Saturday, November 6, 2010

Defending The Pharaoh While Waiting For Superman

In a comment on his blog, Andres Alvarez (developer of the Wins Produced Viewer) essentially blamed the Pharaoh, Rashard Lewis, for Dwight Howard's failure to win a championship already.  Here's what Dre said:

"Poor Dwight, sub out Lewis for an average player and Orlando should have at least won one championship."

I found this comment surprising since Dre's own Wins Produced Viewer shows that Rashard's production was above average (0.130 WP48) for the 2009 playoff run when the Orlando Magic came closer than they ever had to winning an NBA championship in Howard's career.  I decided to investigate further by taking a look at Rashard Lewis' production in the NBA Finals against the LA Lakers.  Was he really the reason the Magic lost 4-1 to the Lakers?

This article will use Win Score and Estimated Wins Produced, statistical models created by Professor David Berri from the Wages of Wins Journal, to measure how much a player's box score statistics contributed to their team's performance. An average player produces an estimated 0.100 wins per 48 minutes (EWP48), a star player produces 0.200+ EWP48 and a superstar produces 0.300+ EWP48. More information on these stats can be found at the following links:

Rashard Lewis is not the reason the Lakers beat the Magic in 2009. He produced an estimated 0.241 wins in the Finals (0.054 Est. WP48) going against Laker power forwards Pau Gasol and Lamar Odom (see the spreadsheets labeled 'MAGIC' HERE). An average power forward would've produced an estimated 0.286 wins (0.064 Est. WP48).  So Rashard wasn't that bad compared to an average power forward going up against Gasol and Odom, as illustrated in the spreadsheets HERE.

If I look at it game-by-game, then an average power forward would have swung Game 4 in Orlando's favor but none of the others. An average power forward would've produced an estimated 0.218 WP48 in Game 4 compared to the estimated 0.047 WP48 Rashard produced. So an average power forward would've gotten the series back to LA for Game 6 but not necessarily gotten Dwight Howard a championship.

The real question is, "Who's to blame for Dwight Howard not having a championship?"

I used 82games.com's Sortable NBA Team 'By Position' PER Ratings for 2009 to determine what level of production should have been expected against the Lakers.  I used the following formula to come up with the expected production for each player against the Lakers in the 2009 Finals:

([Lakers' Opp. PER @ position]/[Average PER @ position]*ADJ. P48) - [AVG. ADJ. P48 for Position] + 0.099 = Expected WP48

Here's the opponent PER for the Lakers at each position, along with the average PER at each position:

POS - Lakers' opponent PER (Average PER)
PG - 16.4 (17.1)
SG - 13.5 (15.8)
SF - 14.9 (15.7)
PF - 16.6 (17.7)
C - 17.1 (18.3)

Based on those numbers, the biggest gaps between actual and expected Wins Produced in the Magic in the 2009 Finals were as follows: 

Most Disappointing Magic Players in 2009 Finals
1. Rafer Alston (-0.458 Wins Produced)
2. Jameer Nelson (-0.427 Wins Produced)
3. Courtney Lee (-0.367 Wins Produced)
4. Dwight Howard (-0.340 Wins Produced)
5. Marcin Gortat (-0.190 Wins Produced)
6. Rashard Lewis (-0.175 Wins Produced)

There were a lot of people to blame before you get to Rashard Lewis.  The only player with an excuse for their decreased production in the Finals was Jameer Nelson who was returning from an injury.

More importantly, Dwight Howard was a bigger disappointment in the Finals than Rashard Lewis.  Howard was expected to produce 0.332 WP48 against the Lakers' defense but only managed to produce 0.256 WP48.  Howard still produced like a star in the 2009 Finals, but he was expected to produce like a superstar and came up short.

Fans of Howard and the Magic can stop blaming the Pharaoh and keep waiting for Superman to come through in the Finals.  Of course, the Miami Heat could make that a really long wait.

4 comments:

  1. Mosi,
    Dang excellent post! You may force me to admit hyperbole. However, last year a real PF instead of Lewis would have been huge. In 2008 even with a real PF competing with LA or Boston would have been iffy. So maybe I can restate as: if they'd had an average PF instead of Lewis then perhaps they'd have made another finals.

    Arturo's made similar comments about Dwight when we've brought him up before. This may be a ripe subject for the podcast.

    ReplyDelete
  2. Dre:

    Glad you liked the post. So, restated, you basically think they should've made the Finals last year for a rematch against the Lakers instead of getting upset by Boston.

    To that, I'll say this - Lewis was battling a viral infection the entire series (see http://www.nba.com/2010/news/05/26/magic.lewis.ap/index.html).

    I can't believe I'm repeatedly defending Rashard Lewis...

    ReplyDelete
  3. I'm confused by these PER numbers. Isn't 15 average? Does this mean that the average league PER is above average? Or are those values for starters only maybe?

    ReplyDelete
  4. I thought average PER was 15, too, but the numbers at 82games.com didn't average out that way.

    It could be the same thing that happens with WP48 when you look at the unweighted average across the league which is 0.080 (see http://dberri.wordpress.com/2009/10/09/the-contending-spurs-fail-to-miss-blair/#comment-65719).

    Since I don't know that much about PER, I wanted to use the numbers that could be replicated by someone else.

    ReplyDelete