The Trends & Nuances of Villanova Basketball's Advanced Statistics

In the spirit of brevity - and in recognition of the words I crammed into the rest of this article - this piece takes a look at a few statistical trends I've been tracking over the course of the season. Most of the analysis boils down to lineup data, shot charts, and the nuggets available from great sites like kenpom.com and hoop-math.com.

Let's get to the analysis.

Who's the team's best power forward?

While there may be a little difficulty in assigning clear ‘power forward' status to anyone on the team, there's far less in defining who the 2^nd tallest player on the floor is at any given time. And the players who have been shouldering the burden for the team at that spot include: Josh Hart, Kris Jenkins, Mikal Bridges, and Darryl Reynolds.

While the small sample size caveat NEEDS to be thrown out there - some of these samples are as small as 44 possessions (which, at Villanova's average possession length, is something like 25 to 30 minutes over the entire season) - I took a look at the on/off numbers for each of these ‘power forwards' while the others weren't on the floor to get a sense of how well they were handling the position. All tables are courtesy of hooplens.com - my public source for on-off data was taken down last year.

Josh Hart has carried over his strong play as the team's de facto ‘backup' power forward even after his transition to the starting lineup. Though his place in the starting lineup is, nominally, small forward, he's logged over 120 possessions on the floor with neither Bridges nor Jenkins in, and only one of Reynolds/Ochefu - the largest sample any of the 4 have as the team's sole ‘2^nd big.' While these lineups have been somewhat hurt on the boards (allowing opponents to rebound 37.1% of misses, while 30.3% are rebounded when the configuration is not ‘on' the floor), it has excelled nearly everywhere else. The defense has been a staggering 22 points/100 possessions better with Hart working as the 2^nd big, forcing turnovers on 27.3% of possessions and LOCKING DOWN both the interior and the perimeter (35.2% on 2s, 25.0% on 3s). The team's net rating during this time approaches 50 points per 100 possessions, which is obviously excellent. Hart has proven up to his increased role as a ‘big' this year, pulling down 18.8% of available defensive rebounds (which outstrips JVP's best efforts, from any year) while maturing as an interior offensive force. For my money, he's been the team's best ‘4' so far - which is sort of unfair, as he's probably also the best ‘3' (and ‘2'). Still, the team has performed very well with him as the 2^nd biggest dude on the floor - a tribute to his skills and willingness to bang inside.

Here's where we start to get to small sample size theater. It's somewhat difficult to tease out exactly who the power forward is in some lineups, so I mainly restricted this analysis to lineups with only 1 of the 4 identified as ‘competing' for a power forward spot. For what it's worth, Sports Illustrated's Luke Winn noted in last week's power ranking article that Villanova does extremely well with the pairing of Jenkins and Ochefu as the nominal bigs of the team - while some of that is certainly tied to the fact they're both in the starting lineup, it does point toward the possibility that Jenkins has been an excellent ‘4' to pair with Ochefu (something I noted in a breakdown last year of what Jenkins brings to the team). And, for what ANOTHER site is worth, KenPom's lineup breakdowns of the past 5 games suggest the only two really in the rotation for ‘power forward' are Hart and Bridges (Jenkins, likely due to his 3PAR and middling individual rebounding totals, is a small forward based on Mr. Pomeroy's lineup algorithms). So, as always, I'll be careful going too nuts on conclusions from a small sample size - you should too!

Bridges has held up well in the 15-20 minutes he's played this season as the sole ‘4' candidate on the floor - the team's offense has scored well, and the defense has done the same .However, there have been some really odd results on the boards - likely due to the size of the sample. The lineups containing Bridges (with no Hart or Jenkins) have rebounded 50% of the team's misses on the offensive end - but allowed the other team to collect 43.8% of their clanks on the other side. While I'm putting this one down to noise due to the size of the sample involved, it'll be interesting to track these two anomalisies as the sample expands, to see what the real line ends up closer to. One last note - the lineups with Bridges (plus no Hart and Jenkins) have taken almost 60% of their FGAs from 3 - way too high, especially with the amount of space that should theoretically be open on the way to the basket.

Jenkins, despite his success in pairing with Ochefu in the starting lineup (which, incidentally, includes Hart) appears to have really struggled (from a team perspective) in his minutes without wing backup. The offense in this small chunk has limped to an offensive rating of 86 points/100 possessions (almost 30 points below the season average) while barely hitting the offensive boards. This lines up with Jenkins' individual rebounding rates, which are somewhat staggeringly low for a starting 'power forward.' And, while the defense has done alright over the same minutes, the data available so far suggests Jenkins needs his wing brethren on the floor for the team to function well with his presence, and the positives/negatives that entails.

