Villanova Basketball Advanced Stats: Taste the Rainbow

Kirk Goldsberry - cartographer, spatial reasoner extraordinaire, professor, Harvard scholar, and, the most relevant,  staff writer at Grantland - has long been providing some of the most incredible and accessible content for basketball fans, casual or otherwise.  It's a topic I've brought up before - his work on offensive rebounds and where they're recovered in the NBA was referenced in another article - but it's coming up here again.

Much of what he contributes to Grantland is based on the data-crunching he details here, in a paper submitted to the Sloan Sports Conference.  He utilizes shot data provided by the SportsVU video tracking system, which utilizes missile tracking technology to log the positions of all players on the court (and the ball), to develop individual shot charts such as these, that show volume and effectiveness of a player's shooting around the half-court.  These charts boil down a season's (or multiples) worth of shooting into one easily digestable half-court, highlighting the strengths and weaknesses of a player, and pointing out things that are really impossible to see in the box score.  His contributions are my second favorite part of Grantland's generally high-quality basketball coverage (Zach Lowe remains the king).

Most of the commenters who've been reading these / been on this site in general know youngBUCK and I generally share the same views;  we're actually good real-life friends with a shared passion for Villanova basketball, and sports in general.  I'll be collaborating with him, as we have in the comment section, on several of the articles/stats I'll be presenting this year.  Though he occasionally occupies the post of president troll around here, he'll be stepping down (mostly) to bring the business to these articles.

In the course of a constant quest to provide you guys with new, totally awesome statistical content (just kidding - it's honestly mostly for me), youngBUCK and I took a page out of Kirk's book and figured out how to develop shot charts for each of Villanova's players.  Then we talked about it - this will form the bulk of the article to follow.

The data for the charts was pulled off publicly accessible websites that display shot charts during and after games, and converted to correspond to a plotted basketball court.  The rest of the process was easy-ish, now-easily-repeatable, and probably boring to you.

The charts themselves match Goldsberry's as close as possible, in terms of setup - smaller squares mean lower volume, bigger = more, etc., and the color scheme used to designated hot/not/maybe just average can be seen in the graphic.  To come up with the colors, I divided the basketball court into 14 'zones,' and compared the player's average to Villanova's average in that zone.  Obviously, more data (from teams that aren't Villanova) would be better in terms of overall analysis, but these numbers can be seen as a means of comparing Villanova's players to each other.

For Villanova, the shot charts were only available from the Kansas Jayhawks game on - so there are something like 6 or 7 games missing from the data.

If you've got any unanswered questions about the actual process of manipulating and processing this data, feel free to ask us in the comments.

On a semi-regular basis, youngBUCK and I will be posting discussions we have via email, mostly on Villanova basketball (and these shot charts), occasionally drifting into the wider world of sports and other random events.  There's some great data and info in here, but also the usual dose of nonsense.

Here's how the basketball court was divided into 'zones,' to provide reference for the numbers (courtesy of AutoCAD).

Ryan Arcidiacono

Region	Makes	Attempts	FG%	expected average
1	5	9	55.56%	63.87%
2	5	8	62.50%	63.04%
3	6	14	42.86%	40.78%
4	2	7	28.57%	44.05%
5	6	15	40.00%	28.57%
6	1	1	100.00%	38.89%
7	6	16	37.50%	20.75%
8	0	0	#DIV/0!	40.00%
9	4	10	40.00%	37.25%
10	3	11	27.27%	40.00%
11	5	6	83.33%	37.97%
12	13	50	26.00%	32.81%
13	13	34	38.24%	40.46%
14	20	46	43.48%	38.31%

From: youngBUCK

Subject: Arch's Shot Chart

To: corollary

Churned out part 1, section A, subsection i ( I have a feeling we're going to stake a claim to the long-winded sports position on the interwebs).

Main takes from Arch's chart:

• Needs to improve finishing around the rim (embarrassing).
• Stop wasting threes from right side. Don't take them. Absurdly high volume for such low numbers. Know your spots.
• Speaking of which, keep taking two point jump shots. Contradicting metrics surest law is a fun way to un-legitimize what we've been trying to do here.
• Needs to be shaded towards strong side of the court, especially off the ball.
• There's no way to see this on the visual, but STOP TAKING THREES IN TRANSITION. A lot of those large blue diamonds on the right wing are pull ups in transition that rob us of easy 2s.

Picking out some of the finer points - with Phil Booth coming in (and HOPEFULLY some sort of Ennis development), we can expect Arch to get more time as the two. Look for him to get some off the ball screen to free him up from three on the left wing. Everyone has memories of Arch both catching and shooting from the left wing, or (more painfully) getting to 'his spot' off the dribble. Both mostly ended the same way: bad misses.

