Power Rankings: Kentucky's press, freshman adjustments, more
In Volume VI of the Power Rankings, big questions are asked -- and answered: Is Kentucky still a pressing team? Which freshmen big men are for real, and which ones have been beating up on lesser competition? What does a shot chart of nearly every Frank Kaminsky non-conference field-goal attempt look like? Can we just move on to the blurbs already? (Yes.)
11Kentucky Wildcatslast week: 1record: 14-0
Kentucky's experimentation with full-court pressure -- which started in Game 1, and was documented in this season's first Power Rankings -- has continued into January. These Wildcats press more than any other Kentucky team in the John Calipari era, but they can't be classified as "all-out pressers"; they press only half as much as teams such as West Virginia, VCU and Louisville, according to Synergy's possession logs:
Would the Wildcats be pressing more if they weren't annihilating most of their opponents? The answer seems to be yes. There wasn't enough time to watch every defensive possession on film, but using Synergy's logs and game-pace data, I estimated Kentucky's half-by-half splits for each game, and came up with a first-half pressing frequency of 23.4 percent, and a second-half frequency of 11.8 percent. Here's the entire breakdown:
Kentucky's platoon depth is the reason Calipari is pressing, but he doesn't employ the same press across all lineups. The scheme hinges on the presence of 7-footer Willie Cauley-Stein, who typically acts as a menacing, mantis-like double-teamer of the inbounds pass, while the other two bigs (in the first platoon, Karl-Anthony Towns and Trey Lyles) drop back to protect the basket:
When Cauley-Stein is on the bench, the Wildcats typically do not double the man who receives the inbounds pass. Instead, they let the second-platoon point guard, Tyler Ulis, go one-on-one with the ballhandler, while a second guard lurks, looking for a possible trap closer to halfcourt. This keeps their less-mobile big men -- particularly Dakari Johnson -- out of the press and in a deep-safety role:
Next up: 1/10 at Texas A&M, 1/13 vs. Missouri
22Duke Blue Devilslast week: 2record: 14-0
Among last week's topics was the alarming split in Texas power forward Myles Turner's performance against good and bad competition; the potential NBA lottery pick had dominated opponents who were outside kenpom.com's top 100 but been an inefficient role player against opponents inside the top 100. DraftExpress' Jonathan Givony suspected that this might be the case for plenty of top prospects. Is it, or are Turner's splits the exception?
To find out -- and to make sure I'm not unfairly picking on Turner -- I ran the same, pace-adjusted splits (through Wednesday's games) on five freshman bigs who could be future first-round picks: Turner, Duke's Jahlil Okafor, Kentucky's Karl Anthony-Towns, Utah's Jakob Poeltl and Gonzaga's Domas Sabonis.
Turner's splits have only become more exaggerated -- a 44.3-point dropoff in offensive rating against good competition, plus a total absence of offensive rebounding and a high foul rate:
And what about Okafor? His offensive rating and offensive rebounding drop off somewhat, but not nearly as much as Turner's. Against good teams, Okafor is still a high-usage, reasonably-efficient star who stays out of foul trouble:
Towns, meanwhile, comes out looking good; nothing about his season-wide stats are deceiving. Aside from a dropoff in blocks and a spike in foul rate, his performance against top-100 competition is just as strong as it is against weaker teams:
In the interest of not ODing on splits charts, I'll hold the Sabonis-Poeltl numbers for later in these rankings.
Next up: 1/11 at N.C. State, 1/13 vs Miami
33Virginia Cavalierslast week: 3record: 14-0
Virginia, like Kentucky, is a defensive juggernaut -- but the Cavaliers do not press. According to Synergy, they've applied full-court pressure on just one possession this season, and it came when they needed to foul Miami in overtime on Jan. 3. Part of Virginia's success -- in addition to forcing contested jumpers in the Pack Line, as Rob Dauster did a nice job of breaking down this week -- is due to its extreme avoidance of defense-in-transition situations. It does not press, and its guards do such a good job of getting back on defense that opponents rarely see quality opportunities to push the ball in non-turnover situations.
Here are the 10 teams that spend the lowest percentage of their defensive possessions in transition; among this group, Virginia allows by far the lowest points per possession:
(Chart source: Synergy and kenpom.com, for possession-length data.)
Next up: 1/10 at Notre Dame, 1/13 vs. Clemson
44Wisconsin Badgerslast week: 4record: 15-1
Frank Kaminsky continues to have the best shot chart in all of college hoops -- a dense cluster of high-efficiency attempts in the paint, the result of mixing post-ups and dribble attacks, and then a cluster of three-point attempts from the top of the arc to the right wing. I finished charting every Kaminsky non-conference field-goal attempt that was available on film -- 151 in all, plus 18 shooting-foul situations -- and wanted to publish that version before starting on the Big Ten version in a few weeks:
(Trivial note: It took until Dec. 28, against Buffalo, for Kaminsky to take his first -- and only -- 3 from the left wing or left corner. He is a creature of locational habit.)
