Friday, August 31, 2007

Inside/Outside According to Pelton

Kevin Pelton originally developed the Inside stat to determine the tendency of any NBA player to play on the wing as opposed to the paint. As Pelton himself described the stat "... It's really amazingly simple - (FTA-3PA)/MIN...(but) not exact...". Applying Pelton's metric to Big East players who logged 32% or more time at their respective positions produced this Top 10 List of Inside players:

Kentrell GransberryUSF65654.2239014922.71
Paul HarrisCuse75854.32172016619.26
Luke HarangodyNDU65851.4290111417.17
Adrian HillRU72661.8169112016.39
Aaron GrayPitt101667.7361016616.34
Roy HibbertGtown97565.7277015315.69
Jeff AdrienUConn102782.5314116215.68
Lamont HamiltonSt. John's88671.22813717515.58
Hasheem ThabeetUConn76261.2121011515.09
Rob KurzNDU86867.82264917914.98

This list contains those old familiar faces -- and in roughly the order I would have expected. But there are, for me anyway, three surprises -- Rob Kurz, Adrien Hill and Paul Harris. These three show up on Pomeroy's list (see my previous inside/outside post, "Inside/Outside According to Pomeroy") too:

  • If Notre Dame produced one name for this list, Rob Kurz is not the one I expected. And for Notre Dame, a program which under Mike Brey's stewardship has not developed a reputation for strong inside play, to produce 2 players for the Inside list is very surprising.
  • Paul Harris not only shows up twice, but he appears prominently on both lists. The multiple listing (Pelton & Pomeroy) re-enforces the preconception that Harris' strongest play is inside the paint.
  • Like Kurz and Harris, Rutger's senior center Adrien Hill appeared on both lists. Of the three Hill was the least publicized. In fact Hill may have been the least publicized big man on either list.

Pelton's top 10 Outside List also contains a few surprises:

Andy RautinsCuse74153.122618824-22.13
Avery PattersonSt. John's90472.630822042-19.69
Frank YoungWVU113095.041526377-16.46
Jamar NutterThe Hall91177.235221871-16.14
Brian McKenziePC42133.71047812-15.68
Colin FallsNDU102780.2314250102-14.41
Melvin BuckleyUSF107088.439122883-13.55
Deonta VaughnCin98981.740721689-12.84
Mike NardiNova96175.130518776-11.55
Dan FitzgeraldMU71975.317212242-11.13

This list contains 6 common names with Pomeroy's top 10 Outside players. The four newcomers, Melvin Buckley (USF), Deonta Vaughn (Cincinnati), Mike Nardi (Nova) and Dan Fitzgerald (Marquette), all contributed more than occasionally to their respective teams last season. Coming from teams that struggled throughout the year, Buckley and Vaughn may have be lesser known players outside their respective fan bases. A few of the surprises (at least to me):

  • Rautins & Patterson appeared in the top 3 of both lists, with Rautins again taking top honors. Neither will play this season. Rautins, who tore an ACL playing for Team Canada, will redshirt as he rehabs post surgery, while Patterson transferred to a DII school in May. I have to think Syracuse and St. John's will miss these guys...
  • Brian McKenzie appears on both Pelton's and Pomeroy's top 10 Outside List. The Friars are touring Italy now and again McKenzie is taking a relatively high proportion of his shots as 3s (8/18 as of Aug. 30).
  • The Hall's Jamar Nutter appeared on both lists as well. The slam on Nutter over the years has been consistency. He is in his fifth year of eligibility and putting together a good run would really help the Hall.
  • Notre Dame contributed two names to Pelton's Inside list but only one name (not surprising that it is Colin Falls) to his Outside List. Brey has cast Notre Dame as a perimeter oriented team, so the sudden prolific inside players is extremely surprising.

Dante was "Mr. Inside" for the Wildcats. The balance of the ranking remained the same, which means Dwayne Anderson remains "Mr. Outside".

Dante Cunningham87968.7189111312.74
Bilal Benn24619.22822910.98
Casiem Drummond13010.2250129.23
Will Sheridan82964.81295626.88
Curtis Sumpter95574.64001411662.62
Scottie Reynolds96275.23531691750.62
Reggie Redding42933.5823825-3.03
Shane Clark77660.618710052-6.19
Mike Nardi96175.130518776-11.55
Dwayne Anderson18914.8584518-14.29

Given what I know of Villanova's roster and play in 2006-07, the ranking order and values relative to each player "feel" pretty accurate.

Wednesday, August 29, 2007

U19 Team: A Four Factors Look at the Team

The team selected for this competition contained a mix of D1 brand-names (Michael Beaseley, Patrick Beverly, Jonny Flynn, Donte` Greene, David Lighty...) and solid, if not All American-level talent (Matt Bouldin, Stephen Curry, Damanian Hollis, Raymar Morgan, Tajuan Porter and Deon Thompson). I have compiled a summary of the tournament data and divided it into two tables (below). The first contains mostly offensive data -- field goals & free throws (attempts & scored), PPWS and points.The second table contains the nuts and bolts...rebounds, assists, steals & blocks.

Team USA
Standard Dev6.337.651.694.277.639.670.1416.47
Standard Dev4.

The news reports (and box scores...) told a story of a roller coaster ride, a couple of laffers followed by a nail biter. The stories and box scores conveyed a some level of inconsistency in Team USA's performance, I decided to apply averages and standard deviations to get some insight.I have included those totals, averages and standard deviations for each category, both for the USA team and their opponents.

Team USA
Standard Dev4.484.
Standard Dev5.295.294.844.185.325.031.694.4

One of the first things that drew my attention was the relatively large standard deviation for both the USA and opposing teams. Team USA scored an average of 93.7 points, but that standard deviation of 16.47 suggests the USA scored well above and below that average. The same inconsistency is noted in Assists and Blocks. For the opposing teams, rebounds (especially offensive rebounds), steals and blocks seemed particularly inconsistent. The average of points scored 93.7 (USA) versus 74.9 for (opponents) had me wondering how the US team lost the gold medal. In reviewing the box scores I decided to break down the opposing teams into a "European" and "non-European" group. The Americans were able to dominate the non-European competition easily due to the large gap in skills (ie offensive rebounding, assists, steals and blocked shots) combined with US athleticism. The American logged whopping 38.3 average margin of victory (MoV) over the non-Europeans (Mali, China, Brazil and Argentina) en route to a 4-0 record versus that competition. The Europeans were a different matter. The USA team did manage a 4-1 mark versus France (2-0), Lithuania and Serbia (1-1), but the MoV was a far most modest 5.3 points. The loss to Serbia ironically was also by 5 points. I divided the games into European and non-European below and use Dean Oliver's Four Factors (plus a few other stats) to decipher Team USA's play.


