How to Bart ethically: Or, 15 2023 statistical queries nobody asked for
Barttorvik is one of the more intriguing tools available to draft analysts, potentially allowing for the construction of a road map to a player’s future out of college. But how do you use it properly?
Part 1: RULES OF THE ROAD
Barttorvik.com is one of the most indispensable websites for a hoop nerd. Back when I cared more about bracketology accuracy, it was my favorite source of information for team performance, filtered through different models and categories, using analytics to drown out the noise. Nowadays, it is equally useful but put towards a different cause: NBA prospect comparisons.
On this site, you can take a library of every college player of the last 10 years and narrow the field by 2-point percentage, block rate, free throw percentage, class, conference, and a hundred other categories. At its simplest, it is a tool for looking at what NBA players — or players who never made the NBA — were, as NCAA players, similar to a current prospect. For example: take 2023 Duke center Dereck Lively II. I can look for freshmen and sophomores, 6-foot-9 or taller, who offensive rebound a lot (10% or higher OREB%), get steals (1.5% steal rate), block a TON of shots (10% block rate), and have an overall defensive impact that is pretty good (Defensive Box Plus Minus of five or higher). It yields a list featuring Lively as well as other standout NBA bigs who defend well and, unsurprisingly, defended and rebounded well in college.
A pretty cool way to confirm that there is a decent probability that Dereck Lively will be a good defensive big in the NBA! It’s no guarantee of course, but it is helpful data.
However… with great power comes great responsibility! Like any statistical metric, the Barttorvik player comparison database can be misused to create misleading comparisons. You can tweak and tweak and tweak, add filters indefinitely until all of a sudden you get a comparison that gives you great optimism. Take this, for example:
Using over filters, some of which are oddly specific (why a usage of 26 versus a more even number like 25?), I have created reason to think that likely late first rounder and Pepperdine sophomore wing Max Lewis is destined for indubitable greatness!
You can do this with most players. Here’s one for Deuce! Similar to Otto Porter, James Harden, and Zion!
Now, I love Deuce McBride. But using this query to imply he has some shot at Harden or Zion levels of greatness is probably a bit wild on its face. Hence the title of this piece: we must use this tool ethically and responsibly.
There’s no singular right way to do it. Some people I know like to use it to input a bunch of different stats a player accumulates to generate a larger list of 25 or more players of different kinds, many of whom contain different “elements’’ of the target prospect. For example, friend of the site Chuck of the Chucking Darts Draft Pod used the query below to get to the heart of Brandin Podziemski, a mid-major breakout sophomore who was a splashy 6-foot-5 shooter and offensive engine for Santa Clara this year. He shoots very well, passes very well, and plays below the rim. If you know Brandin, you’ll see elements of many of these guys below – both stars and non-stars – in his game.
Me, personally? I have one overarching goal when I Bart search, and two guiding rules.
The goal: to compare current prospects to NBA players who resemble good or great versions of that prospect in the future.
The two guiding rules I use are: 1) try to focus on the stats which define a player’s positive uniqueness, and 2) try to use as few filters as possible.
For rule No. 1, you can also do the inverse – use negative stats to show why players with X, Y or Z red flag stats do not make it into the league, or aren’t very good in the league. Take Jett Howard. He’s been mocked everywhere from inside the top 10 to just outside the lottery. And yet… very few players have Jett Howard’s height and awful rebounding and stock numbers. You can use a filter I don’t often use — which is set it to show drafted players who reached your input thresholds (as opposed to all NCAA players), and do something like this to show the only other player to be as bad at rebounding and steals as Jett who still got drafted:
That’s right, none other than Sauce Castillo himself! He was a top 10 pick and had some eerily similar numbers to Jett: slightly positive assist/turnover ratio, elite shooting numbers, solid BPM. Alas, he was never destined to live up to his lofty promise. If you broaden it out to top 60 picks, you get a handful of other nondescript guards like Marcus Thornton and Doron Lamb.
I don’t really use Bart for negative searches, because when it comes to red flags for prospects, we usually don’t need fancy searches to know that something is very troublesome and risky. That’s the nature of red flags. It’s very easy to find reasons to rule players out as less safe, and when a front office is choosing between a handful of favorite prospects, they definitely should consider that stuff. But from my couch, I am not debating which of 2-3 available prospects my team is taking — I am trying to get a grip on 30-50 guys and see what they might become. I’m at Baskin Robbins, trying to sample all 31 flavors for funsies. This is an exercise in optimism and possibility for me. Again, that’s a personal choice. After all, Bart searches are but a tool. Do what you will with them!
PART 2: BUCKLE UP, WE GOING BARTIN’
Now that we know the rules of how to Bart search ethically, I would love to share with you some fun Bart search queries that get to the core of some 2023 NBA Draft prospects. For each prospect, I will present at least one Bart search with a brief commentary. My goal with these, as I said in Part I, is to use a prospect’s strengths to identify players who had similar strengths as prospects who turned out solid (or better!) in the NBA. And as a refresher, here are my two rules of the road for these searches:
1) Try to focus on the stats which define a player’s positive uniqueness, and…
2) Try to use as few filters as possible.
And one final caveat: I did some of these in the first few months of 2023, so players’ final stats may have shifted a wee bit for those earlier searches. And with that, let’s get to it.
