FanPost

SPTSJUNKIE’S 5th Annual Draft Model Projections: Just draft Doncic?

Welcome back to another draft and another visit to the lottery for our beloved Kings. Since the lottery gods have smiled upon us this year, hopefully some models and draft profiles can help us learn a bit more about whomever the Kings select at #2, #37, and any additional acquired picks.

As a quick refresher, last year I moved to 5 models that could help us gain different perspectives into why prospects might succeed or fail, here is a quick one sentence description of each model:

Primary model: The base model I’ve done every year with statistics and scouting variables back to 2005

Modern model: One problem with the primary model is back in 2005 there were players putting up huge Win Shares who simply wouldn’t fit in the modern NBA, so this sacrifices some sample size to start with the 2009 draft

Position Model: This model was run evaluating each position separately, which allows the algorithm to "discover" metrics that are very important for PGs, but not for PFs

Humble: In a circle of life moment, I’m incorporating Layne Varsho’s use of actual draft order / mock draft order as a variable, which defers to scouts knowledge and captures some additional signal my scouting variables alone do not. Exciting to see Kevin Pelton, Nate Silver, and others are using this now

Position Humble: Evaluates each position individually, but also incorporates in the draft order variables

Modern Humble: Cleary using draft order along with the more modern model

One more important note: I included Luka Doncic and Michael Porter in this model – no adjustments were made to either of their stats. They were treated like two college players who had played a full season. Their numbers should be taken with a grain of salt, but given they are both in play at #2, I thought they were worth including. I’m like a kid with chicken pox who can’t help but scratch that itch.

How to interpret the scores:

11.5 and up: Superstar potential

7.0-11.4: Solid starter with all star potential

5.0-7.0: Mostly 6th man / solid starter potential

3.0-5.0: Solid bench contributor

Below 3.0: Lucky to be in the league

The Rankings

Draft_Scores_v2_zpsmemqtf07.0.png

Considerations and Caveats

When interpreting this kind of data and adding to your own personal scouting, it’s important to understand not only what data is important in the model, but also what does it do well and poorly? So having poured over its predictions and retrodictions (fancy word meaning predicting the past, which even with fancy training/testing techniques, models seem to do much better than predicting the future) what have I seen:

How to interpret each model

A quick note on how I would think about interpreting each model:

Primary: Traditionally, this has done very well with guards and big men with more of a mixed record on wing players. It also tends to overrate some of the slower PGs who struggle in the NBA still stand out (e.g., RayMac, Burke) and old school bigs (e.g., Diamond Stone).

Modern: Really seems to place an emphasis on skilled players, but may overlook some of the gifted athletes who can still contribute. Seems to favor big men a bit.

Position: Works pretty well and I can explain some of the differences between positions in the comments, but needed some positions seem clearer than others looking at the retrodictions. However, some players can score highly at a position they could not play in the NBA, so you do need to use some judgement (e.g., some important variables for a center look like guard stats, so Ball scores like a decent center, although he could never play that in the NBA)

Humble: Seems to work really well, but can prop up highly drafted busts a bit by giving a bit too much weight to what the scouts though making them look like decent prospects instead of complete duds like the other models will.

Positional humble: Also seems to work very well, but even with cross validation using testing and training sets, I am a bit concerned this model could be a bit over-fit (although I did dial this down by removing flags for variables like "drafted top 5" that really can have a big impact on the model)

The positional models seemed to struggle a bit this year and may need some revamping unless there are some real flaws to a few prospects we don’t see yet. Will be interesting to follow.

Thoughts on individual players

Are the Suns decisive or suffering from [Miami] heat stroke?

DeAndre Ayton

Model view: With the exception of the problematic position models, the other 3 really liked, but did not love him. He typically ranked 5th-6th, but with numbers that firmly showed him to have all star or superstar potential. The models loved his age and productivity, but struggled with his lack of elite defense or high basketball IQ.

Scouts view: Build like an old school center, but with the agility and potential skill set of a modern center. Ayton is a solid post player, who is strong cutting to the hoop and on the offensive boards. His shooting shows promise, but his form is inconsistent (legs tend to land differently). Strong potential in the pick and roll, but underutilized there in college (as most bigs are). Questionable defensive player – both in college and like Embiid, while he has good lateral quickness for a large center, can he ever guard the perimeter enough to keep up with the elite teams who will put him in the P&R defensively.

Stats view: Averaged nearly 24 and 14 points per 40 and hit a respectable 73% of his FTs. His 0.7 steals and 2.3 blocks per 40 are low for someone with his gifts.

