businessman who has shown he can upgrade a professional sports team. There’s a payoff for success.

    2. We got the #1 pick in the draft. What do stats show about the performance of teams in years n+1, 2, 3 and n+4 who get the #1 pick in the draft in year n?

    3. We’ll have another high draft pick next year — maybe the #1; you give us a good shot at it!

    4. A high % of rookies and other young players — what % of players get better between 19-20? Between 20-21? Between 21-22? Gosh… there’s your table. What is there about “rookie” that influences that data? Is it congruent w/ data for year 2 players who are 19, 20, 21, etc.? Year 3 players? Can’t I use your table to predict some serious improvement for JaVale McGee. How about Yi (who is claimed to be 22)? Or even Andray Blatche? Don’t players tend to improve between 23 and 24. Is poor misunderstood Andray getting any of that statistical love?

    (Ok, the above is intended somewhat tongue-in-cheek — but in fact, it’s worth looking at the stats from a different angle)

    5. Gilbert Arenas — in ’06-07, Arenas had a ws40 of 7.4. Last year, in 32 games, he was at 4.8 — ok, that’s it. He’s cooked. Here we see the usual career arc of an NBA player unfold. Except… when you look at the *actual data* that drive the single-number, we see something very interesting. Almost all the decline between his productivity in ’06-07 and last year comes from 2 phenomena — 1) he took fewer 3 point shots per 40 minutes, and 2) he got to the line somewhat less and shot a little less well from the line (his other numbers varied somewhat between the 2 years, but in ways that canceled out any cumulative effect).

    Given the first 32 games of a comeback from 2 years off, both shooting fewer 3s and getting to the line a little less (esp. less than the fabulous 9.8 times per 40 of ’06-07) don’t seem like indicators of a downturn, though of course one would have to keep an eye on him for reluctance to drive (he had no reluctance the first time he came back from the knee injury, however).

    Again, one of the key problems in deploying wp48 is forgetting that the single number is not *itself* a piece of statistical data. To *predict* off of it in an individual case, you have to look through the number to the data that drive it, right?

    So, while the Wizards are unlikely to be good, it is nonetheless possible that Arenas, Howard and Hinrich return to a previous level of productivity and play those kinds of minutes, that Wall has a rookie season more like a Rondo than a Rose, that Blatche and McGee eat up most of the Center minutes (leaving backup minutes mostly at the 4 and in the hands of Booker) that Booker’s rookie year mimics e.g. Taj Gibson’s, that Yi plays a little less than and no worse than for NJ last year, and that some of the really horrible players on the team (Young, Thornton and Armstrong) combine for no more than 1500 minutes. In which case, the Wizards could win about 44 games! So there.

    More likely, however, your number of 22 comes true.

    • September 16

      Great to hear from you. I got some counterintuitive posted I think :-).
      By point:
      1. I agree. I like their long term outlook. They have only 1 guy with a guaranteed contract in 2012 (Gilbert). They are perfectly positioned to start from scratch. But in the short term they are not built to be any good just cheap and disposable.
      2. The number 1 pick is a hit or miss affair ( You’re better off with later,cheaper picks.
      3. That’s why I like their long term chances. Selling this team as anything other than a rebuild is disingenuous.
      4. The longer their in the league the less likely a leap becomes.
      5. Gilbert is getting old though and has had a whale of a time staying on the court.

      It’s possible that everything will break the Wizards way but it’s not likely. It’s also bad for them in the long term as they’re better off with another lottery pick than at best an eight seed and the mother of all beatdowns in round one.

      • Tom Mandel
        September 16

        An eighth seed? No way — every star in the universe has to combine in a single light focused on the Wizards for us to make the playoffs.

        2. Of course the #1 pick is a hit or miss affair — *every* number pick is a hit or miss affair! In that sense, yes, having more picks is better. But, in fact, we did get 3 round 1 picks, and no matter what is better on paper, it’s hard to field a team w/ more than 3 rookies getting meaningful playing time — hard to sell to fans as well.

        Moreover, you didn’t answer my question — and maybe it’s not easy — about the years n+1, etc. results for teams w/ the #1 pick. It’s important, because events contribute to *emergent* results as part of larger complexes. You can’t just add the player’s stats (I mean you can if you assume it to be unrelated to any other facts) and assume you understand his ‘value.’ That is, you can if you define ‘value’ as how much his play helped win the games that produced the numbers, but he has other effects as well — which should be investigated not treated as anecdotal.

        3. It’s a total rebuild, yes. I’m sure the Wiz would send Gilbert to any team that would give them an expiring contract.

        4. Leap schmeap — a) is there an *average* improvment between age 23 and 24? How does it graph against years experience in the league? You can’t just stop w/ wp48 numbers, only visit them as part of assessment.