The differences in eFG% are huge, representing a nearly 18 point swing against the Americans when playing the Europeans (versus the non-Europeans). The Americans were able to pound the ball inside against their non-European competition (71.9% of their FGAs were 2 pointers; their 2FGM was 61.3%) and score consistently (eFG% of 60.3), if not on the first FGA, then surely the second or third...note the American's OReb% was 49.5% -- they rebounded nearly one in two of their misses. The non-Europeans had far less success shooting (64.7% of FGAs were 2FGAs; their 2FGM was 44.1%) and were not especially strong when rebounding their misses. Note the Europeans were only marginally better than their non-European counterparts when rebounding their missed FGs, but considerably better when rebounding defensively -- the American suffered a steep decline in their OReb% to match their decline in eFG% (49.5 to 33.7 and 63.3 to 51.0 respectively). The Europeans took 42.2% of their FGAs as 3s. And though they hit a modest 33.3% (still better than the Brazilians, Chinese, et. al.), the additional points produced, coupled with their significantly better field goal defense and defensive rebounding, degraded Team USA's offensive efficiency. Coach Wainwright, no doubt in an effort to counter/compensate, sat several here-to-fore successful inside players for wings. Michael Beaseley for example had worked himself into a starting job by the medal round, but after falling behind early to the French and then the Serbs, Wainwright significantly reduced Beaseley's PT in search of a 3 point shooter. Beaseley himself shot 1-10 (0.100) for 3s during the tournament. Wainwright needed a sniper and Beaseley, as versatile as he had been to that point, could not fill that role.

It has been eight years since the USA earned a silver or gold medal in the U19s competition. The achievement cannot be understated. And Jerry Wainwright should be commended for bringing this team through the tournament so well.

Tuesday, August 28, 2007

The Value of Jerel McNeal

Jerel McNeal is one of those players who fills out a boxscore. No matter what stat I track, when I review Marquette's boxscore, I find McNeal almost always has at least one of it. As a member of Marquette's heralded Three Amigos guard class which entered in 2005-06, McNeal has shouldered a major responsibility for playing time, offense and defense. Marquette's fans rave about his style and intensity of play. Which makes me wonder why his possession-based stats, especially those which on the offensive side of the ball, tend to be unremarkable. I have always been curious about his effect on the Warriors since his offensive rating (as computed by Ken Pomeroy), suggest his individual offensive efficiencies are well below those for the team as a whole...

Marquette 105.350.01.06
Jerel McNeal72.430.729.493.445.81.01

The team vs player comparison is only for games in which McNeal played. The comparative field goal efficiency, Offensive Rating and PPWS show McNeal's numbers were well below the performance for the entire team, suggesting McNeal was something of a drag offensively. But how much of a drag? If he were a role player (Min% <40 perhaps with Poss% & Shot% in the 10 - 15 range...) the difference would have been marginal (< 1pt/game most likely...), but McNeal's PT averaged close to 30 min/game, while he tended to take 3 in 10 of the Warrior's possessions and nearly 3 in 10 of their FGAs when he was on the court. Those are more impact player/"star" type numbers, than role player or "regular starter" type numbers. His Poss% and Shot% (combined with his Min% numbers) suggest that had he shot to the team average for eFG, the Warriors would have average another 2 points per game, the margin of defeat in their second game with Louisville on February 17. But how about those less tangible contributions that McNeal might make to the Warriors offense. His assist rate, turnover rate and offensive rebounding for example. Those numbers may infer positive contributions to the offense that would not show directly using shooting & scoring statistics. The next table compares Assist Rates, Turnover Rates Offensive Rebounding and FTRate for both the Warrior team and McNeal.


Marquette did not have the reputation for being a good rebounding team -- McNeal's OReb% is consistent with a backcourt player, possibly a little better than average for a guard. His turnover rate was above the team's rate, suggesting that his tendency to lose 1 in 4 of his possessions was another efficiency drag on offense (you can't score if you lose the ball...). He did however get to line more than the average Big East guard. And He supplemented his field goals with (relative to his teammates) frequent trips to the line. His free throw efficiency was 0.682 (88-129), better than the team average of 0.669 (524-783). His assist rate was Top 500-level (per Pomeroy, he ranks #128), which when combined with Dominic James' ARate (31.5 #53), provides the Warriors with a dynamic offense capable of exploiting man and zone defenses.

If McNeal's offense is (aside from assists and free throws) more a liability than an asset, why is he playing? According to Marquette fans his defense dazzles. That sets him apart from the average Big East backcourt player, and more than justifies his presence on the court. His Stl% (steals per 100 possessions) is indeed eye catching. Pomeroy ranked him #6 in D1 for steals per 100 possessions. While a few might discuss the relative value of a steal ("Does more steals=better Def Rating?" over at APBRmetrics is an example -- a consensus there is of the opinion that the value and ease diminishes as the level of play raises...) and whether it diminishes or enhances a defense, few can argue that McNeal is not very good at it.

His season-ending injury provides an opportunity to see how the Marquette team functioned in his absence. The results were mixed. The team went 2-2, winning the next two games (against Pittsburgh to end the Big East regular season and St. John's in the quarterfinals of the BET), but losing the last two games (a return engagement against Pitt to close out the BET and against Coach Cean's former mentor, Tom Izzo and Michigan State in the first round of the NCAAs) to close out the season. Plotting the offensive ratings of McNeal and Marquette over the last 10 weeks of the season (see chart below) indicates Marquette's offensive rating did indeed increase in McNeal's absence (about a point). David Cubillan and Wes Matthews' ORtg%s rose as they received (and converted) more scoring opportunities in McNeal's absence, suggesting he was probably a 2+ per game point drag on their offense.