What you have to believe for him to be a star: He can become a P&R screening threat on offense and continue to hone his jumper. He will become at least passable defensively after being one of he worst defensive players being drafted (by DBPM).

Verdict: Given his size, athleticism and passable skill level already, it’s hard to imagine he isn’t one of the top 5 players from this draft with the upside to be a top 10 player in the league. Comes into the league a bit more underdeveloped skill-wise than bigs like AD and Towns. And his defensive awareness in particular needs to improve greatly. My gut says he’s a strong player for the Sun and likely an all star given the dearth of legitimate centers in the league, but my gut says there will be 2-3 players from this draft who have better careers – especially if we judge them by more advanced stats (RPM or the next improvement on RPM) instead of pure points and rebounds.

NBA comp high: KAT Part Deux – At his peak, another inside-outside mobile big with passable defense

NBA comp medium: Taller, leaner Zach Randolph – A double double machine who is always a step slow on defense and settles for too many mid-range jumpers, but can carry an offense

NBA comp low: Andrew Bynum – Talented big man who has stretches that tantalize and fits next to another 1-2 alpha players who can take over when his motor wanes. Not bad character, but also a lack of desire prevents from reaching his ceiling.

Whose number two and who’s a number two? The Kings’ supposed final three choices.

Luka Doncic

Model view: So again, grain of salt, but at least one of the models thinks his Euro numbers make him the best prospect in the history of the database (college players 2005-Present). I strongly doubt that, but the worst model still has him as a potential all star. The models love his high IQ, assist to turnover rate, high levels of productivity per 40, combination of points and assists, and steals.

Scouts view: Very high basketball IQ and ability to run a team. Great vision and passing with a near perfect command of the screen and roll – able to pick great angles to hit to roller or secondary target. Good rebounder and shooter. Very creative dribbler, but doesn’t blow by most defenders with speed or a great first step. Good team defender…… But, is he athletic enough to be great and not simply good offensively and good and not bad defensively.

Stats view: 23-9-7 per 40 with 1.7 steals. As importantly for his potential shooting, an 80% FT shooter two years straight and 35% three point shooter on fewer attempts per game in 2017.

What you have to believe for him to be a star: He has the athleticism to defend and create offense . His ancillary numbers are a better indicating of his three point shooting than this year’s three point percentage.

Verdict: I have to be honest, I have a very hard time watching European games and putting the performance in context like I do with the US. So this is a hard call. I think Doncic is a good, but not great shooter. I always side with skill / IQ over pure athleticism, so long as a player has a positon and enough athleticism. And I think Doncic can slim down, as we have seen with some college players like Harden, and get a bit quicker. Ultimately, I think his ceiling is very high, but at worst, he’s essentially Lonzo Ball with better shooting. And despite his flaws, Ball did well in LA this year and was perhaps the third best rookie from his draft. The ability to run the pick and roll, orchestrate an offense, and make offense easier for your teammates is very valuable, even if the player cannot ISO or turn the corner on elite players like a Harden or CP3. I think Doncic is a safe player, who even if he’s only good, he still fits the modern NBA and will be easy to plug into a team. However, if he can get quicker and improve his jumper, he has the potential to be a superstar.

NBA comp high: Brandon Roy – Harden improved so ridiculously, I just can’t say Harden. But Roy is another big guard with a good, but not great outside shot and the ability to create for his teammates. He was more athletic than Doncic is today, but he did not dominate with elite athleticism, so with a NBA regimen it’s possible Doncic could get closer to that level.

NBA comp medium: Lonzo with a jumper / Toni Kukoc / Hedo – Take your pick – big G/F who can shoot and is a great playmaker for his position. A #1/#2 option at times, but not someone you want to rely on carrying your team. Best as a cog with a great supporting cast (as opposed to a 2 stars and scrubs team).

NBA comp low: Mike Dunleavy – Absolute worst case, likely a 10 year starter who will be able to start for playoff teams, even if he is more of a role player than a star.

Marvelous Marvin Bagley

Model view: Generally 3rd-5th except in the positional models. Lower in the "modern model" than the ones including some of the years where older school big excelled. The models loved his age, productivity, and his elite athleticism. They did not like his A/T relative to other players at his position, his turnovers, and lack of steals and blocks.

