Villanova Basketball Advanced Statistics: Breaking Down Transition Basketball

Editor's Note: Corollary has joined the VU Hoops staff. He's out of the FanPosts and on to the front page full time. Congrats and happy to have you on the team!

It starts off like just any another play. Seton Hall slowly brings the ball up the court, swinging it around the wing, searching for an opening. About 20 seconds into the possession, Fuquan Edwin tries to take it past James Bell, and dribbles off his foot. Arch dives in for the ball, forcing Edwin to pull it back and try to pass it off. Sensing the opportunity, Darrun Hilliard jumps the passing lane, snags the lazy pass, and riding the loud and gradually increasing sound of the crowd, dribbles the length of the court for an easy lay-in.

It's just one easy bucket in the larger narrative of the game - a game Villanova would eventually end up losing. It can't mean that much.

It can seem like the importance and role of the transition game, on both ends, gets short shrift when it comes to analyzing the game. Announcers, pundits, and commenters might remark on fast break points, or maybe manufacture some version of the phrase ‘defense creating offense,' and move on. The value of the transition basket is often lost in the maelstrom of the game - and there's plenty to be seen behind the curtain.

Shaka Smart knows it well. He and his VCU team have quickly shot to national prominence on the back of their signature ‘Havoc' press and the onslaught of transition and 3 point attempts it produces. Since taking his 'Cinderella' team from the First Four to the Final Four in his second year coaching, Smart's teams have led the country in both steals and forced turnovers in each of the last two years. VCU's success in forcing turnovers led them to the top of the charts in transition attempts (as a percentage of overall shots) - in 2012-2013, they were at 17th overall; this year, they ranked 2nd, taking 32.70% of their shots on the break. For comparison's sake, the median team took 20.90% of its shots in transition this past year.

So why is this a good thing? Well, it hardly takes a seasoned basketball observer to see the benefits of shots generated off steals. More often than not, the team that gets a steal has numbers and a clear path to the basket for an easy layup or dunk - they lead to some of the easiest points in basketball.

But steals are not the only way to generate opportunities on the break. Pushing the ball after a long rebound, or even starting quickly off a made basket, can let an offense attack the defense before it has a chance to get set, potentially opening up driving lanes, creating mismatches, and forcing bad decisions. Transition possessions result in a significantly higher percentage of shots at the rim, the most efficient one in basketball, and are converted at a rate much higher than the average attempt at the rim. These chances also create a number of free throw opportunities, again at a significantly higher rate than any other type of possession. The benefits of these chances are strong enough for entire philosophies of team-building to be constructed around them.

Very successful teams have been built on the principle of pursuing transition baskets; the aforementioned VCU Rams and the 2013 national champion, Louisville, among many other college teams, as well as a number of NBA teams. The Phoenix Suns of the mid-to-late 2000's made 3 conference finals in 6 years, and finished as one of the top 2 offenses in the NBA in every year from 2005 to 2010 on the back of an offensive philosophy designed to push the ball on each possession and find a shot within the first seven seconds up the floor - the 'Seven Second or Less' era. The Denver Nuggets, in 2013, built a team around no stars and the value of the transition game (along with the thin air and home court advantage Denver provides), pushing the ball up the floor, in towards the rim, and out for 3 pointers at every opportunity. They posted the highest win total in franchise history - although they eventually fell in the first round of the playoffs.

The value of the transition game isn't necessarily in the extremes, though the success of teams who build their philosophy around pursuing and creating transition baskets helps highlight the merits. Offensive chances in transition - and how well the defense defends and prevents them - are often integral to the success of all teams. Especially teams who have trouble producing shots at the rim in the halfcourt.

As with most of the other articles I've written, this will be an analysis of the numbers behind one of the many aspects of the game. If it's not your type of thing, feel free to jump around, skip ahead, or just skip over - if you stopped right now and said ‘Of course I know fast break points are easier to come by than half court points- I've been watching basketball longer than this dude. Plus, what an asshole' - you'd hit most of the major themes of the article. There's a lot of interesting things in the numbers, though - and not always what you'd expect.

Methodology

For the entirety of this article, a ‘transition' possession will be defined as the actions in a possession that occur within 10 seconds of the change of possession. It's not a perfect method to define transition possessions, but it's the best way to do it based on simply play-by-play data, rather than reviewing each possession via video.

