Things are clicking for UK...what's changed?
I look at what's made the difference for UK after 2 straight blowouts, plus what should UK do at power forward?
John Calipari has said frequently this season that he likes his team and now it seems like the fanbase is starting to agree with him. The early season wins looked like a menu designed by a 4 year old; plenty of cupcakes, but nothing of any real substance for roughage. Back to back blowouts over quality-ish opponents (a likely tournament team in UNC and a solid mid-major in WKU) showed that UK can put impressive results against teams with a pulse.
The Wildcats are about to start conference play, where nearly everyone except Georgia and Missouri is probably going to give them a test. Per KenPom, UK is only projected to be more than a 10 point favorite in 3 conference games (the aforementioned Terrible Two, plus home against Ole Miss). It is critical than UK identify anything sustainable from their last 2 blowout wins while also addressing any weaknesses. With that in mind, my latest edition of Hoops Insight focuses on 2 questions:
What was the key to the improvement against UNC and WKU?
What should UK do at the power forward position?
Before I get to that, I wanted to let my subscribers know that my stats engine has a couple new features. First, I’m adding data on some SEC opponents that UK will play twice this season. LSU’s data is live on the site, with others coming over the next week. Second, I’ve added the ability for users to download formatted play by play data in .CSV files for their own analysis. I hope this will allow more data-driven users to discover things they won’t get anywhere else!
HUGE improvement against quality teams
I’d consider 5 UK opponents so far this season to be “quality” opponents: Duke, Ohio, Notre Dame, UNC, and WKU. Those are the only 5 games UK has played that wouldn’t be considered Quad 4 games (according to BartTorvik.com), although the Ohio and WKU games would only count as Quad 3 since UK was at home. Let’s use these as our comparison set to evaluate how UK played against “quality” teams.
In the first 3, UK scored 108.6 points per 100 possessions and allowed 107. Neither is a great number, and adjusting for opponent quality we can estimate that UK played about 17 points better than an average team in those games. That would rank Kentucky about 30th in the country, on par with Florida or Memphis.
In the last 2, UK has scored 138.9 points per 100 possessions and allowed 85.2 (excluding garbage time). Those are both stunning figures, but especially the offense. According to BartTorvik.com, there has not been a game against a top 150 team this season where an offense scored 138.9 points per 100 possessions. Even including garbage time, the only game where a team scored more points per game against a current top 50 team was Purdue against Florida State. Purdue is the #1 offense in the country, so UK did something special.
So what is Kentucky doing exactly?
Getting a bit lucky on defense, but making the right moves on offense
I’ll start on the defensive side because the story there is less interesting and I want to save the good stuff for a second. Two things have really changed from UK’s earlier struggles against quality teams:
UK is forcing more turnovers
UNC and WKU lost all ability to hit threes
The first one is pretty easy to explain. Among UK’s 5 quality opponents, 3 of them are really good at avoiding turnovers (Duke, Ohio, and Notre Dame). They all rank in the top 64 in lowest turnover rate, and turned it over on 12% of possessions vs UK. 2 (UNC, WKU) are outside the top 100 in committing turnovers and they turned it over on 18% of possessions vs UK. In SEC play, Florida, Alabama, and LSU are turnover prone but other SEC contenders are not. This one probably won’t persist all that much.
The three point shooting is more complicated. Duke, Ohio, and Notre Dame combined to bit 30% of their threes in halfcourt offense against UK, and that was with Ohio throwing in 4 of 7 three point attempts towards the end of the shot clock. UK held these teams to 23% from three in transition, and has largely limited opponent three point shooting all season with a total of 27% opponent 3pt FG%. That went to another level against UNC and WKU however as they combined to hit 4 of 33 (12%), including just 14% in halfcourt offense and 11% in transition. Western is not a good shooting team, but UNC is 12th nationally at 39%. UK’s 3 point defense is good, but nobody holds opponets to sub-20% for any real stretch. This won’t persist, but luckily UK gave themselves a nice margin for error thanks to their offense.
That offense, my word. UK has been on fire, posting an effective FG% of 60% in the last 2 games. Including the Robert Morris game earlier this season, UK has 3 games at or above a 60% effective FG% this season; they had 4 total over the past 2 seasons. Kentucky’s gotten to that figure by being excellent in almost every area:
UK has taken 37% of their shots in transition the last 2 games, up from 31% in their earlier quality games, and has a 68% eFG in transition. That % of shots in transition would rank 2nd for the season and the eFG in transition would rank 4th
UK is hitting 48% of their threes. That is amazingly good.
UK is hitting a mind boggling 84% of their shots at the rim in halfcourt offense in the last 2 games. UK now leads the nation in FG% at the rim for the season at 76%.
Kentucky has made some tweaks to the way they’re getting shots in the last couple of games as well. Some of these are due to playing against the WKU zone (fewer shots out of pick & roll) but they’ve made some other nice adjustments. Per Synergy data, UK has started getting more shots off of cuts and dribble handoffs, and fewer shots off of pure spot-ups. Kentucky has been pretty inefficient on their spot-ups this season (42% eFG%, per Synergy) so they’re shifting into more efficient actions (69% eFG on cuts, 60% on handoffs). Kentucky has also reduced their propensity for jumpers off the dribble (which tend to be long twos) and take a larger share of shots at the rim.