In order for him to be considered a legitimate guard (instead of merely a volume shooting facilitator better suited for a mid-major), he has to step up his efficiency around the rim. His numbers are absolutely horrible in close. This was a huge problem for the entire team this year, and Arch was arguably the biggest perpetrator. He showed flashes of being able to penetrate (and at times make remarkable passes that led to mostly made lay ups). But not being able to score when he gets there is far too Chennaultian for a starting guard that logs the most minutes on the team.  On the bright side - he can't get any worse. Here's hoping he starts seeing shades of red... or at least gets out of the blue. This is high level division one basketball and you simply can't excuse those numbers on two foot shots.

Arch's ability to make the long two pokes a hole in the tired ‘least efficient shot in basketball' argument. His right baseline jumper is superb, when he shoots it, and his long two from straight away comes in very handy (especially when we are in need of a make to end a scoring drought). These are the types of shots that he can get whenever he wants (and will only be able to get more of next year, as our offense spreads out even more as Jenkins becomes a perennial 3 point specialist). Similar adjustment here, as it was with his threes: know where you make them at a high rate, and find those spots on the court.

Arch's three point breakdown very glaringly screams: STAY AWAY FROM THE RIGHT WING AND STRAIGHTAWAY THREES. It has become an exercise in futility. If you consistently miss shots (and it's not the last 2 minutes of a close game where you transform into Pistol Pete 2.0), don't take them from those spots. Way too many possessions were ruined when Arch pulled up from three, contested, in a spot we've seen him miss from thrice as much as make from. He is extremely efficient from the left wing - TRY TO GET THERE MORE. Know your strengths, and shoot to them. That's smart basketball, which is expected of Arch (because he's a PG, not because he's white. Also, because he's white). But really, considering what a polarizing player he is, it's only appropriate Arch has such a bipolar shooting chart. There's hot, and there's cold - almost nothing in between.

Takeaway in 10 words or less: Avoid the right side, improve at the rim. Get better at counting.

From: corollary

Subject: Arch's Shot Chart

To:  youngBUCK

Good to get these rolling.  Spot-on analysis, as it's where I would have gone - which I consider spot-on.  As a side note for both of us, I'm now able to get these shot chart splits in transition and for the various lineups - so we can prove he took his shitty shots from shitty locations on the break.  Obviously, the fact he shot 25% from transition and 40% in the half-court (and his only solid areas are on the left wing) kind of points directly to that, but still.

While you're killing him for his inefficiency from in close (and rightly so), don't forget the volume and actual numbers.  Volume was super low -  he took 17 shots in all the counted games from the paint area.  Ridiculous for a point guard. And, as another side-note, because 90% of all layups/dunks are logged at the exact same pixel spot in the shot charts, I made the diamonds in the paint all equal to each other (on their respective sides), so the chart would have more of a visual at the rim - otherwise it would basically be 2-3 diamonds on each side.  A little messy, but probably necessary.

He wasn't bad at straight on 3s this year - he hit 38+% of them, after all - but it was worse than the team as a collective shot from that area (about 40.5%).  It's a shame his highest volume area doubled as his worst - he was actually a very good 3 point shooter for most of the sampled games and zones on the court - just under 37% from 3 overall.

Legitimately lol'd at this bit:  instead of merely a volume shooting facilitator better suited for a mid major -where would our comments/articles be without a bit of Arch-baiting?  Every point guard was so horrible at getting to and finishing at the rim - really need that out of a distributor, as dribble penetration is what bends the defense and creates open looks - incredibly important on a team that shoots as many 3s as we do.  If there's no real reason to fear the penetration, there's no reason to rotate and contain it.  Arch may be able to find the pass if bigs/wing help do rotate, but the smarter teams just won't bother, and will dare him to finish on his own - which he never does, and isn't great at anyway.  Arch played more games than he took shots at the rim, which is absurd.

His excellent shooting from the mid-range was a trend all year - think me and 'Night Train' had a convo about this about halfway through the season, when he was hitting 60% of his jumpers that weren't 3s - crazy unsustainable, but crazy nonetheless.  It's always a good weapon to have in the toolbox, and offenses can be built around efficient mid-range jump shooting guards (for this exercise, but LMA and Dirk as bigs work too) with good shooting around them - a la Tony Parker, or the resurgent Monta Ellis in Dallas.  If he could build his game around penetrating and dishing, and taking the pullup when the defense sags off to prevent penetration, he could be an absurdly poor man's Parker for us - which would still be a great thing.  Needs to improve at the rim in attempts/efficiency for that to become a real thing.

The main point here is that mid-range shots are not bad things if you have someone who can consistently hit it at a high level. It's when a team (or a player) takes high volumes at average to meh-efficiency that it's a problem/waste of possession.  No one fears a mid-range jumper that's being hit at a 40%-or-less clip, because there's no reason to.

Darrun Hilliard