Inside/Outside According to Pomeroy

Ken Pomeroy introduced a stat which he called "Inside" as part of his Big 12 2006-07 Preview. The formula, =100*(FTA-3FGA)/FGA, quantifies the common elements of perimeter-oriented play on the collegiate level...

  • A ratio of 3 point to 2 point FGAs.
  • Lack of contact & resulting fouls. Offensive schemes which feature close in scoring via penetration or passing into the low post often result in a foul and FTA as the possession-ending action.
The larger the positive value the greater that player's tendency to play "inside", while the larger the negative number, the greater the tendency to play on the perimeter (the "outside"). Using playing time (about 32% minimum) as a filter, the ten "most inside" players were:

Hasheem ThabeetUconn762121011595
Hamady N'DiayeRU3875604275
Adrian HillRU726169112070
Paul Harris'Cuse7582172016667
Darryl Watkins'Cuse961180011664
Zach HilleslandNDU62812107663
Kentrell GransberryUSF656239014962
David PadgettLville807186111360
Dante CunninghamNova879189111359
Rob KurzNDU8682264917958

Pomeroy had suggested that for regular players the range should be about +70...0...-70. Thabeet and N'Diaye fall outside of that range. Neither however, played a significant role their team's offense. N'Diaye's Poss% & Shot% were 15.2 and 11.1 respectively, suggesting his biggest contributions were on the defensive side of the ball. Thabeet's numbers, 15.0 & 11.0, respectively are strangely similar to N'Diaye's, and show that he is not a focal point for the Husky offense. Aside from the range the names and reputations are consistent with frontcourt players -- David Padgett, Darryl Watkins, Dante Cunningham and Kentrell Gransberry. Two elements did surprise me though. Rob Kurz, Zach Hillesland appeared among these 10, both are from Notre Dame, not a program with a big frontcourt reputation. Paul Harris and Rob Kurz are decidedly not center, bf-type players. But the formula suggests they are more "inside oriented" than I might have thought. I also noticed that Roy Hibbert, Luke Harangody, Aaron Gray & Terrence Roberts, players who were large contributors to their team's offenses are not in this list.

Andy Rautins'Cuse74122618824-73
Brian McKenziePC4211047812-63
Avery PattersonSt.John's90430822042-58
David CubillanMU70214211137-52
Colin FallsNDU1027314250102-47
Dan FitzgeraldMU71917212242-47
Ronald RamonPitt88420916468-46
Frank YoungWVU113041526377-45
Alex RuoffWVU117130319870-42
Jamar Nutterthe Hall91135221871-42

"Mr. Outside", Andy Rautins, also falls slightly outside of the range identified by Ken Pomeroy. Unlike N'Diaye & Thabeet however Rautins's numbers, 17.1 and 21.1, do represent more than an occasional contributor to the Orange offense. Finding two players from John Beilein's system (Alex Ruoff & Frank Young) is not surprising. Finding Providence's Brian McKenzie on the list is. McKenzie, a rising freshman is another unheralded backcourt find for HC Tim Welsh. David Cubillan and Dan Fitzgerald of Marquette provided the wing dimension in HC Tom Crean's offense. Fitzgerald played the #4 (Steve Novak's role for the Warriors over the past 4 seasons), appeared in all 33 games and started 11. Cubillan was an all-purpose backcourt backup who took additional minutes when Jerel McNeal broke his hand. If the list is to be trusted, both Andy Rautins (torn ACL) and Avery Patterson (transfer) will be missed this season. The Villanovans:

Bilal Benn2462822996
Dante Cunningham879189111359
Casiem Drummond1302501248
Will Sheridan82912956244
Curtis Sumpter9554001411666
Scottie Reynolds9623531691752
Reggie Redding429823825-16
Shane Clark77618710052-26
Mike Nardi96130518776-36
Dwayne Anderson189584518-47

The order of the players may be a little surprising, but as a classification, each player appears to be "typed" (Inside/Outside) properly. Bilal Benn's ranking is no doubt due to the distortion of marginal play.

Pomeroy credited "Inside" to Kevin Pelton a Sonics sportswriter and moderator for APBRmetrics. Pomeroy explained that he modified the stat by using field goal attempts in place of minutes played. After reviewing the Top 10 lists, I was curious to see the differences if I used Pelton's original formula.

Sunday, August 26, 2007

U19 Team: The Players

The team selected this time showed a mix of all-star (Michael Beaseley, Patrick Beverly, Jonny Flynn, Donte` Greene, David Lighty...) and solid, if not household-name-familiar talent (Matt Bouldin, Stephen Curry, Damanian Hollis, Raymar Morgan, Tajuan Porter and Deon Thompson). Player summary and per game stats, sorted by minutes played, are given below for the nine game tournament.

Patrick Beverly25094379133330461824117
Jonny Flynn218931667202446172286
David Lighty213931455112634122579
Raymar Morgan188928695192350223083
Deon Thompson18694071004071101690
Stephen Curry1619276012341526192485
Michael Beasley1539498911048791218111
Matt Bouldin118914355159203336
Tajuan Porter115917499318186849
DeAndre Jordan789213000213081650
Donte Greene688102751551291234
Damanian Hollis526916247123423

The table above contains statistics for scoring during the tournament, while the table below records the rebounds, assists, turnovers, personal fouls, steal and blocks. The table below is also sorted by minutes played.

Patrick Beverly143448291213231
Jonny Flynn21820251713022
David Lighty1520357162317
Raymar Morgan1524399173029
Deon Thompson243155618201314
Stephen Curry62834201426022
Michael Beasley20284842320414
Matt Bouldin81321762202
Tajuan Porter3581112506
DeAndre Jordan1113240131784
Donte Greene481236823
Damanian Hollis167011214

The mix seemed to work well through the pool play, but ran into problems during the medal round. Wainwright had a nucleus of Patrick Beverly, Jonny Flynn, David Lighty, Raymar Morgan and Deon Thompson, a group whose Min% ranged from a high of 69.4 (Beverly) down to 51.7 (Deon Thompson). Behind this group was a "regular" rotation that included Michael Beasley (42.5) and Stephen Curry (44.7). These seven saw most, but not necessarily all, of the minutes. All 12 players had minutes, and all but Greene and Hollis appeared in every tournament game. But among those 5 the Min% varied from 32.8 (Matt Bouldin) down to 14.4 (Damanian Hollis). The table below shows possession-based stats (except PPWS) for the individual players.

