Wednesday, January 31, 2007

Back to Seventh Grade Biology

Don't ask me why, but I had a hankering while shaving this morning to do a blast from the distant past when I wore parachute pants and my greatest fear was breaking wind in class. Thus, I'll revisit the Scientific Method for the first time in ages for the lofty purpose of demonstrating Joe Johnson's value to the Hawks.

Observation - I've watched a number of Hawks games this year and have the distinct impression that the team is totally dependent on Joe Johnson. When he plays well, they're capable of beating most teams in the league. When he isn't playing well, they can be blown out by the Bobcats. Thus, despite the presence of three lottery picks (Josh Childress, Marvin Williams, and Shelden Williams) and a fourth player who would be a lottery pick if his Draft could be replayed (Josh Smith), the team behaves like a one-man team.

Description - To test the observation, we'll look at Johnson's numbers in the Hawks' wins and losses. This science stuff is just so easy! I should stop lauding my youngest brother for taking organic chemistry and physics at Michigan in contrast to my efforts to elude the science requirement with "The History of Medicine," "Biological Anthropology," and "Biology, Society, and Culture." (No joke, my term paper for the latter class was "Reproductive Success among Fashion Models." It was a really weighty application of principles of evolutionary psychology...and you should have heard the responses from modeling agencies when I contacted them about my potential research inquiries. I digress.) We'll use stats on a per 48 minute basis instead of raw numbers because we don't want the results distorted by increased or decreased minutes.

Prediction - Joe Johnson's points and assists will be higher in the Hawks' wins.

Control - We're dealing with a binary proposition, so I don't see the need for a control. B&B's Official Youngest Brother can correct me if my methodology doesn't work. That said, we can look at the other Hawks players in wins and losses to determine whether Johnson is the key or whether he rises and falls on the same tide that affects the rest of the team.

Falsifiability - The hypothesis can be easily disproven. If Johnson's numbers go up in losses, then I'm barking up the wrong tree. Fortunately, they don't:

Johnson in wins - 34.7 points, 4.9 rebounds, 6.2 assists, 2.0 assist/turnover ratio, 1.49 points per shot, 54.8 FG%, 51.1 3P%

Johnson in losses - 26.5 points, 4.7 rebounds, 4.1 assists, 1.0 assist/turnover ratio, 1.13 points per shot, 44.0 FG%, 29.6 3P%

Causal Explanation - There is a major correlation in almost every category. When the Hawks win, Johnson is an excellent scorer and shares the ball well with his teammates. When the Hawks lose, Johnson struggles to score, he doesn't get as many assists, and his turnovers go up. The disparity in three-point percentage is perhaps the most shocking disparity. That is evidence that Johnson's game is dependent on his shot. Johnson can't score in bushels when his shot isn't falling like Dwyane Wade. He doesn't have the same ability to blow by defenders and get to the line. Thus, everything for Johnson depends on being able to hit from outside, which sucks defenders towards him and allows him to get to the basket, break down defenses, and create for his teammates. As Ron Burgundy would say, it's science.

And now that we've established Joe Johnson's value to his team, for your viewing enjoyment, here is the promotional video that the Hawks created for his all-star campaign:

A few other notes on the disparity between the Hawks' wins and losses:

Zaza scores more in the Hawks' wins, but rebounds more in their losses. Could the latter be explained by the fact that the Hawks do better when Zaza gets help on the boards?

Josh Smith in wins: 21 points and 12 boards. Josh Smith in losses: 16 points and ten rebounds.

Josh Childress scores less, but rebounds more in the Hawks' wins. Marvin Williams is the same, only the disparity is really pronounced for him, as he scores 19 points/48 minutes in wins and 12 in losses. Marvin and Childress are also more efficient scorers in terms of points per shots in the Hawks' losses. Shelden Williams is also better in losses: 11/11 in wins and 13/14 in losses.


Anonymous said...

How about a picture of the parachute pants?

Steve S said...

Nice application of the method, but methinks you've misstated your prediction. Does the fact that JJ scores more points in win demonstrate that he drives the Hawks' success, or is it simply that the Hawks' success leads to higher point totals for all players? Correlation != causation.

Michael said...

The parachute pants have been figuratively burned.

Steve, I think I sorta answered your question. Johnson's point totals go up in wins more significantly than any other players on the team. In fact, several players (Childress and the Williamses) score better when the team loses. Thus, you can't really argue that the rest of the team produces better in wins and that causes Johnson's numbers to go up. I suppose you could argue that Josh Smith's numbers are much better in wins and therefore that Johnson's better numbers are the result of Smith giving him an effective wingman.

Chg said...

Any chance you would post the term paper?