And, finally, a peek at the dreaded ‘two big' alignment. Typically, only one of Ochefu and Reynolds is on the floor - they've generally been splitting time as the team's ‘center.' When they've both been on the floor, though, the results have been pretty disastrous. The team actually has a net negative rating, having been outscored by 9 points per 100 possessions, in the small amount of time the two-big alignment has been out there. The team is shooting a horrid 10% from 3 point range during this time - though they're also defending from range quite well ('beating' us with a 12.5% rate). Ultimately, the ‘2-big' experiment has not worked during the small sample it's been used - while it's probably something that should still be tried in small doses, it's trending toward more and more likely that it just won't work for this team as currently constructed.

Bridges IS That Good

Some historical context for just how good Mikal Bridges has been so far:

First, a quick explanation of Box-Plus Minus - one of the top public metrics out there - measures player quality by attempting to tie box score production to actual team contribution. This analysis ‘relies on a player's box score information and the team's overall performance to estimate a player's performance relative to league average.' This metric is a per-100-possession stat; +5 means a player is 5 points better than an average player over 100 possessions. A full explanation of the regression, calculation, and accuracy of the metric can be found here. Mostly, though, it's one of the better ways to compare player quality across years, and even eras.

Mikal Bridges, in terms of this metric (though admittedly, the data only goes back to 2010-2011 on Sports-Reference), is currently having the 13^th best season on record for a freshman (for players that log >15 minutes per game). Check out this list:

This list is littered with NBA players. Of the 19 players who have graduated/aren't freshmen this year, 12 were chosen in the lottery (14^th pick or earlier), with all but one (Javon McCrea, who went undrafted) going in the 1^st round. The players who haven't graduated yet? One's Ben Simmons, the presumptive #1 pick in next year's draft, while the rest are either on draft radar or superstar walk-ons - Nicholas Baer FTW. It must be stressed that Bridges is logging extreme efficiency while playing just over half of the team's available minutes - but the optics of that comp list are incredible nonetheless. The guy has played at an extremely high level for a freshman, and has a bright future both at Villanova and (likely) beyond.

What he does best is generate steals and finish easy buckets in transition - this certainly props up his efficiency. His outside shot has started to look more confident and efficient during conference play, however, and he's flashing signs of a deadly dribble-drive game (three steps from the 3 point line to the basket, anyone??) every other game or so. He's started to corral his athletic gifts and length into a skilled on-ball defender, and is terrifyingly effective as a weakside help defender at the basket. Durant+T-Mac+Rondo+Lillard = Bridges, indeed.

Bridges' 2015-2016 Shot Chart

On a similar note, how have the tiny lineups looked so far?? As anyone who follows the general thread of my comments knows, I've been intrigued by the possibility of super small lineups (featuring da STUD Mikal Bridges at center) as a potential look for this team, plagued as they are by a lack of true backup size. The returns have been small in sample, but intriguing.

The offense has certainly been goosed by the ultra small look, jumping in effectiveness by 13 points/100 possessions over the 93 possessions (and ~60 minutes) Jay's deigned to leave Bridges on the floor without the cover of Reynolds or Ochefu. But the defense and rebounding have also clearly suffered - the 2-point conversion rate jumps 6% when Bridges stands as the lone rim protector, and the 3 point percentage surprisingly climbs by an astounding 12.1%. Meanwhile, the opposing team rebounds 36.7% of its misses when Bridges stands as the center - not a killer rate, but certainly not helping. My somewhat educated guess is that the 3 point percentage is a bit noisy, due to the sample size - there's no logical reason a defense with 5 perimeter defenders should be that  bad at defending the 3 - and the problems with the 2 point defense are more real. Of course, the poor perimeter defending could point to an overeagerness to prevent dribble penetration with no real protection waiting on the back end, and could be there to stay - it's tough to really know.

At this point, lineups with Reynolds at center have been more effective than those with Bridges at center - but not by a large margin. I personally remain of the mind that Jay could use small doses of the crazy-small lineup to throw a different look at teams, with a manic pressing style (on defense) and a willingness to get out and run, with a seriously spaced floor, on the other end. We'll see if Jay experiments for extended periods - seems unlikely at this late point, but I'm staying hopeful.

The Emergency 'Chief Banged His Head' Section

As this article had been drafted prior to the reporting of Chief's concussion, I had concluded the above-section right there. I'll add in a quick look at lineups with Reynolds at center, and what the team's missed in minutes without Chief so far this year.