Patrick Beverly69.417.617.9139.762.71.2930.422.8
Jonny Flynn60.618.117.1118.752.31.1233.325.8
David Lighty59.214.812.0118.374.41.3955.626.7
Raymar Morgan52.222.320.895.144.21.0043.531.9
Deon Thompson51.720.221.6110.556.31.1522.514.1
Stephen Curry44.722.521.1122.455.01.1940.031.7
Michael Beasley42.532.232.9101.855.61.1420.213.5
Matt Bouldin32.813.616.8108.847.10.998.68.6
Tajuan Porter31.920.724.1102.243.90.9316.312.2
DeAndre Jordan21.723.821.8117.170.01.3353.326.7
Donte Greene18.922.422.5104.946.31.0444.433.3
Damanian Hollis14.412.317.4149.462.51.2825.018.8
Team Credited0.70.7

The table above contains most of the offensive/scoring statistics calculated on individual possessions. The statistics are the standard ones used for analysis by Dean Oliver and Ken Pomeroy. The Offensive Rating (ORtg%) measures a player's offensive efficiency by calculating the number of points scored per 100 possessions. PPWS, developed as the "True Shot" statistic by John Hollinger and modified for the college game by John Gasaway, estimates the number of points scored (by field goal and free throw) whenever the player takes a field goal attempt.

Michael Beasley3.922.314.319.2
Patrick Beverly31.4136.114.2
Matt Bouldin20.417.57.411.5
Stephen Curry26.318.44.118.2
Jonny Flynn30.
Donte Greene9.218.46.412.3
Damanian Hollis07.02.112.1
DeAndre Jordan033.015.417.4
David Lighty10.624.27.79.8
Raymar Morgan10.219.38.713.4
Tajuan Porter21.923.92.94.6
Deon Thompson7.622.814.117.4

Players who had a good tournament included:

  • Patrick Beverly -- started all but the first game of the tournament, and played 30+ minutes in those last eight games, giving him a team high Min% of 69.4. Beverly finished with a PPWS of 1.29 (very good) and an eFG of 62.7 (sensational). This rising freshman will kill for Arkansas next season. He hit 39.4% of his 3s (equivalent to hitting 59.1 of his 2s...), and 65.2% of his 2s. And the kid knows how to give it up; he averaged 3.2 assists per game and had an assist-to-turnover ratio of 2.42:1 (not a typo). He also had a very respectable %DReb of 14.2. This from a guard. Razorback John Pelphrey has a terrific team quarterback and scorer next season.
  • David Lighty -- may have started all 9 games, but his playing time was limited during the early tournament games. Coach Wainwright increased his playing time over the course of the tournament. While he did not shoot often (he averaged about 5 FGAs per game -- also note his Shot% was 12.0, suggesting he did not have even a "role-player" level status in the team's offense.) he was deadly accurate when he did. His PPWS was 1.39; his eFG% was 74.4. Lighty shot an amazing 45.5% on his 3s (the equivalent of 68.3 on 2s), translating into an ORtg% of 118.3. But he attempted a paltry 1.2 3FGAs per game. That must have driven Wainwright crazy. The Americans badly needed a prolific scorer (with Lighty's accuracy) in their second game with the Serbs. But Lighty responded with an accurate but paltry 3-4, 1-2, 2-2 9 performance. He will be a very good complement scorer to Kostos Koufu next season at Ohio State.
  • Stephen Curry -- played largely in the rotation off the bench. While he started no games and played, more often than not <20 minutes in a game, he earned about 44.7 of the Min% at his position. His PPWS was a very respectable 1.19 while his eFG% was 55.0. His DReb% was a very strong 18.2.
  • Deon Thompson - who showed some accuracy from his frontcourt position. Thompson posted a very respectable PPWS of 1.15 and an eFG% of 56.3, all without benefit of a 3 point shot. His OReb% was a very strong 14.2 (anything over 10.0 excellent) and his DReb% is a respectable 17.4. He managed to start 4 games over the course of the tournament. He averaged 1.5 blocks per game.
  • Jonny Flynn -- the second of two members from Syracuse (incoming freshmen) had a respectable PPWS of 1.12 and an eFG% of 52.3. His 118.7 ORtg% was very strong -- he would be ranked in the Top 500 next season if he posts a rating like that while playing for the Orange. Coach Wainwright played Flynn to the tune of 60.6 of his %Min, so the Big East coach clearly thought Flynn had something to contribute. Flynn did appear to have his best days against the weaker competition like Mali (7 assists) and China (14 points on 5-8 shooting, 3 assists) and Brazil (10 points on 4-4 shooting). In fairness Jonny's single best game came against the French (1st game) in which he scored 18 points on a 7-14 shooting effort that also featured 4 defensive rebounds and 5 steals. But overall he also seemed to disappear, a problem most especially felt in the last two games where he shot 3-12, 0-4, compiling 6 assists and 5 turnovers (1.2:1) with a single rebound.
Several of the more recognizable names did have a less than spectacular tournament:
  • Michael Beasely -- was the second highest scorer on the team, but scored with considerably less efficiency than Flynn, Beverly & Lighty.
  • Donte` Greene -- a non-factor for much of the tournament. Coach Wainwright kept him out of the second game (versus China) due to injury. He played double digit minutes in only 3 of the remaining games (no more than 14 minutes -- that against France in their first tilt). His PPWS was a (for this tournament) modest 1.04, while his eFG% was a disappointing 46.3.
  • DeAndre Jordan -- Certainly not an undiluted disaster as several others experienced. Jordan managed to start 2 games (Mali & the first Serbia game) and play for 18 minutes in a third one (vs. Brazil). But as the USA team went into the medal round Jordan's minutes were cut drastically. While his ability to score was not in doubt, his ability to hold onto the ball was. A turnover rate of 33.0 is very large for even a frontcourt player who handles a lot of passes into the paint. We will see how well he does at TAMU next season.
  • Damanian Hollis -- saw very little action. He played a total of 52 during the nine game tournament, appearing in only 6 games. The first game he did not play carried the notation DNP - Coach's Decision. Hollis was shooting pretty decently when he played. He just did not play.

All in all a very successful tournment. Congratulations to Coach Wainwright, his staff and most of all the U19 players.

Friday, August 24, 2007

Recruits: Hoopmaster's Summer Report Card on the Class of 2009

The summer camps and AAU tournaments are winding down and the scouting services are preparing their summer Report Cards. One of the first out is Van Coleman's Hoopmaster ranking. Coleman has been working through the CBS affiliate CSTV for several years now. A Summary of Changes & Highlights is available, while the bare rankings are given here. I have listed those players of particular interest to Villanova below, along with their rank and comments...
Isaiah Armwood12wf#5 wf
Kenny Boynton6sg#3 sg
Dominic Cheek13wf#6 wf
Nolan Dennis22sg#5 sg
Gregory Echenique35c#5 center
Wally Judge24pf#7 pf
Dalton Pepper99wf#27 wf
Dexter Strickland9sg#4 sg
Maalik Wayns29pg#3 pg top ranked Philadelphia player
Most are already known to Villanova fans. Nolan Dennis is rumored to be relocating to New Jersey and St. Patrick's in Elizabeth.

Thursday, August 23, 2007

Pace Notes: Thoughts About the 2006-07 Season

Another season and another decline in the average possessions per game. In his 2005-06 season retrospective, "State of the Game", Ken Pomeroy observed that "...The dramatic slow-down during this time (the 1970's and 80's) prompted the NCAA to experiment with, and eventually add, the shot clock and three-point line. There was an initial increase in tempo in response to the shot clock, but the game has been slowing down since 1994, and reached an all-time low in 2005...". That trend continued into the 2006-07, as the year-end raw pace slowed from 67.6 to 66.9, making the year-over-year average decline another 1.1%. The season started optimistically; the average for games played through November 28 (the first week when all D1 teams played at least one game) was 69.2, well above 2006 ending pace. I had loosely tracked Ken Pomeroy's pace statistics during the 2005-06 season and realized that the average pace for all D1 members had declined by about 2% from New Year's (the start of most conference play) to the end of the season. Taking weekly snapshots of the Pomeroy's efficiency stats I was able to track the raw pace for the entire D1 season (graph below).
D1 Average Possession/game
After starting above 69.0, the raw pace began to decline dramatically as all of the teams began their season with home games and invitationals. The pace took a dead cat bounce during the fall exam period and resumed its decline through the end of the season.

If, as Pomeroy suggested, D1's declining pace is tied to the increasing use of the 3 point shot, pace may make a larger adjustment this season as the 3 point line was moved back a foot for men's play. The question is "Will it rise?" as teams decide to take play inside rather than risk declining 3FGM percentages, or "Will it decline (again, only more drastically)?" as teams work harder (and longer...) to set up the 3 point play. Widening the lane to encourage dribble penetration and enlarging the "no charge" area under the basket might work to speed up play as well. Abolishing zone defenses, as the NBA has attempted to do for the pace three decades will most likely lead to the same Prohibition Era-style cat-and-mouse game that characterizes NBA play today. Teams disguise their match-up defensive play as referees, tired of calling the illegal defense foul that disrupts game flow, ignore all but the most blatant violators.

As I track pace again this season I will also keep in mind the season-long pattern shown in the chart above. When I plotted the average D1 pace at New Years, in late January (or early February -- about mid-season in conference play), at the end of February (the end of the regular season and conference play) and season-end (post NCAA & NIT), I noticed the same pattern in the 2005-06 (blue line in chart above) and 2006-07 seasons (the red line above). The number of possessions declined by between 1 - 2 possessions per game as the college teams moved through that portion of their season. For the 2005-06 season, the points on the trend line were 69.02-->68.32-->67.81-->67.61, while for the 2006-07 season they were 68.34-->67.43-->66.98-->66.85. Reasons for the decline include:

  1. Familiarity between coaches and opposing team members translates into more strategy during play.
  2. The relative importance of conference play motivates coaches to exercise greater control over each possession.
  3. Shorter player rotations diminish disruptions to each teams playing rhythm.
  4. Player fatigue due to the length of the season.

Tuesday, August 14, 2007

Big East Freshmen: What Ever Happened To...?

The Villanovans excepted, the players listed below were typically the first or second listed when recruiting lists were reviewed. Each player was classified anywhere from "impact player" to " major contributors" when mentioned with their team. Their numbers are typical for a D1 freshman, but these were supposed to be the difference-makers. The Villanovans (Drummond and Redding) are included because they are, well, Villanovans (and I was curious to see how they compare to some of their better known counterparts...). The others were supposed to play larger roles for their teams, whether offense, defense or both. They were supposed to be starters (or at least fixtures in the rotation) this past season. Each player's %Poss & %Shot percentages (Caracter excepted), at around 15-17, indicate they were by and large role players in their team's offense, or (if < 13...) no more important than occasional contributors. Other players took responsibility for scoring, even when these players were on the court. These nine were sorted by %ORtg (Offensive Rating -- points per 100 possessions):


And the rebounding, assist, turnover and free throws/field goal attempts stats:

Walker2.18.425.011.312.4Started strong &..
Drummond14.717.046.44.416.8Mid-season surge & then...
Caracter19.613.852.95.622.5Better at end
Robinson8.817.540.39.524.2Peaked in January...
Kelly11.915.853.74.830.6TOs killed
Barwinkel0.05.610.011.919.2Never found niche
N'Diaye9.217.675.03.027.2Good rebounder
Redding5.812.228.411.623.6Moved past sophs in rotn
Rivers2.911.719.613.539.9<--- Ouch!

An explanation for PPWS (points per weighted shot) is provided by the Big Ten Wonk. Ken Pomeroy provides explanations for each of the other columns on his "Individual Stats Primer" page.

Defense Before Offense? - Hamady N'Diaye was Rutger's only pickup in the summer of 2006. And the Scarlet Knights needed (much, much) more than a project going into last season. N'Diaye never broke into the starting lineup, despite Joynes' redshirt (and the obvious need). He did not progress enough during the season to grab more than 38.6% of the PT he drew at the beginning of the season. Adrian Hill (the starting center) has graduated, so N'Diaye will compete with Joynes for PT next season. Looking for a longer learning curve may be reasonable.

