Thursday, April 15, 2010

What The Numbers Can And Can't Tell Us About The Series

There's no team in the league for whom there cumulative season statistics are more representative of who and what they are than the Atlanta Hawks. The Hawks suffered no major injuries and made no changes to their roster or their rotation during the season.

I'm confident saying...
  • The Hawks are a team that scores 113 points per 100 possessions and allows 108 points per 100 possessions.
  • The Hawks are a team that rebounds 28.2% of their own misses and 72.8% of their opponents' misses.
  • The Hawks are a team that turns the ball over on just 13.3% of their possessions and forces turnovers on 15.4% of their opponents' possessions.
  • The Hawks make a free throw for every 4.7 field goals they attempt and send their opponents to the line once for every 3.6 field goals they attempt.
I'm especially confident that these numbers are representative of the team's characteristics considering that no member of the team had a season wildly out of character given his age and career.

I'm not at all confident that the following numbers tell us much about the team the Hawks will face in the first round.

MIL 09-10
Poss Off Eff eFG% FT Rate OR% TO%
Offense 90.2
48.3 18.1
1.048 48.6

Injury limited Michael Redd to playing 12% of Milwaukee's minutes but in those minutes he used a lot of possessions (usage rate: 23.7) despite being rather ineffective (again presumably due to injury). We're talking about just 12% (492) of the season's minutes but those minutes are reflected in the numbers above.

Redd 09-10PossOff Eff
On Court96494.4
Off Court6496106.3

Redd's ineffectiveness ground Milwaukee's offense to a devastating halt* but that has no bearing on this playoff series.

*As a point of comparison, here are the on court offensive efficiencies for the end of Atlanta's bench: Randolph Morris, 97.4; Jason Collins, 96.6; Mario West, 96.1; Othello Hunter, 94.5.

Nor will Andrew Bogut, could-have-been All-Star and serious candidate for the All-Defensive Team, have any impact on the series. Bogut played 55.8% of Milwaukee's minutes, was the team's best rebounder, and trailed only John Salmons in scoring* rate.

*Bogut gets more attention for his defense but he scored at a rate eight percent higher than Josh Smith and almost twenty percent higher than Al Horford this season.

Bogut 09-10PossOff EffPossDef Eff
On Court4095105.84106101.2
Off Court3365103.53360105.9

Milwaukee were seven points better per 100 possessions with Bogut on the court this season. Kurt Thomas can reasonably be expected to provide production approaching Bogut's defensively in the post and on the glass but Milwaukee will need someone (or someones) to replace Bogut's offensive production. Maybe it'll be Jerry Stackhouse, who found less resistance from the Hawks than the rest of the league since joining the Bucks in January.

Regardless of who (if anyone) steps up to replace Bogut, Milwaukee's first offensive option in the series will be John Salmons. He appears to be the only Buck guard capable of both getting into the lane and finishing there. I have Salmons scoring 28.1 points per 100 (estimated) on court possessions since his trade to Milwaukee. As a point of comparison, Joe Johnson averaged 30.3 points per (estimated) 100 on court possessions this season. Jamal Crawford averaged 31.3. Salmons, though, was more efficient, posting a 58.1 TS% versus 57.3% for Crawford and 53.8% for Johnson. Salmons also earned fewer assists and committed fewer turnovers per 100 possessions than both Johnson and Crawford.

Salmons, though, has relatively little impact on Milwaukee's team stats for the season. He played 37.6 minutes per game after joining the Bucks but still played just 28.3% of Milwaukee's minutes on the season. Kurt Thomas played 26.3%. Hell, Hakim Warrick has almost as big an impact on Milwaukee's season profile as Salmons. He played 25.6% of the team's minutes before the swap with Chicago.

Charlie Bell, for further example, played 40.3% of Milwaukee's minutes this season but just 5 seconds of the three games against the Hawks. Those minutes during which he scored at almost half the rate of Salmons while posting a 48.6 TS% likely have no bearing on this series either.

Here's how Milwaukee's minutes would be distributed in a 48-minute game based on the percentage of the team's minutes they played this season:

Mbah a Moute22.4

That's 52.6 minutes (21.9%) taken by players no longer with the team and another 20.7 minutes (8.6%) taken by players who will (likely) not participate outside of garbage time. Milwaukee's season stats maybe describe 70% of the team that the will face the Hawks.

The same table but with realistic expectations for playing time in the second column:

Mbah a Moute22.423

The results posted by the 2009-10 Milwaukee Bucks are not representative of the team the Hawks will face. Certainly not to the extent that the results posted by the 2009-10 Atlanta Hawks are representative of the team the Bucks will face. Before acquiring Salmons, the Bucks were getting outscored by two tenths of a point (0.2) per 100 possessions. Since acquiring him, they've outscored opponents by 5.7 points per 100 possessions. Neither sample size is comfortably large. The trick will be to identify which deviations from that record are anomalies of a given game and which are representative of who the Bucks presently are.

No comments: