Enboldened by the front page treatment that Julio & Bos were kind enough to
provide for my last statistical discussion, I will attempt another.
In this discussion, I will propose a new statistical category in order to
fill a gap that I believe had existed in the current box score data
categories. The new category is the following: Two point field goal
percentage.
Existing data formats give the overall FG % and it separates out the 3P FG%.
But what is missing is an indication of how well the team shot its two point
field goals.
A simple illustration. Assume that a team shot 50% on two pointers, 40%
from beyond the arc, and takes 2000 shots in a season, a reasonable
assumption. If the team shot all two point shots, then its FG% is 50%. If
however, another team shoots the exact same number of shots but 30% of the
team's shots were treys (31% of the shots taken by the Duke teams from 94 to
99 have been treys), then their FG percentage goes down. The results are
illustrated below:
Team FG FGA 3P 3PA 2P 2PA FG% Total points
A 1000 2000 0 0 1000 2000 .500 2000
B 1000 2000 240 600 700 1400 .470 2240
Team A shoots all 2P field goals while Team B shoots 30% of its shots from
downtown.
Herein lies the need, in my eyes, of the 2P FG%. If you compare the
shooting % of both teams, Team A looks better because its FG% is three
percentage points better. Three percentage points is not trivial as
typically, the middle pack of 4 or 5 ACC teams shoot within 3% points of
each other. But Team B is actually shooting better because 40% on
treys is a scorching rate. Last season, only four teams in the _country_
shot over 40%. And as a result of shooting those threes, they got an extra
240 points over the season.
Thus, while the separate shooting percentages, of 40% for treys and 50% for
2P field goals, are informative, the combined percentage reveals very
little. It fails to answer the critical quesiton: Is the team shooting
lower because of taking more treys or is it because the team is shooting
worse overall?
Now for a historical example. An excellent comparison is between two of the
best ACC teams ever, Duke 86 and Duke 99. Their relevant stats are as
follows:
Team FG% 3P 3PA 2P 2PA All FG Points Games
86 51.3 0 0 1245 2429 1245 2490 40
99 51.4 293 739 951 1683 1244 2781 39
This is just about the most perfect example possible. Both teams shot
within 0.1% and came within a single FG of each other for the season. Each
played just about as many games as possible under the best of circumstances
as 40 games is still the NCAA single season record. The 86 team shot no
treys obviously because this was the year before the rule was adopted in the
NCAA.
Their separate, 2P and 3P, field goal percentage is as follows:
Team 3P % 2P %
86 0.0% 51.3%
99 39.6 56.5%
The 39.6% rate from three point land was exceptional. Duke ranked #7 in the
country last season in this category. What the cursory FG% stat hides is
the massive gap between the two teams in 2P field goal shooting. The 99
team shot the 2P field goals 5.2% percentage points better than the 86 team,
a difference that is far from trivial.
The reward for taking the shots from downtown are of course, more points.
The 99 team, while shooting one FG less than the 86 team, scored 291 points
more.
I hope that readers, and especially the sportswriters, can take a good look
at my case for the importance of the 2P FG percentage and reflect it in
their future thinking and writing.