Leadership Positions Available, Apply Within - Watching the Connecticut rotation the past two seasons reminded of the times I have watched baseball card collectors analyze a holding for possible acquisition: They flip incessantly through the deck, pausing only very occasionally to check the player or condition of the card. I had the sense they knew what they were looking for; they just hadn't found it yet. It was painfully clear by the middle of January that the returning players (Adrien, Johnson, Austrie and the multitude of bench players...) did not have the confidence of the staff to lead the team. The search for a leader and rallying point followed the path of their search for a scorer. For Kelly the learning curve has proved to be longer than his press clippings suggested. He was nearly invisible during the OOC, accumulating enough PT to meet Ken Pomeroy's 10% requirement for stat tracking only after the Big East regular season started. By early February he had established himself in the rotation for about an 8 minute turn. And that is about where he stood through the end of the season. His numbers suggest a typical freshman trying to make his way (89 %ORtg, a solid eFG - about 55%...since his FGAs were 2s, this was also his FGM percentage). He had a promising PPWS of around 1.07, some very strong offensive rebounding and promising defensive rebounding statistics. He was, however, extremely turnover prone, with a TORate which hovered around 30% during the Big East season. He lost about 3 in 10 of his possessions (due to fouls, having the ball stolen, lost out of bounds, etc.). Stanley Robinson is one of only two players on this particular list who saw his PT (Min%) actually increase as the Big East season progressed. At the end of the OOC Robinson was logging about 37% of the PT at the 3 (behind sophomore Marcus Johnson). After Johnson had an especially disappointing outing versus Marquette (0 points in 11 minutes), Robinson became a starter. His PT increased from about 14.0 mpg to 18.5 mpg over the first month of the Big East season. Despite several promising outings (20 points against Indiana, 13 points against St. John's, 9 points against Providence), he too seemed to slide into a scoring funk. His 2FGM was a disappointing 37.8%, while his 3FGM was only slightly worse 36.6%. With shooting numbers like that (especially his 2FGM) his eFG remained in the low/mid 40's. Despite the disappointing output offensively, Robinson proved a strong rebounder (especially on the offensive boards). Like Kelly and Thabeet, Robinson's off season "To Do List" should include cutting down (significantly) on turnovers. He will also need to develop into a more consistent scorer.

Never Have so Many Made so Much About so Little - Louisville's Derrick Caracter posted some paradoxical numbers to go along with his paradoxical freshman year. Slapped publicly by the coach for an unauthorized interview and sent home before the end of the fall semester, his return just before the beginning of the Big East regular season did not mark a reconciliation with the staff. On the contrary after playing 17 minutes in the Big East regular season opener against Notre Dame (a double digit Louisville loss) he and fellow freshman Earl Clark were bumped from the travel squad as the team journeyed to Tampa for their road opener at South Florida. After a 5 minute run versus Marquette (another loss), he was DNP for the next 6 games. And then played in every game through the end of the season, averaging 14.3 minutes per game. His stats are a testament to his love of offense and scoring. His %OR was 19.8 -- nearly double the average %OR for Big East players, while his eFG% was 55.8 (on 53-93 shooting, 53-92, 0-1) and his PPWS was an impressive 1.13 (if he shot it he would score...). But his overall Offensive Rating (ORtg%) was a strangely low 97.4, due in part to his terrible free throw shooting (27-50, 0.527) and turnovers. His defensive stats, particularly defensive rebounding (DR%) was a lackluster 13.8, below average for a Big East big.

Running Backwards...Jerry Wainwright designated Will Walker the point guard of the future from virtually the moment Walker committed to the Blue Demons. Walker started 13 of the Demons first 18 games and by the beginning of the Big East season he was logging 55.0% of the minutes, had a %ORtg of 121.0 (Top 100), an eFG% of 55.2, a PPWS of 1.19 & a TORate of 10.1% (also Top 100). Compared to eventual ROY Scottie Reynolds' numbers at that point (%Min - 56.1, %ORtg - 109.3, eFG% - 47.0, PPWS - 0.98, TORate - 26.3) Walker looked pretty solid. There were, however a couple of caveats...his ARate was 9.5, shockingly low for a point guard who usually is the team's quarterback, and his %Poss & %Shot were below 15.0 (12.7 & 14.4 respectively), indicating he was not really part of the offense or team flow. His minutes (and the rest of his stats...) imploded after DePaul's second game with St. John's in which he played 16 minutes. After St. John's, Walker did not see double digit minutes again; he recorded 4 DNPs over the course of the rest of the Big East regular season and saw only a minute of play (garbage time) in two other games. He averaged 21.7 mpg during the OOC, but only 9.9 during the Big East regular season. Meanwhile, DePaul signed another pg (Mike Bizoukas) for the 2007-08 season, and Jeremiah Kelly for the 2008-09 season.

Saturday, August 11, 2007

Big East Freshmen: Ten to Watch

These ten rising freshmen were not the headliners on their respective teams. They were not the first option on offense nor the defensive stopper. Bozeman was a starter from virtually the beginning of the season. With exception of Hayward and Calhoun, who worked themselves into their respective team's starting lineups, the others established themselves in the rotation, which explains how virtually all share these two common elements...

  1. Playing time in the range of 20%-59% (Solomon Bozeman excepted) -- the player was not a starter, but in the rotation for most of the season.
  2. %Poss & %Shot numbers may be <19, (but usually >13...) -- the player had a "role" in the offense, but was not the focus of the offense.

While the All Rookie Team members usually have higher numbers in those categories, the common thread for this group is that each has promising numbers with limited playing time. Each of these ten have the potential to assume a larger role for his respective team while simultaneously improving his efficiency. The All-Rookie Team members are going to grab a good many headlines and eyeballs next season, but a few on this list will surprise...


Rebounding, assists per 100 possessions, turnovers per 100 possessions and free throws per 100 FGAs are in the table below...