Despite the certainty of the eye test suggesting Darryl Reynolds is not quite as good as Chief (to put it far more nicely than some have!), the team has honestly been just about as good with Reynolds swapped in for Ochefu - by the numbers. The offense clearly falls off a bit - down by 8 points per 100 possessions - though that's mainly related to a big drop in 3 point percentage during those minutes - the team shoots about 7 percentage points better from distance when 'Reynolds sans Chief' is not on the floor. While watching the game today, I pondered a bit on why the offense looks so much worse without Chief, and I couldn't come up with a great answer. Could be a product, as the announcers suggested, of teams helping hard off Reynolds to double on the perimeter - though it's not like Ochefu is much more of a threat as a jumpshooter. Could be that the Chief sets much smarter screens, and moves more intelligently on the offensive end. But I don't know/can't prove, and anecdotally (at least today), the guards themselves just looked more sluggish and dumber with the ball. Need to be much better than that with the big guy out.

And, a peek at the on-off numbers for Ochefu.

His lineups kill it on defense, and from 3. The 1st is obvious; the second, maybe not as much. The team certainly looks more than capable surviving time without Ochefu; we'll see if they can survive entire games. I'm optimistic, but also (like everyone) want him back as soon as he's healthy.

Spreading Those Minutes Around...And The Trip PG

Jay's been generally VERY liberal with the spread of minutes so far this year - only one five man group has logged more than approximately 35 minutes through 20 games this year: the starters, of course, at about 140 minutes. And even then, only 6 other groups have been together for more than 40 possessions. Jay's been quite experimental with his combinations, relying on a surfeit of quality rotation players more than any one combination of bodies that works. The most effective of these groups has (sigh) been the triple PG look, with Hart and Ochefu as the ‘bigs.'

Lineups	Search Score	PPP Differential	Off Pos
Arcidiacono, Ryan--Brunson, Jalen--Hart, Josh--Jenkins, Kris--Ochefu, Daniel----	166	0.37	230
Arcidiacono, Ryan--Booth, Phil--Brunson, Jalen--Hart, Josh--Ochefu, Daniel----	260	0.58	61
Arcidiacono, Ryan--Booth, Phil--Hart, Josh--Jenkins, Kris--Ochefu, Daniel----	191	0.43	57
Arcidiacono, Ryan--Booth, Phil--Bridges, Mikal--Hart, Josh--Ochefu, Daniel----	128	0.29	57
Arcidiacono, Ryan--Brunson, Jalen--Hart, Josh--Jenkins, Kris--Reynolds, Darryl----	45	0.1	57
Arcidiacono, Ryan--Bridges, Mikal--Hart, Josh--Jenkins, Kris--Ochefu, Daniel----	13	0.03	46
Arcidiacono, Ryan--Booth, Phil--Bridges, Mikal--Hart, Josh--Reynolds, Darryl----	90	0.2	43

While I continue to hate the theory behind the triple PG lineup - we have wings capable of handling the ball and shooting, and it guarantees a size disadvantage at 3 positions (we've got a permanent one at the 4) - it's getting harder and harder to argue with the results.

In the minutes with Arch/Booth/Brunson on the floor (vs. those with the group either out or incomplete), the team is 14 points/100 possessions better, mainly fueled by an excellent conversion rate inside (on offense) and a jump in the forced turnover rate on defense. The team's also been much better at generating looks at the free throw line (and not fouling on D) during the same minutes. And, perhaps most surprisingly, the negative differential in rebounding percentage is only about 3% on either end - pretty incredible for a lineup with 3 guys under 6'-3" and (presumably) one traditional big.

While some of these stats (mainly the high free throw rate) could point to this lineup's use in late-game-with-a-lead situations, where foul shooting and being able to beat a press are at a premium, the effectiveness of this lineup might be the thing I was most wrong on this year. It's been very good - though it still continues to scare me come tournament time.

The Evolution of Arch

SPEAKING OF things I was wrong on - Arch!

My biggest problems with Arch's point guard game over the years have been his lack of a dribble-drive game and volume jump shooting at mediocre conversion percentages. Surprise! He's improved in both areas this year.

Here's a look at this shot chart from last year vs. this year:

Arch's 2014-2015 Shot Chart

Arch's 2015-2016 Shot Chart

Arch's 29 conversions on ‘layups, dunks, or tip-ins' (generally, at the rim) through 20 games this year are already more than he logged through the entirety of his first two years in a Villanova uniform. At his current pace - assuming (conservatively! conservatively!) Villanova will play 15 more games - he will score more buckets at the basket this season than he had in his career through his junior year. So what's he doing better this year?