Scouts view: Perhaps the best pure athlete in the draft – certainly the best for a PF/C. Marvin has a quick first step. Great body control and ability to finish inside. Fantastic hands – making him a great fast break option, cutter, or potential screen and roll man. Great "double jump" ability. Very good lateral speed, which gives him defensive potential. Great motor. Poor passer – both touch and vision. Needs to improve right hand (shoots with left on layups on some moves that should be right handed). Poor rim protector. Great potential for defending the perimeter, but poor defensive awareness and technique (crosses his legs, closes out poorly, takes bad angles defending the pick and roll).

Stats view: 25-13-2 per 40 with a 64% TS%. However, 1 steal and 1 block per 40 and 63% FT (33% mid-range, 14% off the dribble via Synergy Stats)

What you have to believe for him to be a star: He has outlier skill development to match his physical gifts – improving some combination of his right hand, handles, passing, defensive IQ, rim protection, and shooting.

Verdict: Between his athleticism, motor, and the skills he has – he should be a good player and he certainly has upside. He may be a unique player who needs the right system, but it’s hard to imagine he can’t at least be a rich man’s Julius Randle. He should immediately be one of the best rebounders iand finishers in the NBA, even if he initially struggles to create for himself, except in opportunistic matchups. But get Bagley in space and he’s going to be an offensive weapon. Additionally, if he does experience outlier development, with his gifts, he could be a top 5 player.

NBA comp high: Amare with better lateral quickness – Take Amare and give him the lateral speed to defend the perimeter and that’s a great weapon.

NBA comp medium: Rich man’s Julius Randle – Great motor and athlete, who is fantastic finishing and scoring inside, but struggles defensively and limits spacing offensively with his limitations.

NBA comp low: Leon Powe – There’s no way he is as bad as Throb. Worst case, he’s a bench big who can provides energy, boards, and scores some points as the ultimate garbage man (TM Cedric Ceballos)

Michael Porter

Model view: Probably not a surprise the only one that liked him was the Humble version – mediocre college stats playing hurt made it hard for him to pop in any other model.

Scouts view: Near ideal physique for a modern SF/PF. Very strong shooter. Quick first step and strong athleticism. Potential three level scorer. Plays upright and a bit passive on defense. Not a bad passer, but doesn’t have great vision or involve his teammates enough.

Stats view: 23-15-1 per 40 with 2.3 steals per 40, but a 0.35 A/T and 44% TS.

What you would need to believe for him to be a star: Elephant in the room? How’s his back. Smaller elephant hiding in the room? Even in high school there were questions if he was more than a one dimensional scorer and non-factor defensively. Without any healthy college tape, is he a great player who just focused too much on scoring in high school and could get away with lazy, upright defense?

Verdict: GMs must be sweating this year. Last year felt easy, now we are talking about trying to contextualize Euroleague games and high school / AAU games. It’s clear that Porter can shoot. He has the potential to be a three level scorer, but his handles are even more raw than Tatum’s and the Celtics brought him along slowly (allowing him to play off Horford and Irving). I’m a bit concerned about his drive and personality compared to Doncic and Bagley. He plays softer and more selfish than either. Here’s prima donna potential with Porter and I could see him being more of a one dimensional scorer like Wiggins or to a lesser degree Barnes.

NBA comp high: The Rudy Gay we always wanted – Athletic SF/PF who can guard big or small and score from anywhere, but with greater drive.

NBA comp medium: Andrew Wiggins – One dimensional shooter / scorer who doesn’t defend or create for others

NBA comp low: Walt Williams / Ed O’Bannon – Going old school. Porter can shoot and should get opportunities, if healthy, he could be the Wizard, a pretty one dimensional scorer and non-defender, who was less effective than Wiggins. Or if his back is an issue, poor Ed, NCAA legend with chronic injuries.

The players we should consider, but don’t appear to be


Wendell Carter Jr. (for you RORDOG)

Model view: Ranked between 1-7 in every model. Such a versatile big man, he scored well regardless of the criteria – good passer, shooter, rebounder, solid steals and blocks, A/T for his position, etc. His strength in all of the models demonstrates he is a well rounded player.

Scouts view: Perhaps a bit hidden at Duke offensively next to Bagley. Good passer. Solid defender with some lateral movement. Has good post moves, a good jumper, can put the ball on the floor, and is an adept passer. Generally plays tough and not afraid to mix it up. However, can play passive and does not demand the ball / rise to the moment in a bad way. Good, but not great athlete, loads up on two feet to dunk and can play a bit below the rim at times.