Next up: 1/11 vs. Rutgers, 1/15 vs. Nebraska
55Arizona Wildcatslast week: 5record: 13-1
One of the Wildcats' most noticeable changes from last season to 2014-15 is a drop-off in production on the offensive glass; they've fallen from 27th in offensive rebounding percentage to 164th. Some of this is due to losing Aaron Gordon to the NBA, but their 4-5 men have noticeably low rebound rates for players of their length and skill. Brandon Ashley grabs just 7.0 percent of available offensive boards, and Kaleb Tarczewski, who gets only 6.5 percent, has just one offensive rebound in his past 107 minutes of play. On a team that doesn't rely heavily on three-pointers -- and thus shouldn't have an overwhelming amount of long rebounds -- that's a very low number for a 7-foot center.
Next up: 1/8 at Oregon, 1/11 at Oregon State
66Gonzaga Bulldogslast week: 6record: 14-1
Fun With Freshman Splits, Part II:
Domas Sabonis is pretty much the same rebounder against good teams as he is against bad, but he's the most foul-prone of all the freshman bigs, averaging 7.0 fouls per 40 minutes (pace-adjusted) against elite competition.
His offensive drop-off is more due to increased turnovers than poor shooting -- the fact that he still makes 65.4 percent of his shots against top-100 teams is promising.
Next up: 1/8 vs. San Francisco, 1/10 vs. Santa Clara
77Utah Uteslast week: 11record: 13-2
Fun With Freshman Splits, Part III:
As for Jakob Poeltl, his offensive splits are Turner-like: when the Austrian 7-footer faces good opponents, his usage rate drops from star to role-player level, and his efficiency drops by nearly 40 points due to a spike in turnover rate. However, on defense -- his area of biggest value to the Utes -- Poeltl's rebounding and block rates improve in bigger games. He's a major reason why they're ranked sixth nationally in defensive efficiency.
Next up: 1/15 at Arizona State, 1/17 at Arizona
88Villanova Wildcatslast week: 7record: 14-1
I saved this Villanova play from the second half of its 90-72 win at St. John's on Tuesday for a few reasons.
It was a momentum-killer, tying the game at 50-50 and ending a 7-0 run by the Johnnies. It was an excellent, everyone-on-the-perimeter set to run against a 2-3 zone, as it engaged the top two defenders in the Ryan Arcidiacono-Daniel Ochefu ballscreen, and forced the bottom-left defender, Sir'Dominic Pointer, to choose between guarding Josh Hart (on the left wing) or Kris Jenkins (in the left corner).
It also made me wonder: Why guard Arcidiacono, a 22-percent long-range shooter, and Ochefu, a non-shooter, that way on a ballscreen? If the top-right defender in the 2-3 just goes under the screen, since Arcidiacono isn't much of a shooting threat, and the back-center defender (Chris Obekpa) keeps an eye on Ochefu's roll, then you have two defenders available on the left side, and Hart and Jenkins (who eventually makes the shot on Arcidiacono's feed) are accounted for. Oh well: This is hardly the first time St. John's has done something schematically confounding.
(GIF source: Fox Sports 1.)
Next up: 1/10 vs. DePaul, 1/14 vs. Xavier
99Notre Dame Fighting Irishlast week: 19record: 15-1
While Kaminsky and Okafor are the clear frontrunners in the national player of the year race, Jerian Grant might be the player with the best chance of catching them. Notre Dame's point guard has been ridiculously efficient (his 132.1 Offensive Rating is tops in the country among players who use at least 25 percent of their team's possessions) and he already has a signature dunk; what he still needs is a signature game against one of the ACC elites -- Virginia, Duke or Louisville -- to get into the serious NPOY conversation. In the meantime he'll have to settle for being the starting point guard on my as-of-Jan. 8-All-America team:
PG: Jerian Grant, Notre Dame
SF: Justin Anderson, Virginia
PF: Frank Kaminsky, Wisconsin
PF: Jahlil Okafor, Duke
C: Willie Cauley-Stein, Kentucky
6th (CG): Delon Wright, Utah
Next up: 1/10 vs. Virginia, 1/14 at Georgia Tech
1010Louisville Cardinalslast week: 9record: 14-1
The Cardinals' oft-maddening point guard, Chris Jones, is a frequent topic here, due to him being a drain on their offensive efficiency. In the interest of balance, I'm taking this blurb to celebrate his impact as a pressure-applicant in one of the country's best defenses. Jones consistently makes a big impact on D, and he continued that in Wednesday's win over Clemson, ball-hawking for 36 minutes, recording five steals and helping force the Tigers into their second-worst turnover percentage of the season. This edit is titled Chris Jones Causes Problems (For The Other Team):
(Edit source: ACC Network)
Next up: 1/10 at North Carolina, 1/13 vs. Virginia Tech
1111Kansas Jayhawkslast week: 12record: 12-2
Freshman wing Kelly Oubre Jr. used to cause problems for Bill Self, but now Oubre, like Chris Jones, causes problems for opposing offenses. A sampling of what Oubre's telescopic wingspan did to Baylor on Wednesday:
When Oubre was in his "doghouse" period -- the seven games prior to a Dec. 10 win over Georgetown -- his steal percentage was just 1.1. From the Georgetown game forward, as he's taken on a bigger role and cracked the starting lineup, Oubre's steal percentage is almost FOUR TIMES better, at 4.3 percent. He has the potential to be the Jayhawks' best turnover-creator in Big 12 play.
