Consumable Analytics

• Date: 01 May 2017
• Published date: 01 May 2017
• DOI: http://doi.org/10.1002/jcaf.22264

C

o

m

m

e

n

t

a

r

y

65

© 2017 Wiley Periodicals, Inc.

Published online in Wiley Online Library (wileyonlinelibrary.com).

DOI 10.1002/jcaf.22264

Consumable Analytics

Tim Chartier

Basketball entails running

up and down the court.

I remember practice in

middle school where we’d run

up and touch the baseline and

then back to touch the other

baseline. We’d run up and back

again and again. We could

have easily tracked, as possibly

my coach did, how far we’d

run. We’d also practice chang-

ing direction on the coach’s

whistle. Some players could

alter direction quickly, while

others seemed more concerned

about staying on their feet than

the rate of their acceleration.

Some players could run and

run. Others could dart from

spot to spot. Game strategies

were often designed to monop-

olize on such skills. How well

did we actualize those strate-

gies? Remember, it was middle

school basketball, so we also

had many other fundamental

objectives—like consistently

scoring on open shots.

Even at the professional

level of basketball, players’

abilities to stop and start with

the ball or how far they might

run in a game could, at best,

be estimated. Today, with the

availability of spatial data

tracked in every game, we can

examine parts of the game

with a high degree of accuracy.

In the fall of 2013, SportVU

cameras were installed in every

NBA arena. Every 25th of a

second throughout the game,

the cameras record the (x, y)

location of every player on the

court and the (x, y, z) loca-

tion of the ball. As such, every

NBA team has had the motion

of every player and the ball

captured for every game since

2013.

While interesting, SportVU

data in raw form, without

further processing, is limited.

Does it help to know Stephen

Curry was 15 feet behind the

three-point line at 4:24 in the

third quarter of a game? Pos-

sibly. If you additionally find

that he made a three-pointer

while covered closely, you have

more information. If you com-

pare this type of shot against

every game he played this sea-

son and find a trend, you have

more information. Note how

the data makes insight possible

but analysis of the data leads to

insight.

At the core of analytics is

data. Data, while the basis of

analytics, is generally not inher-

ently insightful. The analyst

must know who will view the

analysis, the decision or action

that will result from seeing the

analysis and the timeline of

the person receiving the work.

When factors such as these are

integrated into the research,

consumable analytics can result.

This isn’t a guaranteed result

of analysis. As an academic, I

often hear concerns that analyt-

ics produced from academia can

be esoteric, even if interesting.

Again, careful attention

must be paid as to the goal

of analysis. The businesses

that consult with my analytics

group at Davidson College take

action on our work. Sports

teams use our research as part

of their game preparation.

How do we keep from

producing narrowly defined

insights? First, we communicate

with whomever will make deci-

sions from our analysis. Even

in sports, a player has different

needs from a coach. A team has

different interests than a fan.

Knowing who will receive the

work impacts what we analyze

and create. Second, we learn the

types of decisions being made

from our analytics. Will a coach

use our analysis to help strat-

egize for a game? Will the work

be viewed in a few seconds

at halftime to devise correc-

tions to game time play? Will

a business use our analytics to

directly make a decision, or will

the insights be integrated into