Dariq Whitehead
Dariq has had an up and down year, marred with injuries and rehab both before and during the season. As a result, his advanced stats are mostly mediocre to outright bad. This is especially unfortunate, as his high school tape is the stuff of legend and why he was arguably the top recruit in the class. However, despite his bad luck on the health front, he gutted out an important role for Duke: the floor spacer and bailout shot taker. Despite him only scratching the surface of his potential, the core brand of Dariq remained strong at Duke, as this query shows: he’s still a true wing who shoots the fucking lights out as a freshman, not unlike AJ Griffin and Klay Thompson, even as his game and body are still rounding into form.
Julian Strawther
This query is longer than I would like, but each added filter has some intention. The goal was to capture the efficiency at the heart of Strawther’s game. His offense is defined by it — he’s never had a true shooting percentage (TS%) below 60 at Gonzaga, and is a lights out shooter from three on high volume, as well as an efficient long 2-point shooter and floater shooter who can also finish at the rim. And he has wing size. There’s some real talent on this list, and Strawther — despite being a junior — is only 20 years old and younger than most of the list.
Noah Clowney
This one is simple. Clowney’s defining skills are his ridiculous rim protection from the 4 and 5 spots, and his semi-theoretical shooting. So this search is a look at freshmen and sophomores who take threes, block a lot of shots, and have high impact metrics. The list is pretty short and pretty impressive!
Taylor Hendricks
This search is similar to the Clowney one above, except I ratcheted up the 3-point rate numbers even higher to reflect Hendricks’ more advanced comfort taking lots of threes. The result? A rare two-person list including Tay himself and none other than Sacramento sharpshooting big wing Keegan Murray.
Brandin Podziemski
Brandin Podziemski has some pretty insane stats, so I decided to go for a search which reflects that. His box plus minus (BPM) impact is among the highest in the class this year at over 10, so I searched for younger prospects who had similarly crazy high BPM, elite 3-point shooting, and a also willingness to shoot long twos (to weed out shooters who can’t put it on the deck at all). Finally, I also added in a 1.5 assist to turnover ratio threshold to incorporate his high feel as ball handler. The results here speak for themselves — there might be a special player lurking within the Santa Clara guard.
Cason Wallace
Cason is a ridiculous defender, one of the best in this class. His steal rate is among the higher ones you’ll see for a guard. His defense is also reflected in his BPM, which is pretty solid. Combine that with a decent willingness to shoot threes and, well, you see a few different analogs of 3-and-D guards with off-the-dribble skills ranging from role players to stars.
Colby Jones
Wing-sized defenders who can pass well and are willing to shoot are NBA currency. The best teams have them in spades. Colby is, clearly, a solid bet to snag one of these players in the late first or even the second round. Teams, take note!
Anthony Black
Ant is an uneven player. He can’t shoot, but his free throw rate blows everyone else in this class out of the water. He has no touch, but he processes the game better than even the point guards in this class. He’s skinny, but deceptively strong and athletic. This is one of the rare occasions where I included a few negative parameters — specifically, a lack of willingness to shoot and a poor rate of converting shots — along with some of his positives like assist rate and strong BPM. The appearances of De’Aron Fox, Ja Morant, and Kyle Anderson give credence to the idea that despite the ups and downs of Ant’s game, we should ride it out and see where the strengths eventually take us.
Cam Whitmore
Simple queries are my favorite, as they’re both clear and precise. How’s this: freshmen who dunk a LOT and shoot threes a LOT. The list is pretty short, which shows how hard that is to pull off in your first year of college play. You have to be athletic as fuck, and have the skill and clout to shoot a lot of threes. There are concerns about Cam, but he is definitely a hooper.
Brice Sensabaugh
Brice is probably my favorite player in this class, so I had to include multiple searches. The first one is possibly my favorite in this whole piece: extremely high usage, extremely high production freshman. The second one is just age and production. Both show clear star potential. Don’t overthink it, folks!
Kyle Filipowski
Flip is an insane statistical prospect. This tweet has a bunch of queries, because he’s just so weird and fun to compare to other prospects. Most 7-footers don’t shoot as often as he does, or get steals like he does. He compares statistically to wings and guards in that respect. Yet he rebounds and draws fouls like a 7-footer, which places him at a sparsely populated weird middle of a hoops Venn diagram of bigs, guards and wings.
Jordan Hawkins
Hawkins’ 3-point rate is absurdly high, among the highest you’ll ever see. He shoots threes off screens more than anyone in the nation, and yet unlike most college shooting specialists, he has decent overall impact metrics and isn’t a stranger to drawing fouls in the paint. This query reflects that, and shows a few NBA guards who utilized their shooting gravity similarly in college.
Jett Howard
Finally, Jett. I wanted to pull a list of not-short players who had his combination of high feel (assist to turnover ratio), insane willingness to pull the trigger from three and the athleticism to dunk a bit despite being arguably a shooting specialist.
Hopefully you found these a little bit insightful and can now envision versions of these players who are successful. The top of this draft class features an unusual number of non-NCAA players in Victor Wembanyama, Scoot Henderson, and the Thompson twins, Amen and Ausar. Elite talent often speaks for itself, and in my opinion digging into those slightly less certain players, who might contain hints of stardom if you squint hard enough, is the most fun part of draft analysis. Bart searches are an excellent way to explore that, and hopefully now it will be something you all know how to enjoy too.