Solomon Bozeman1.18.887.023.028.3
Qa'rraan Calhoun9.810.442.84.112.5
David Cubillan1.78.526.214.016.4
Larry Davis6.712.229.813.816.4
Jerry Smith6.613.122.916.822.2
Dwain Williams1.73.226.913.921.0
Lazar Hayward12.513.237.13.816.5
Earl Clark9.617.241.94.914.6
Larry Wright3.06.729.89.114.6
Vernon Macklin10.68.242.611.121.0

Pomeroy provides explanations for each of the columns (except PPWS, which you can find at the Big Ten Wonk site) on his "Individual Stats Primer" page.

He was robbed. Compare Solomon Bozeman's numbers to those on the All-Rookie team and the others on this list. He has All Rookie numbers but no recognition. He ought to have a chip on his shoulder, and if he plays next season like he does he should make an All Big East team. USF's Robert McCullum handed Bozeman the ball at the beginning of the season and Bozeman ran the team for nearly 3 months, logging about 33 mpg during South Florida's OOC. When sophomore Chris Howard completed rehab in early February, McCullum worked Howard back into the starting lineup, eventually handing the team back over to the more experienced pg. Bozeman meanwhile went to the bench remaining, nevertheless, in the rotation. That experiment lasted about 5 games. Bozeman worked his way back into the starting lineup, this time at the 2, displacing a freshman wing player, Amu Saaka. Bozeman's eFG is good, but not great and his turnover rate, even for a point guard (who handles the ball a lot), was high; he compensated partially with a Top 500 assist rate (ARate - 23.0) and a knack for getting to the free throw line (FRate - 87.0 -- a Lowry/Krauser type number). His PPWS (1.21) suggests that his accuracy at the free throw line (88.4%) was very solid also. If he took the shot he would score. South Florida, stocked with frontcourt players like Gransberry, Buckley and Mattis last season to which Bozeman deferred, meant he was not a first/second/third option on the offense. South Florida will have a nice set of back court players in Bozeman, Howard, and next year. If Bozeman can get his turnovers down and Heath can find a way to use those guards more effectively than McCullum did, then Bozeman will get a bit more ink.

The All-Purpose Backup Marquette's HC Tom Crean shuffled a number of players through Steve Kovak's (#4) spot in the rotation. Kovak played in the paint for defense (rebounding, etc.) but would flash out to the wing to shoot 3s on offense. Crean's original idea may have been to toggle between Fitzgerald and Hayward, but coming out of preseason practice, it turned out that St. Benedict's David Cubillan got the first look. Cubillan's strengths did not really match the requirements (heigth, build, etc.) and the competition eventually did narrow to the preseason favorites, sophomore Dan Fitzgerald and freshman Lazar Hayward (see table above). But Cubillan proved useful enough to keep in the rotation. His strength, as attested to by his %ORtg, eFG% and PPWS, was clearly his shooting and not (again from his numbers...) his rebounding (OR% & DR%), getting to the line (FTRate) or ability to distribute the ball (ARate). In the James/McNeal focused offense (between them they had 58% of the Warriors' possessions and took 57% of the shots when they were on the court...) however Cubillan was little more than a role player until McNeal's injury. While his role grew very slightly after that point he was not a regular option on offense even at the end (see his %Shot), despite his Top 100 %ORtg.

Another Friar Find? When Sharaud Curry was suspended for four games at the cusp of the OOC and Big East regular season, I wondered if this was the beginning of the end for the Friars' 2006-07 season. To that point they had lost to Brown, but beat Boston College, struggled (but won) against Holy Cross and Harvard, then were routed by Florida, but turned around and hammered Rhode Island. With Curry, their point guard and sparkplug out they had to finish the OOC against Florida State and Longwood and then open against Marquette and Seton Hall. Freshman Dwain Williams took Curry's position, and after a rough start against Florida State (another rout), the Friars ran off 3 straight wins before Williams turned the ball back over to Curry and returned to his place on the bench. Coach Tim Welsh must have liked what he saw, because Williams earned a slightly more active role in the rotation for the next month or so. The wins over Marquette and Seton Hall were significant. Marquette was ranked in the Top 20 at the time and the Pirates and Friars were tabbed to compete for the same ranking within the conference. Williams was a late signing from the Spring of 2006 who had earned some national publicity as a junior out in California and then fell off the radar screen as a senior. HC Tim Welsh found a solid contributor who, despite missing 4 Big East games saw his minutes per game grow after he short stint as a starter. Hopefully Welsh will be able to keep Williams at PC and participating in the basketball program.

The One That Got Away. When Louisville signed Earl Clark the Nova Nation let out a collective groan. The 6-5 point guard from Rahway, NJ had been on the Wildcat's radar since his sophomore year. Fans loved his handle and ability to penetrate and score (but maybe not his 3 point shot...). A diamond in the rough, Villanovans believed he could have joined the 'Cats and would have seen regular minutes at the 2 or 3. Instead he became the "...and also" in a four man Lousiville class that included NYC (and all BE Rookie) Edgar Sosa, sharpshooter Jerry Smith (see list above) and manchild Derrick Caracter (another New Jersey player and friend of Clark). By the end of the OOC Sosa was consolodating his hold on the point, Smith was shooting the lights out of Freedom Hall, Caracter and Pitino were locked in a "Duet of Prima Donnas" ...and Clark was searching for a role. His numbers were paradoxical, but promising -- his %ORtg was a disappointing 95.1 even as his eFG% & PPWS were 53.1 (very good) & 1.10 (also very good). If his shooting was not the problem, why was his offensive rating so low (those numbers suggest an %ORtg over 100, possibly higher...)? A high turnover rate (TORate -- 28.6) and low assist rate (ARate -- 6.1) were contributing factors. Clark's PT grew over the course of the Big East season, going from 11.8 mpg during the OOC to 16.9 mpg in the Big East regular season. By the post season (BET & NCAA) his PT had grown to 26.5 mpg. And his %ORtg was up to 106.7. RP will find a more prominent place in the rotation and offense for Clark in 2007-08.

Friday, August 10, 2007

Big East Freshmen: The 2007 All-Rookie Team

Beginning in late December of 2006 I extracted the weekly player data for Big East freshmen, sophomores and transfers from Ken Pomeroy's website. I wanted to monitor the progress of those specific groups of players. I had a number of questions about the freshmen in particular:

  1. How would those players tabbed as elite (including the preseason ROY) actually fare during the season?
  2. Who among the relatively unheralded players would emerge for honors?
  3. From the numbers which players may be flying under the radar for next season?
  4. Is freshman fatigue a measurable phenomena?