The noise and potential inaccuracies you can expect in the data mostly stem from free throw rates and turnovers. I did not filter out free throw attempts in the last 2 minutes of close games Villanova was winning - obviously, a lot of intentional fouling in those situations will happen within 10 seconds of a possession change. Additionally, turnovers that happen within the first 10 seconds won't always be a true ‘transition' possession, as they'll include steals off inbounds passes, or from early presses, and the like. Generally, these two factors will be ignored as much as possible in the analysis - though points from free throws, and possessions lost to turnovers must be included in calculations of offensive rating.

Play-by-play data for 33 of Villanova's 34 games this year (Lafayette is yet again the missing culprit) was gathered from ESPN's publicly available database. Using this play-by-play data, I found the ‘end' of each possession - and subsequent beginning of the opponent's next one - filtering out the trickier definitions, such as possessions ending in an and-1, or a missed free throw leading to an offensive rebound. This allows the establishment of a game time for the beginning and end of every possession in the game. Each action in the game - made shots, missed shots, assists, turnovers, rebounds, free throws, etc. - was then defined by its relationship to the game time at the beginning of the possession. Anything that happened within the first 10 seconds of a CHANGE in possession was defined as ‘transition,' while anything after that first 10 seconds was defined as ‘half-court.' Missed or blocked shots that resulted in offensive rebounds simply resulted in the continuation of a possession, and not as an entirely new one.

These possessions, after being split into ‘half-court' and ‘transition' possessions, were further broken down by how the possession started. Possessions that started with a made basket, missed basket, steal/defensively recovered block, and a dead ball turnover were further separated by their respective category. The same thing was done for Villanova's opponent in each of these games. How Villanova performed, on both offense and defense, was then analyzed for each of these cases.

I also went through the same data to parse out what each major rotation Villanova player did in transition and the half-court. Finally, Villanova's performance on both offense and defense was split up, for all of the possession types - as described above - by the time from the start of the possession. The splits were done on 5-second intervals - from 0-5 seconds, 6-10 seconds, 11-15 seconds, and so on to 31+ seconds.

Some overall rankings and stats, based on a table including every team's stats from the last 3 years, will be included in the overall analysis to point out some of the positive aspects of transition offense and defense, and some other surprising factors. Data for these was gathered from hoop-math.com and kenpom.com.

The data work in this article owes heavily to what Texas alum and just-a-slight-bit-more-advanced-stat guy Jeff Haley, who writes for Burnt Orange Nation and runs his own stat site, hoop-math.com, both generates (in terms of stats) and writes. The numbers he posts on his own site will be used, in addition to the things I did. There will be minor differences in shooting stats between his and my own, in terms of ‘transition' and ‘half-court' defined attempts and shots. These differences owe to the way he defines transition shots - only INITIAL shots within the first 10 seconds of the shot clock count - and the fact Lafayette will be included in his data set. Possessions after a block+offensive rebound are also included, in my data, in the same category as steals.

Most of the ‘advanced statistics' used here - though generally, they're simply an arithmetic crunch of your typical counting stats - have been utilized before in my FanPosts. This first one has a relatively complete rundown of everything that'll be used here - odds are, though, if you've read this far, you already know the type of stats I generally use.

Though I wrote this as one gigantic block of article, I've split it into a few easier-to-digest sections that'll be periodically released through the next week or so.  This first section will focus on the overall value of transition chances vs. half-court, and a few nuances of the differences between transition chances.

Transition Offense - An Overall Picture

Quick Summary

• Attempts in transition are taken at the rim far more often than shots in the half-court
• All types of 2 pointers are hit at markedly higher percentages during transition.
• The percentage of 3s taken and made is largely unchanged between transition and the half-court.
• Villanova's transition offense accounted for less than a third of its total attempts at the rim - but over 37% of its made shots at the rim.
• ‘Nova takes way too many 3s in transition, and doesn't hit enough of them.

Simply watching a couple basketball games would be enough to tell you that shots in transition, especially off missed baskets or steals, seem to end in easy baskets at the rim far more often than your average transition possession. Quantifying this, though, helps give some really strong indicators of the positive value of a shot in transition.