One player who’s been used more effectively is Kellan Grady. Earlier in the season he was used almost exclusively as a spot up shooter or in transition, but he’s actually a pretty effective shooter on the move. He’s now taking more shots off of handoffs or off screens (6 the last 2 games, after 8 total in the first 9 games). I don’t know if UK is purposely creating these actions, if he’s finally getting the ball off of them, or if he’s starting to shoot off them, but it’s working.
A dilemma at power forward
It has always struck me as odd that Kentucky fans seem to undervalue the contributions of Keion Brooks. Last season he averaged 17.5 points and 11.5 rebounds per 40 minutes. This season he’s at 18.8 and 7.4. He’s been a quiet contributor without much flash, but on Twitter most fans seem to think he’s eminently replaceable. I’ll be honest, he’s been one of my favorite Wildcats over the past few years for the way he’s grown his game in quiet ways. He’s always had very good block and steal numbers for a 6’7” 205lb forward, he’s increased his 3 point shot volume to try to stretch the floor, and has developed into a 70%+ shooter from the foul line. He’s still never hit a good percentage from three, but has hit 44% of his 2 point jumpers this season (up from 33% last season).
Last season Jacob Toppin was a bundle of athletic potential, but not as effective as Brooks. This season however Toppin has been able to be largely effective for UK in his minutes, while Brooks has regressed in a couple of key areas. Coming into this season I thought it was absurd to have a Kentucky starting 5 without Keion Brooks, but Jacob Toppin has done enough this year to make it a possibility.
Let’s first look at how UK performs with each:
With Brooks in, UK scores 129.2 pts/100 poss and allows 97.8. They have an adjusted margin of +31.6 pts/100 poss
With Toppin, UK scores 123.7 per 100 and allows 90.3, for an adjusted margin of +39.1
Kentucky shoots better with Brooks in, but allows a lower eFG% with Toppin. They shoot much better from midrange (thanks to Brooks himself), and a bit better at the rim and from three
Kentucky’s rebounding is virtually identical with either
Kentucky opponents get to the rim much more often with Toppin in, but shoot worse when they get there
The numbers are pretty comparable honestly. Brooks’ ability to generate offense is more than what Toppin brings, but Kentucky has been better on defense. This plays out in their individual stats as well:
Brooks takes 26% of UK’s shots when in the game with an eFG% of 46%. He hits 67% at the rim and 40% on other twos. 58% of his shots are twos away from the rim.
Toppin takes 15% of UK’s shots with an eFG% of 43%. 61% of his shots are at the rim, where he hits 65%, but he’s 1-11 on any other shots including just 0-1 from three.
Per Synergy, Toppin has been excellent when guarding spot up shooters (they are 0-10 from three when he’ the defender) while Keion’s assignments have dne 8-23 from three. Toppin also has slightly higher block and steal rates
Their offensive styles also affect their teammates in unique ways. Brooks has taken a higher share of shots this year and has regressed a bit as a passer, assisting on only 2% of teammate baskets compared to 15% last year. Toppin shoots less often and has added value as a passer, assisting on 15% of UK baskets. You can see their impact on teammates:
Kellan Grady has a larger role with Toppin in the game, taking 19% of shots at a 73% eFG vs 13% of shots and a 68% eFG% with Brooks
Oscar Tshiebwe is very effective with either player, but takes 26% of UK shots when playing with Toppin vs 21% with Brooks
TyTy Washington is more effectiuve with Brooks and the spacing he provides for drives, with a 64% eFG on 24% of UK shots compared to a 31% eFG% and 19% of shots with Toppin. Washington gets to the rim far more often with the spacing Brooks provides
Sahvir Wheeler is the one player who has almost no change with either player, with a 48% eFG% on 20% of UK shots whether Brooks or Toppin is in
Although Toppin has played more often against opposing bench units, the stats for Brooks and Toppin don’t change materially when looking just at minutes played against 4 or 5 opposing starters. There’s not a clear separation between the two like there was earlier in the season. Brooks is a more dangerous offensive player, but Toppin does bring more of an ability to challenge shooters. Both are solid rebounders. If Brooks was passing like he did last season, or if his three pointer starts to fall, he’s probably the better option. Brooks has not played poorly this season and I don’t believe he deserves to see a reduction in his minutes, but Toppin has been effective enough to deserve minutes of his own. Unfortunately neither player can play small forward, and neither probably defends well enough to play center unless it’s an emergency. Toppin is a good option to defend power forwards who are outside shooting threats, but Brooks brings more on the offensive end. It may end up being more of a question of matchups. I’m still a Brooks guy, but it’s not unreasonable to think Toppin is a good alternative.
UK has two power forwards who are both good enough to play. That’s a good problem to have. UK has only had 2 games where they were outscored in Brooks’ minutes this season (Duke and Notre Dame), and only 1 where they were outscored in Toppin’s minutes (Duke). During the last 2 games Brooks’ minutes have been the lowest they’ve been all season, but UK has been +21 in 36 possessions with him and +37 in 72 possessions without him. Either way, UK is playing at an incredibly high level.