Verdict: He’s been on my underrated list all season. He may not be a traditional star who scores 20+ PPG, but has the potential to become an Al Horford like player who you can run a lot of offense through in the low and high post and who can play off ball and stretch the floor as well. He’s likely to be a steal and wouldn’t mind trading back up into the lottery for him with one of our guards if we draft Doncic.

NBA comp high: Al Horford or Andrew Bogut part deux

NBA comp medium: Brad Miller – Highly skilled offensive big who is a good defender though timing and IQ as opposed to pure athleticism.

NBA comp low: Cody Zeller

Jaren Jackson Jr.

Model view: High variance – as high as #3 in the modern model and as low as 16th in the primary model. The models love his age, steals and blocks, and elite defense. They do not like his lack of points and assists or his assist to turnover rate relative to his position.

Scouts view: Monster defensive potential with the lateral speed to defend the pick and roll and perimeter. Good three point shooter with a quick release, but odd form. Ideal physical profile for a center, although only a good athlete, has to load up on two feet to attack the rim. Lacks defensive discipline – tries to block / steal everything, which leads to foul trouble and gives up a lot of rebounds to opposing teams. Poor at creating offense. Can blend in and disappear for long stretches.

Stats view: 20-11-2 with 5.5 blocks per 40. 80% FT and 39% on threes.

Verdict: A rich man’s version of Isaac from last year. Monster defensive potential, but will really need significant development to become a star player offensively. JJJ’s numbers get a bit overrated, because his per 40 numbers are scaling up from 22 mpg, since he fouled a lot and was benched by Izzo. About 33% of his games this year were dud stat lines. On tape, you can see Isaac stumble around and play a bit wild, he still seems to be getting comfortable with his body. I think he gets there defensively. I think he’s a complementary offensive player. He has the potential to be a Ben Wallace, Draymond, Ibaka type of player whose impact on the game is far higher than he gets credit for, since he’s unlikely to ever be a top 2 or even 3 option.

NBA comp high: Next Gen Serge Ibaka – bigger, longer, and uncut… motor… yeah that ends the sentence

NBA comp medium: Myles Turner with better perimeter D

NBA comp low: Higher motor Raef LaFrentz – Look it up kids

Mo Bamba

Model view: Really struggles in most of the models with his lack of points, assists, and A/T. His blocks are nice, but have never given much more than a small boost in my models.

Scouts view: Unicorn who could be a defensive force. Potential to shoot, but still very theoretical. Extremely long and nimble enough to not be a liability on the perimeter. Solid hands and good finisher, but little ability to create his own shot and gets pushed around at times due to thin frame.

Stats view: 17-14 with 5 blocks per 40 and an 89.6 D-Rating are good. 68% FTs and 0.7 assists per 40 with a .35 A/T are not.

Verdict: Much more raw offensively than I think a lot of people realize. Bamba is likely going to need a year or two to contribute much of anything offensively. Players like DJ and Drummond had far more bulk and explosiveness and still took a few years to really become efficient. Has some touch on this shot, but empty gym videos aside, it really hasn’t shown up in games. With that said, he will instantly improve the defense the moment he steps on the court and shows far more restraint than JJJ – too much at times. Should be an impact player, but may be even more limited offensively and have less realistic upside there than JJJ.

NBA comp high: Mutombo with a three point shot – I think moves better laterally than Gobert and has somewhat similar per 40 numbers to Mutombo in college.

NBA comp medium: Marcus Camby – The original board, block, and shoot player. Camby was more skilled offensively at the same age, but was more complementary in the NBA.

NBA comp low: Samuel Dalembert – If his shot never develops, should still be a factor in the paint, although even Sammy would struggle a bit more in today’s NBA

Other players:

Trae Young – My models do tend to overrate highly skilled, slow footed PGs – but the models do love him and his college numbers were fantastic. I think he’s too short / slow to really be another Steph. But I do think people are sleeping on Young and he is going to be a starting PG and shooter who could be another Lillard.

Mikal Bridges – The models treat him as a solid starter or backup. I think that’s fair. He can shoot and defend, which should guarantee him a long career. But he’s not going to be able to defend PFs, he really cannot create for himself or others, and I really think Villanova maximized his skills instead of hiding them. He was below average for a college player at shooting off the dribble. Wright had players like Brunson running a lot of P&R and ISOs (via synergy sports). Mikal was 4th on the team. If Bridges had the ability, I don’t think Wright had any motivation not to take advantage.

