Game Charting Insights: Event Locations

Project scope: we tracked the beginning and end location for possession events (shot, foul, turnover).

Basically zones 1 to 5 are three point locations, 6 to 10 are two point zones outside the paint, and zones 11 to 14 are in the paint. For any backcourt events we use a "-1" zone, and sometimes a "zone 0" for deep frontcourt events (like a foul crossing halfcourt or something). 'Away from the ball' fouls or other events just have an end field.

So if on a play a guy catches a pass in zone 4 and then drives down to zone 10 for the shot, then the shot comes with a 4 start zone and 10 end zone.

Event Location data for shots/turnovers/fouls

We'll examine only the "end" location for a play...the zone on the court where the possession event actually happened. Since we're trying to just get an understanding of league wide averages at this point, we'll also combine the mirror image zones into one (i.e. the Zone 1 and Zone 5 locations are both corners threes).

Zones	Zone Type	FG%	Ast'd	Blk'd	FTA/ FGA	S-Foul	Foul	T/O%	Pts per 100 Poss*
1+5	Corner 3's	.425	95%	1%	0.04	0%	2%	5%	118.8
2+4	Wing 3's	.349	80%	0%	0.08	1%	9%	14%	88.7
3	Straight-up 3's	.388	81%	0%	0.10	1%	13%	10%	105.3
6+10	Baseline 2's	.439	69%	2%	0.13	3%	13%	11%	81.9
7+9	Wing 2's	.385	62%	2%	0.11	2%	11%	15%	68.6
8	Straight-up 2's	.453	50%	0%	0.20	2%	8%	14%	85.7
11+12	High Paint	.450	38%	8%	0.45	15%	22%	16%	85.2
13+14	Low Paint	.600	54%	11%	0.46	19%	23%	8%	117.1
0	Frontcourt/Missing	.256	45%	0%	0.40	4%	34%	41%	28.5
-1	Backcourt	.000	0%	0%	5.83	0%	48%	43%	53.1
All		.469	59%	5%	0.28	10%	17%	13%	93.6

* Points per 100 Possessions does not account for offensive rebounds in this instance
** Shooting Foul, Foul and Turnover percentages are expressed as a percentage of possessions for the zones in question

The most lethal spot on the floor? That would be in the corner for the short three pointer! Now of course our sample may not represent very well a full season of play, so it's clear that the money zone is basically in the "deep paint" right underneath the hoop, but the corner three territory is also important.

In terms of where fouls occur that depends a bit on if you limit it to just shooting fouls or all (defensive) fouls. For shooting fouls, you're looking at the in the paint zones, where the vast majority of physical contact occurs, there are however a lot of non shooting fouls that occur in the two point jumper zones and behind the three point line.

Turnovers are infrequent deep in the paint (since there's not a lot of dribbling), but fairly constant in all other zones excluding the almost pure catch and shoot corner threes (95% of the corner threes are assisted).

Let's look at the distribution of events a bit more closely:

Zones	Zone type	FGA	FTA	Sfoul	Foul	T/O	Poss
1 to 5	Three's	21%	5%	1%	8%	16%	18%
6 to 10	2-pt Jumpers	32%	15%	7%	18%	31%	29%
11 to 14	In the Paint	46%	74%	90%	62%	37%	48%

So for those people lamenting the NBA is just endless three pointers or slam dunks, we could counter that in fact there's an almost "perfect balance" with a near 50/50 split between the number of possession events taking place inside the paint, and outside the paint!

Of course 90% of the shooting fouls do occur in the paint, but not with a similarly lopsided turnover rate, although it should be noted that with passing turnovers we were charting the location of the passer when he threw the ball for this exercise, not the zone of the intended receiver of the pass.

All right, that's the end point -- the actual location -- of possession events. What about the starting point though, where the player got the ball on route to the action?

Player with Ball/Possession Event -- STARTING location

Once again we will use the same fourteen zone scheme but we will mention that away from the ball events are treated as having an initial 'null' starting zone and just the end zone...which means when we lump them into the zone 0/frontcourt class, you find a lot of free throws "started there" but also a lot of turnovers (3 seconds, moving screen, etc)

Zones	Start Zone Type	FG%	Ast'd	Blk'd	FTA/ FGA	S-Foul	Foul	T/O%	Pts per 100 Poss*
1+5	Corner 3's	.423	82%	1%	0.10	2%	6%	7%	107.5
2+4	Wing 3's	.419	52%	4%	0.19	6%	12%	12%	91.1
3	Straight-up 3's	.446	61%	3%	0.15	6%	11%	11%	95.3
6+10	Baseline 2's	.476	77%	6%	0.23	9%	15%	8%	94.1
7+9	Wing 2's	.442	65%	6%	0.18	7%	12%	11%	86.3
8	Straight-up 2's	.502	64%	4%	0.25	6%	7%	11%	98.0
11+12	High Paint	.498	77%	7%	0.51	19%	24%	10%	97.7
13+14	Low Paint	.643	61%	8%	0.36	15%	19%	7%	121.3
0	Frontcourt/Away From Ball	.360	28%	4%	1.63	10%	47%	32%	61.7
-1	Backcourt	.462	9%	7%	0.43	10%	22%	21%	86.7
Total		.469	59%	5%	0.28	10%	17%	13%	93.6

So in general the expectations are pretty similar on a league wide look getting the ball in most zones, with the exceptions again being the close in zones 13 & 14, as well as the corner three's where catch and shoot is the order of the day. Otherwise starting a move from the wings is slightly less optimal, but this varies considerably from player to player.