        5. Gilbert kept himself off the court last year w/ behavior. While on the court, as I think I sortakinda demonstrated, his play (not the single-number result) seemed to be more or less at the level of ’06-07. No? The next investigative step would be to see whether his rate to the line was increasing game to game, as he wore the rust off. Ditto FT%, number of 3s he shot, etc.

        This is not to say that had he managed to behave like an adult and not gotten himself suspended that he would have held up for the season — we don’t know that, I agree. But my main point was that his lower ws40 # was a place to *start* investigating him. On its own it had very little to teach. Doesn’t mean it was useless, but you see what I mean…

        • September 16

          “An eighth seed? No way — every star in the universe has to combine in a single light focused on the Wizards for us to make the playoffs.” I laughed hard at this.

          As for the draft. I’m pretty sure I’ve posted this. Hold on. Let me look. Got it! . The graph that applies is this one

          With the three picks and their location the Wiz have a good chance at finding a star, a starter and maybe a bench guy (but it’s not a given). Ther’s a good chance that one of those picks was wasted.
          3. And I would take an NBA GM job if offered in a cold second 🙂
          4. There is. There’s also a Leap Phenomenon among younger players (You’re just going to have to wait for me to post this though :-)) and a Fall for older players
          5. I’ll take a deeper look at Areanas as we get closer to the season. I’m a little Wizarded out right now though.

  6. Tom Mandel
    September 18

    Thinking again about your remark that ‘you’re better off w/ later, cheaper picks” — you must mean, obviously, *more* later picks. That is, surely a team is better off with a #1 pick than a #19 pick for example — even though the sliding pay scale does somewhat equalize the “value” of picks.

    I keep trying to get points across about the *limits* of statistical roll-ups — the limits are different from, but are as real as, the usefulness. Hence, in a case like this we have to find an instance of a team trading e.g. one high-lottery pick for other picks.

    One example occurred in ’05 when Portland traded the #3 to Utah for the #6, #27 and the ’06 #31. Looking at numbers only, that seems like a steal for Portland, but it wasn’t as Utah nabbed Deron Williams, who is good, and Portland picked up Martell Webster, Linus Kleiza and Jerome Anthony.

    What’s a counter-example that does instantiate your point?

  7. Tom Mandel
    September 18

    One other comment. The idea that more, later picks is better than fewer, higher picks relies on Dave and crew having demonstrated the low correlation between performance of a player in the NBA and his draft position, right?

    Alas, there seems to be a methodological error there. Actually, many — but let me start with one: how do you include in this analysis the “productivity” of guys who didn’t make it in the league? If you simply measure the numbers for the guy who *did* you are assuming a good bit of your conclusion.

    Another methodological problem: if e.g. #n is worth #n+4 and #n+12 (i.e. positions later in the draft), then what happens when you repeat the operation several times? Very quickly you get into situations that, even looked at numerically, can’t be matched to anything that would work in the real world. This should provide a hint that in real conditions of choice even the first operation might not work. Dig?

  8. […] a net of -22 Wins for a 27 win team.Good god! If this actually happens, I may have to apologize to Ted Leonsis. If only I had some way to quickly confirm the possible impact of this trade. Turns out I do. If I […]

  9. Chris McEwan
    October 20

    So you are predicting that Arenas will produce 2 wins based on his previous 3 seasons (two where he missed the majority of the games due to injury and 1 due to suspension)? I mean really? I am not saying that he is going to produce 20 wins but 10 wouldn’t be unreasonable. I agree that the over/under of 32 is probably high based on the supporting cast but expecting the Wiz to win between 4 and 16 wins is just as laughable. Your own model predicts that the Toronto Raptors will win more games than a team that has better talent, veteran leadership, defensive players and potential. I am sorry but I think you need to rethink your reasoning here.

    • October 20

      I am assuming Gilbert is more like the player of the last three years, you are assuming that he is the player of 4 and 5 years ago. I believe I am being more practical and reasonable. Five of the 48 worst players currently on nba rosters play on the Wizards (Thorton,Armstrong,Young,Collins, Yi see here). Adam Morrison will see significant minutes. Howard is coming back from a serious ACL injury and Gilbert is already acting like a basketcase. I’m really not seeing it. Even if Wall is more Rondo than Iverson (and they’re some signs that he might be this is still an epically bad team. Some food for thought: the 76ers won 22 games in Iverson’s rookie year, the Celts won 24 games in Rondo’s rookie year. Both those rosters were better than the current Wizards by a country mile.

      As for the Raptors? They’d be middle of the road except for the fact that Bargiani is the worst starting center in the league. The big worry with them is tanking (this is why they’re a stay away)

Leave a Reply

Your email address will not be published. Required fields are marked *