(Edit Source: ESPNU.)
Next up: 1/10 vs. Texas Tech, 1/13 vs. Oklahoma State
1212Oklahoma Soonerslast week: 21record: 11-3
I have my doubts that this will hold up, but in a development that pretty much no one expected in the preseason, the Sooners' defense is ranked fourth nationally in adjusted efficiency on kenpom.com -- one spot ahead of the Virginia defense everyone's been fawning over for the past few months. Lon Kruger's previous three Oklahoma teams never ranked higher than 89th, but the interior defense provided by the Ryan Spangler-TaShawn Thomas duo, combined with the newly foul-averse nature of the Sooners' entire starting lineup, has resulted in major defensive improvement.
Next up: 1/10 vs. Kansas State, 1/13 at West Virginia
1313Maryland Terrapinslast week: 8record: 14-2
The Terps' defense has the second-lowest turnovers-forced percentage in the Big Ten, at 17.0; only Indiana, at 16.5, is worse. Maryland has been successful without needing to gamble to create takeaways, but the common theme of its two losses -- to Virginia on Dec. 3 and Illinois on Jan. 7 -- is that its defensive TO% fell below 12. The Illini didn't shoot all that well from the field -- just 40.6 percent on twos and 30.4 on threes -- but cruised to the upset because they controlled the ball.
Next up: 1/10 at Purdue, 1/14 vs. Rutgers
1414Wichita State Shockerslast week: 15record: 13-2
We're nearing mid-January and the Shockers' highest possession-user, at 27.3 percent, is not Ron Baker, or Fred VanVleet -- it's center Darius Carter. How big a surprise is that? Well, Carter was a non-starting role player last season, and now he's using more possessions than Cleanthony Early did in '13-14 (26.8). Whether this is good for Wichita is TBD; it obviously needs an interior presence to keep defenses from keying too hard on its guards, but Carter hasn't been hugely efficient as a high-usage guy. Baker, who can score off the catch and the bounce, seems to be a more ideal focal point.
Next up: 1/11 at Loyola (Chicago), 1/1/14 vs. Southern Illinois
1515Iowa State Cycloneslast week: 10record: 11-2
All the attention on the passing of Monte Morris and Georges Niang, the scoring infusion of Bryce DeJean-Jones andthe (expected) defensiv infusion from Jameel McKay made me forget, on occasion, just how valuable Dustin Hogue is to Iowa State's offense. The 6-6 junior college transfer is a legitimate inside-outside threat who keyed the Cyclones' Big 12-opening win over Oklahoma State on a night where Niang was slumping. Hogue had largely abandoned his long-range shot in the lead-up to conference play, taking just five attempts in his previous eight games, but he went 3-of-4 from deep against the Cowboys, who put far more defensive attention on containing Niang than they did chasing Hogue to the perimeter.
Next up: 1/10 at West Virginia, 1/14 at Baylor
1616West Virginia Mountaineerslast week: 17record: 14-1
In the "Most Frequent Pressing" teams chart from the Kentucky blurb, West Virginia (at 39.0 percent of defensive possessions) and Ohio State (35.5 percent) are ranked Nos. 1-2. Their style of defensive pressure, however, could not be more different; the Mountaineers are interested in creating total chaos, and thus their average defensive possession length is the shortest in the country, at 14.7 seconds, according to kenpom.com. The Buckeyes, who apply pressure and then fall back into a grind-you-down zone, rank 333rd in defensive possession length, at 19.6 seconds. Their early defense seems more designed to kill a portion of the shot clock than it does to speed up their opponents.
Next up: 1/10 vs. Iowa State, 1/13 vs. Oklahoma
The Next 16
19. Ohio State
20. George Washington
21. Northern Iowa
24. Seton Hall
26. Old Dominion
29. North Carolina