Miles Bridges – It’s almost easy to miss that Bridges scores well in all three non-positional models. This is one of the deeper drafts in a while according to the models. His slow shot concerns me. And I think he may need to slim down some to be a true SF/PF. But Bridges has talent and the ability to create and set up his teammates. I struggle to see a star, but looks like another 10 year starter with more upside than his counterpart (Mikal).

Zhaire Smith – Rates at a solid starter in just about every model. In some early versions, he was rating much higher. An uber-athletic wing who should be an elite perimeter defender and is a capable shooter and monster dunker. His handles need significant improvement, but one of the highest upside picks with a realistic shot at hitting his ceiling in the mid-first round. He was a late bloomer who has been praised for having a strong work ethic.

Troy Brown – Anther favorite sleeper of mine. Another gym rat who scored well in nearly every model (at least relative to his projected draft position). Brown is a high IQ player who is a good ball handler and strong defender. His main issue is his three point shooting (29%), but he hit 74% of his FTs and is not even 19 yet. I think we are going to see a savvy franchise target Brown and potentially even trade up to get him. He has 15 year San Antonio Spur written all over him.

DeAnthony Melton – Every year there is one name that scores really high despite relatively little draft buzz. Melton had to sit out 2018 due to the NCAA scandal hitting USC, but in 2017 he averaged 7 assists, 4 steals, and 2.2 blocks per 40 minutes. He’s not much of an offensive player, but if he’s not a poor ball handler, so if he can improve his three point shot, he can be an ideal 3-D combo guard. A good second round target for us, although I don’t think he lasts that long.

Jacob Evans - Scored as a starter and even a potential all star in some of the models. Evans is a high IQ player who is an average athlete on paper with good applied athleticism stats e.g., he gets the most out of what he has). He could also fit the mold of a modern 3-D player. The "backcourt" version of Draymond, Green averaged 17-6-4 with 1.7 steals and 1.3 blocks per 40. Evans is a good spot up shooter, who hit 37% of his threes on over 4.5 per game.

Jevon Carter – Scores well in a couple of the models due to his high productivity and high volume of steals – Carter is another decent second round target. Hard to tell with upperclassmen how much of their improvement will carry over to the NBA, but Carter is another pitbull defender, who shot 39% from three on nearly 5.3 per game. If we draft Doncic in the first round, adding another strong perimeter defender / shooter could help. Has potential to be another Derick Fisher [ducks].

Gary Clark - Clark is a senior, who averaged 18-12-3 with 1.7 steals and 2 blocks per 40. His outside shot has improved as he strives to be a stretch four, but he's going to have to prove his improved FT and three point shooting wasn't just a college senior converting against inferior competition. Clark is a tough, hard nosed and high IQ player, who has a chance to carve out a role in the league.

Jared Vanderbildt – I am always intrigued by the modern model and who pops when we focus in on the newer players whose scores were earned in the modern era. JV fits that bill and is another intriguing player at #37. Another player who is a jumper away from having a defined role. JV is a 6’8’’ defender who can defend anywhere on the perimeter and many bigs. A fun, high motor defender who hit 0% of his 13 attempted threes and 63% of his FTs. He may win the annual JP Takoto award for player who looks like an NBA player except for a dreadful shot.

For the Kings, the answer may be clear. While I am not sure about Doncic, I would be overestimating my abilities to pretend I could really put his European game film into context, he seems to have the minimum amount of athleticism and the size needed to be a legitimate NBA player. I think his floor is very high and very conducive to winning basketball and his ceiling is just as high as anyone in the draft, unless you only equate ceiling with pure physical attributes.

I am much less bullish on Bagley. I am not against him and think with a lower pick, he offers the right amount of reward to justify the risk. I just find the amount of development he needs a bit risky for my tastes at #2 for even a Generic NBA Team. I find it even more daunting when it’s on Vlade, Joerger, and their staff to develop him given our track record. With Bagley, if he improves somewhat, you run the risk of having a very unique player you have to build a roster around to cover his deficiencies, but a player whose impact on the court doesn’t merit those resources. But at #2, there are too many players with near equal potential, but much safer floors. So at #2, I do believe we should "just draft Doncic."

At #37, you never know who slips, but there’s a number of interesting names including Okogie, Evans, Carter, Clark and Melton who could push Mason or our wings for minutes and offer a lot of upside in the second round.

Post your thoughts. I am happy to answer questions or hear if here's another interesting interpretation of the data someone's knowledge of these players or scouting can add to the conversation.

(This is a FanPost from a member of the Sactown Royalty community. The views expressed come from the member, and not Sactown Royalty staff.)