His ‘post game' has certainly gotten the press, but that's only been a minor part of his improvement. He's gotten significantly better (or, at least, the team has gotten better at delivering) at cutting to the basket - as a primarily perimeter-oriented player, he can really catch his defender off-guard with timely backdoor cuts - and he's done a great job of it so far. In addition to scoring a bit more on those, he's gotten smarter about leveraging his defender's (typically correct) assumption that he won't drive toward the basket into easy buckets at the rim. The play used at Creighton last year (and the end of the 2^nd Seton Hall game this one), where he pauses his dribble momentum for a fake DHO on the right wing then drives the baseline for a layup, has been used in more than just that situation this year, to excellent effect. He's even started to pass up PUJITs for higher quality looks in transition - taking 38.5% of his shots in transition at the rim this year marks the first time he's been above 20% (in the same measure) during his 4 years at ‘Nova. I love it!

And finally, his 86 assists so far (and 26.3% assist rate - the highest he's logged at Villanova) put him on pace to finish with the most he's ever logged at the school, by a pretty significant margin. Perhaps even more encouragingly, nearly half of those assists are coming on shots at the rim - Arch had never cracked 40% in his first 3 years. Shots at the rim have the highest expected value - dishing for those attempts is a great thing for the team's offense, and Arch has been creating them at a high level so far.

All told, he's putting together his finest season as a point guard despite the fact he's nominally playing primarily as an off/combo guard in most lineups this year.

Considering the main things he's improved on this year are his distribution and ability to attack the basket, perhaps I should be taking credit for his improvement. As I have said, getting to the basket (and distributing from drives) are the things I look for most in a point guard, especially for Villanova's offense - and it's what he hadn't been doing before! Man, I'm good.

You're welcome, guys.

Wing U

And, here's a sneak peek at how ‘Wing U' may fare next year - though Paschall, Spellman, Divincenzo, and others will be available, the trio of Hart-Bridges-Jenkins will form the veteran core of the evolutionary version of ‘Guard U' - lineups of interchangeable wing pieces standing 6'-4" or more. Man, I'm so excited - with apologies to the excellent team currently plying their trade this year. Love you, too.

The 3-man lineup of Bridges-Hart-Jenkins has held the opposition to a 25% offensive rebounding rate - nearly 5 percentage points better than the team in minutes without this trio. While there have been some minor struggles offensively and defensively within this group, the differences are not enormous from the minutes without them.

It's tough to call this breakdown ‘encouraging,' but it certainly points toward the sustainability of putting more wing size (and less true PG handles) on the floor, and is another intriguing combination Jay Wright can think of stringing together more on the floor. And as a final filler, here's the numbers with this trio and the only big man on the team that will remain around next year (Reynolds).

This group, in a very limited sample, has been crushing the opposition on defense and on the boards. The small sample obviously prevents any sweeping conclusions, but it's not a bad data point as we look toward the type of success next year's squad may be able to have. Make sure you stick around for it, Josh!

Wrapping Up

Jimdribbleless - unfortunately, there's no way to determine what defense the team is in at any given moment on the floor (short of charting each possession by hand), so I was unable to look at the effectiveness of the zone so far this year compared to the man scheme (for Villanova).

And, for KenPom's POY rankings - Billy Vinci provided the link to the full explanation here earlier - but I'll work in a small paraphrasing of the same. It considers the offensive efficiency (measured by offensive rating, a metric delineated by Dean Oliver in 'Basketball on Paper') and usage (the portion of a team's possessions a player uses while on the floor) of a player, and the individual 'defensive rating' (another metric introduced by Dean Oliver, that tries to translate the limited box score stats that measure defense - rebounds, steals, and blocks - into 'stops'). Then, it adjusts the result of those measures of a players' individual game and weights them by how good his team has been that year. Per Ken's article, it'd be nearly impossible for a player on a team outside the top 20 to win it. As far as I'm aware, the exact formula itself has not been released.

Just for the additional info/visual - the team shot charts, on offense and defense, from the year so far:

Villanova 2015-2016 Defensive Shot Chart

Villanova 2015-2016 Offensive Shot Chart

Discuss amongst yourselves.

That about ties up the notes I put together for the season so far. There are other things to investigate and write about, and it may happen again soon, or closer to tournament time - as always, let me know if there are specific things you're interested in,and I'll let you know if it's something I can look into with the data available. Thanks for reading.