I was able to track a total of 39 BE freshmen on a weekly basis. Pomeroy posted the possession-based stats for players who earned at least 10% of the playing time for their position. The average Offensive Rating for those 39 "notable" Big East freshman was 99.6. It is not uncommon for freshmen to have offensive ratings in the 80s/90s. Kyle Lowry for example had a rating of 97.8 as a freshman and virtually exploded to 113.4 his sophomore year. The table below contains the eleven members of the All-Big East rookie team sorted in descending order by Offensive Rating (%ORtg). Reviewing the playing time (Min%) and Offensive Rating, one can build an argument for including these players. While a 14% increase in offensive efficiency (%ORtg) is not typical, note that among this group, even a modest growth of 6% - 8% would place most of these All-Rookies among the Top 200 in D1 ball...


Pomeroy provides explanations for each of the columns (except PPWS, which you can find at the Big Ten Wonk site) on his "Individual Stats Primer" page. The numbers shown confirm in most cases, the legitimacy of each player's selection to the All-Rookie Team, despite the fact that several were not household (the Big East household anyway) names at the beginning of the season.

  • The Surprises (good ones) -- Luke Harangody, Da'Sean Butler, Scottie Reynolds, DeJuan Summers, Tory Jackson and Deonta Vaughn. Harangody, Butler, Summers, Jackson (especially) and Vaughn were certainly not tabbed in the preseason as candidates for post season honors. Reynolds was an MDAA, so the surprise here is that he took the ROY honors. A few remarks about several of the pleasant surprises:
    • Luke Harangody -- his offensive numbers are set up by his %OR (percentage of offensive rebounds). 14.8% is among the best in D1. His offensive rebounding set up a good many close in shots which fed his (see his line above) possessions (%Poss) and shots (%Shots). Virtually all of his FGAs were 2s, so his eFG% is probably also his FGM%, which means he hit almost ½ of his FGAs. Not bad. Note that both his OR% and DR% are comparable to Paul Harris & Hasheem Thabeet, 2 players who received a good deal more recognition (Ken Pomeroy excepted) for their defense than did Harangody. There have been quite a few posters who claim that Notre Dame does not play defense (I mean before the Winthrop game), but with their offense how much defense did they really need? The Irish frontcourt rotation of Kurz (starter), Harangody (starter), Hillesland and Zeller provided "Just Enough" defense to win, for most of the season. His %OR, TORate and %Shots are Top 100 level for D1, while his DR%, %ORtg and %Poss are Top 500. The kid is solid.
    • Da'Sean Butler is one of the two most surprising names on the All-Rookie team. While well regarded locally as a senior HS baller in Bloomfield, NJ, Butler was an "also..." on most WVU recruiting evaluations. That he as a freshman logged 57.6% of the minutes at his position (4) in Beilein's system (a coach with a penchant for redshirting freshmen) suggests he impressed JB, no easy task. I believe that playing time and the "role" nature of most positions in Beilein's system hurt Butler when ROY candidates were discussed. With a reputation for athleticism, Butler may get a chance to show a lot more as an individual in Bob Huggins' system.
    • Tory Jackson -- logged 38% of the minutes in the Irish back court before Kyle McAlarney was arrested for illegal drug possession at the start of the Big East season. Through the end of the season Jackson's minutes shot up to 69.6% because he was called on to replace McAlarney in Brey's system. After stepping back the first month (fairly common when playing time and the player's role within the team increases dramatically...), his numbers began to recover and by season's end (with a bad loss to Winthrop excepted) he was slightly ahead of his pre Big East season efficiencies (but with nearly twice the PT). McAlarney will re-enter Notre Dame next month, but I suspect Coach Brey will play both in Notre Dame's backcourt next season.
  • The Surprises (bad ones) -- Jerome Dyson, Paul Harris and Hasheem Thabeet. Granted these players had good years by freshman standards. But much more was anticipated.
    • Paul Harris -- the near unanimous preseason ROY selection, based largely on his ability to lock down the player in front of him. He was far more likely to produce offensively from a steal and break away than off a set shot developed in a half court set. For a player who was going to a team noted for it's zone defense that ought to have been a warning sign. Ultimately, Harris started a single game (Colgate) and watched his minutes decline during the Big East regular season (from a 25.5 mpg in the OOC, to 18.6 in the Big East regular season) only to recover slightly in the BET and NIT. While definitely an asset when the ball was loose on the court -- his OR% is Top 500; his DR% is Top 100 -- his offensive production declined about 10% through the first month of the Big East regular season. He, like Lowry, compensated for a weak eFG% by getting to the line fairly regularly (FRate of 76.5 indicates that for every 4 FGAs he took 3 FTAs...).
    • If Hamady N'Diaye had Hasheem Thabeet's numbers Rutgers fans would be ecstatic. The Scarlet Knights would have played in the BET (no more, no less). But Husky fans expected more from their team. Thabeet went to Storrs with the reputation for shot blocking defense and inside scoring through athleticism. The defense part was correct, as Thabeet finished with 15.3 %Blk, (good for a #6 rank in D1) and Top 500 rankings for OR% and DR%. His offensive stats by contrast are a paradox: stats for field goal efficiency (eFG%) and ability to get to the free throw line (FRate) suggest he should have been more dominant on offense, but his %ORtg is a surprisingly low 96.2 (for every 100 possessions Thabeet scored 96 or so points). He struggled to convert free throws (0.513 FTM) and turned the ball over nearly 1 time in 4 when he had possession (24.5 TORate). Two areas to work on over the summer. Given that his role on offense did not expand over the course of the season, his %Poss and %Shot remained within 1% of their post OOC value throughout the Big East season, it was clear the staff and his teammates never regarded him as reliable option on offense.
  • No Surprise At All (thankfully) -- Eugene Harvey and Edgar Sosa. Two New York City guards who helped their respective teams. Sosa, with more to surround him than Seton Hall's Harvey, provided a sure handle and reliable (but not offensively dominant) shot for the Cardinals. Seton Hall was decimated by graduations and transfers in the wake of the Orr-to-Gonzalez coaching changeover, but Gonzo will bring in complementary talent for Harvey.