Could it be very valuable to teams to have charts like the above for specific players? Ah, now that would be telling...

Moving on then, another logical area to look at is what happens when a player starts in a certain zone and then moves to another zone for the event.

Possession Efficiency based on Change in Location from start to end

Zone Movement with the ball during a Possession

Start Zone	End Zone	FG%	Ast'd	Blk'd	FTA/ FGA	S-Foul	Foul	T/O%	Pts per 100 Poss*
3-Pointer	3-Pointer	.390	89%	0%	0.00	0%	5%	9%	103.3
3-Pointer	Two-Point J	.393	37%	2%	0.00	3%	11%	15%	65.9
3-Pointer	In the Paint	.527	38%	11%	0.05	15%	19%	11%	92.2
Two-Point J	3-Pointer	.375	95%	0%	0.00	3%	5%	6%	103.3
Two-Point J	Two-Point J	.427	78%	1%	0.05	1%	7%	10%	78.3
Two-Point J	In the Paint	.517	53%	13%	0.05	15%	18%	10%	92.1
In the Paint	Two-Point J	.444	63%	0%	0.00	0%	4%	17%	72.7
In the Paint	In the Paint	.596	66%	8%	0.07	17%	22%	8%	108.2

Some interesting things turn up. When a player with the ball starts in a three-point zone, the most efficient move he can make is none at all -- shooting a three works out better overall than moving in for a two point jumper or even driving into the paint.

When a player gets the ball in Two-Point Jump Shot territory, the most success has come from backing up and taking a three!

When a player gets the ball in the paint, if we discard the tiny sample where the player winds up shooting from long distance, then looking to score inside is the best basic scenario.

What's the common thread in all this? Two-point jump shots stink on a league wide basis. So yes in some ways the complaint that "three's and dunks" dominate play in the NBA is true -- the two most effective options on offense are trying to score/draw the foul in the paint, or getting a good look from behind the line where the extra point payoff makes such a difference.

One last point we'll address in this article is that despite the angst often heard when players take a pass and go up with the shot right away ("they shouldn't settle for that shot..."), the reality is players in the NBA are more efficient overall when they don't change zones after receiving the ball:

Possessions that begin and end in the same zone

Start+End	Zone Type	FG%	Ast'd	Blk'd	FTA/ FGA	S-Foul	Foul	T/O%	Pts per 100 Poss*
1+5	Corner 3's	.392	98%	1%	0.03	0%	2%	4%	112.3
2+4	Wing 3's	.366	85%	0%	0.07	0%	5%	9%	99.5
3	Straight-up 3's	.408	93%	0%	0.04	0%	5%	7%	113.4
6+10	Baseline 2's	.464	86%	1%	0.09	1%	8%	7%	89.1
7+9	Wing 2's	.424	79%	0%	0.05	1%	7%	9%	78.4
8	Straight-up 2's	.445	84%	0%	0.29	2%	2%	7%	93.3
11+12	High Paint	.439	75%	5%	0.50	14%	24%	14%	82.6
13+14	Low Paint	.653	60%	8%	0.39	16%	20%	7%	123.8
	No Zone move	.465	77%	2%	0.17	5%	10%	8%	100.8
	With Zone move	.479	46%	8%	0.35	12%	19%	14%	92.8

Looking at league wide averages tells us a lot about the nature of NBA basketball, but of course there are many individual players who defy the conventions and norms, so we're not suggesting Dwyane Wade should give up trying to beat his man off the dribble! No, rather this is a first look at some general rules about what the value is in having the ball in a player's hands at a given point on the court. The variance from player to player though can be enormous!

Also see: Game Charting: The Value of a Good Pass (03/20)
Part II of floor locations will bring some other elements into play -- offensive rebounds, and counter attacks!

Game Charters for the "Floor Location" Project:

Adrian Lawhorn, Alixander Lim, Andre Warner, Brad Burnett, Brendan Gildea, Brett Steele, Brian Cole, Cameron Tana, Charles Floyd, Craig Ward, Dana Henderson, Daniel Kelly, Dave Paolino, David Mintz, Demetrius Burrows, Dmitri Salcedo, Don Ellis, Dwayne Killings, Eric Patten, Eric Wallace, Erich Hothem, Frank Mantesta, Greg Freeman, Greg Humphreys, Husamettin Erciyes, James Stevenson, Jeremy Brown, John Magee, Justin Kimbrell, Kevin Bartlett, Matthew Powell, Miguel Cuaron, Mike Wolf, Noah Libby-Haines, Noah Purcell, Norling Kho, Patrick Clark, Patrick Sheehy, Phil Edwards, Raj Kannan, Ray Maze, Rich Schmidt, Rob Ireland, Rob Stewart, Sandy Weil, Shawn Krest, Susan Nelson, Thomas Loden, Tom Lore, Tom Powers, Zach Ellin

Thanks as always to the noble efforts of the charters -- they are the ones who will be pushing NBA statistical analysis to new heights!

Interested in doing some game charting? Send a message to: charting@82games.com