Using data gathered by hoop-math.com, the shot selection profiles of a transition attempt and a half-court attempt - as an aggregate of every NCAA team in Division I - are easily broken down. Check out this table:

Those who've read before know I've already spent thousands of words talking about shot selection profiles, but indulge me for just a couple hundred more. The essential shift shown by this table, in terms of transition vs. half-court, is that an additional 15% or so of the shots taken in transition are at the rim - the most effective shot in basketball (shots at the rim, in play-by-play data, are listed as dunks, layups, or tip-ins) - with a corresponding 15% drop in shots taken from 2 point jumper range - the least efficient shot in basketball. Not only that, but there are significant jumps in FG% at the rim (almost 7 percent in transition) AND 2 point jumper range (just over 1 percent). Interestingly enough, the 3-point shot - in terms of how many are taken, and the percentage at which they're hit - stays pretty constant

These improvements, in the efficiency of both the type of shot taken and how well it's hit, combine to provide an increase of almost 7 percentage points in effective field goal %. The difference between the two is vast. Illustrated in terms of expected points: 100 shots taken in transition, at the average percentages in terms of distribution and makes, would net a team 110 points, while 100 shots taken in the half-court, applying the same exercise, would net a team 97 points. While, obviously, possessions don't always end in simply a made or missed shot, and this rough calculation ignores any potential effect transition attempts have on TO% or offensive rebounding, the 13 point/ 100 possession swing is the equivalent of moving from the 67th ranked offense this year to the 299th (per KenPom).

In short, transition attempts can be very, very good for your team's offense - including Villanova. Using the play-by-play data and methodology referenced above, Villanova's splits in transition and the half court were generated. Check out their distribution chart:

Interestingly enough, Villanova's offense was equally effective this year in the half-court and transition, as judged by the tempo-adjusted Offensive Rating, rating out around 1.14 points per possession (unadjusted for the strength of the opponent's defense, as KenPom's stats are). This is mostly influenced by the jump in turnover percentage (4.92% more possessions result in turnovers in transition, for Villanova), and the cratering of their 3-point percentage in transition (they shot 31.50% on attempts in transition, and 37.42% on 3-bombs in the half-court).

One of most important section of the chart covers the first 5-6 columns, though. Villanova, a rim-attempt challenged team for much of the second half of the year (they were better, though still not great, about it during the first half), got to the bucket on 42.23% of its attempts in transition, while converting at an absurdly high rate (74.09%). They also shot the lights out on 2 point jumpers in transition, hitting over 50% on a shot that's normally converted closer to the mid-to-high thirties - it's a relatively small sample (under 50 shots), so it was likely more luck than an absurdly high amount of skill. Still, their hot shooting from jumper range and above-average conversion rate at the rim combined to let them shoot almost 70% (68.48%) from two in transition - an incredibly high rate. Villanova's 143 baskets at the rim in transition accounted for 37% of its total on the year.

On the negative side of those splits, though, is the reality that they're well below average in pushing attempts to the rim in transition (the half-court, too, but that number makes far more sense, with their relative lack of an interior presence). As we saw above, the average team takes about 48.70% of its transition shots at the rim - Villanova managed a measly 42.23%. Especially considering their excellent conversion percentage (the average is around 70% in transition) at the rim, it's not likely personnel is holding them back from more attempts at the rim in transition.

Villanova took FAR too many 3s in transition, especially for the percentage they were converting. While they were an unconventionally dedicated 3 point shooting team last year - over 45% of their total attempts came from distance - much more emphasis should be placed on taking the ball to the rim when transition opportunities arise, considering the respective conversion rates of their shots at the rim and pull-up 3s in transition (74.09% and 31.50%, respectively) as well as their general difficulty/disdain for producing shots at the rim in the half-court. There's simply no reason to take over 40% (actual number was 43.76%) of your shots from 3 in transition - rim attempts are so much easier to come by and convert before the defense gets set, and come with other juicy benefits (free throws, demoralization, WINNING). This should be an easy coaching point to drill in; although it's unlikely Jay will pull the go-ahead signal, he should be able to emphasize how much easier and better going at the rim is in transition (looking at you, Arch. Hilliard and Hart, too, though to a lesser extent). Bringing that percentage of 3s taken down (in transition, mostly) to about 35%, while turning the rest of the looks into rim attempts (or close 2 point jumpers) would greatly benefit the transition offense.

Effects of Possession 'Start Type'

Quick Summary

• The way a transition chance starts (steal, block, defensive rebound, opponent's made basket) has a large effect on its effectiveness.
• All transition possession types are more effective than their half court counterparts.
• Way too many 3s are taken after an opponent make, with a low percentage converted.
• Offensive rebounding percentages actually go up, on all shot types, with the length of a possession

Now, not all transition chances are created equal. Using some additional data downloads from hoop-math.com, it was easy to slice up the shooting percentages by possession type for all teams over the 2013-2014 season, as defined by Jeff Haley's parameters. It should be noted these field goal attempts will not include offensive putbacks, defined as shots taken within the first four seconds of an offensive rebound. Additionally, the %'s of initial attempts will not (and should not) add up to 100%. The data from hoop-math.com, as far as these splits go, does not include possessions starting off dead ball turnovers or recovered blocks.

Clearly, steals are the Cadillac of transition chances. Chances off steals, in the ‘transition' phase, have the healthiest shot selection profiles and field goal percentages in every single category. If a team took 100 shots at the average percentages for the transition steal category, they'd score 126 points in those 100 shots. This is a large part of the reason team philosophies based on full-court pressure defense and generating chances off steals work in college - they lead to the best shots. Your offense doesn't need to be great in most other facets when it's taking a lot of shots as easy and productive as these, and a well-run press is almost always a great defense.

What is interesting to note is the effective field goal percentage of each type of transition chance, as defined by hoop-math.com, is well above those in the half court. Additionally, each transition category has a minimum of a 6.39% advantage in percentage of shots at the rim over their half-court counterparts. Steals and the offensive chances they can lead to remain the standard, but it's hard to coach ‘get more steals, guys' without sacrificing or changing the defensive principles a team is built on. Rather, coaches/players/the 10 guys reading this article should note there is always some value to be found in aggressively pushing the ball up the court, searching for an easy shot at the rim or exploitable gap in the defense before they're able to get set.

Also interesting enough to point out - though the use of the eye test would certainly suggest it before you run the numbers - is that there are few functional differences between different possession ‘start' types (rebounds, steals, makes) once the possession crosses the initial 10 second barrier. The rate and efficiency with which different types of shots are made, and their end result (eFG%) are nearly equal across all categories.

Here are Villanova's numbers, split up by category of possession start and transition/half-court.

One caveat/note when it comes to analysis of overall transition chances is the disproportionate number of 3s (and a frighteningly low conversion rate) that accompanies transition possessions that occur after an opponent make. Across the NCAA this past year, approximately 41.37% of the shots taken in transition after an opponent make were 3s - and only 31.84% of them, nearly 3 percentage points lower than the national average, were made. This can likely be chalked up to a psychological tendency/need to ‘stop' or ‘match' an opponent run after a made basket, when the big brass ones start bouncing around the hero-ball arena. This isn't a new trend, either, as Jeff Haley made a note of it in a previous analysis he had done (on his own stats) for the 2011-2012 season.

This article, diving more in-depth into his own numbers from 2011-2012 regarding transition chances, held a number of interesting conclusions and notes on transition and half-court possessions that you haven't seen here. The not-quite-touched upon-here SparkNotes:

• The way a team takes possession of the ball has a huge effect on the type of shot taken in transition. 70% of all attempted shots in the first 5 seconds after a steal occur at the rim; it drops to 40% for the same after a miss, and 30% after a make. The numbers from 2011-2012 are very similar to those from this year.
• Shooting percentages on each shot type also vary significantly according to the way the possession started. You can see similar effects in the numbers crunched from this year.
• Offensive rebounding percentage, interestingly enough, increases with the duration of the possession. ‘Initial misses at the rim taken between 10 and 15 seconds into the possession are rebounded 40% of the time by the offense, whereas initial misses at the rim taken after 30 seconds of a possession result in offensive rebounds 48% of the time. Similar increases in offensive rebounding percentages occur on missed jump shots.'

That's it for the first part. For the next article, we'll look at Villanova's offense broken down by 5-second shot clock chunks (across the entire season), and the effectiveness of each of Villanova's players in transition and the half-court.

FOR THOSE ASKING:

Here are Villanova's offensive rebounding stats on Transition 3's vs. half-court 3s.  The OReb% is slightly imperfect, as the total number of available rebounds is assumed as missed 3s (usually, it's defensive rebounds plus offensive rebounds).  It's certainly close enough for our purposes, though.

Big jump in offensive rebounding percentage when the 3s are taken in the half-court, which is to be expected based on the numbers we already had.