Touches -- The Overlooked Predictive NBA DFS Advanced Stat
Consider for a moment some of the statistics you hear most frequently used when describing a player's outlook on any given night of DFS. Some that probably come to mind are usage rate, minutes, or per-minute Fantasy point production. Whether you get your information through podcasts, articles, or daily Fantasy tools, the stats that are referenced remain similar.
Any volume-based stat that correlates strongly to Fantasy points is one that we should pay attention to. I'm not entirely sure why a stat like usage rate is thought of as the most relevant for DFS purposes, while most people make no mention of a player or team's touch breakdown. A bit of digging will show that touches clearly correlate more strongly than usage or other commonly referenced stats.
Per NBAstuffer.com, "touches estimate the number of times a player touched the ball in an attacking position on the floor. The theory behind the formula is that once a player gets the ball, he can only do one of four things (aside from dribbling, of course): pass, shoot, draw a foul, or commit a turnover."
This Rotoworld article breaks down the correlation between usage rate and Fantasy points, as well as the correlation between Fantasy points and counting stats like points, rebounds, and assists. Usage rate was shown to have a relatively high correlation (0.7069) with Fantasy point value.
Using 226 qualified players from the 2018-19 season, I found the correlation between a player's touches and their DK points to be higher (0.8597) than any stat other than a player's actual points (0.867). You can find player touch data updated daily on NBA.com.
If touches correlate so strongly with Fantasy production, why aren't more people using this free and readily available data? I believe the simple answer is that it is not easy to keep up with the touch trends for teams on a daily basis. We don't have convenient tools that quickly tell us how a player's absence affects the touches of other players on their team. Not many are willing to take the time to do the digging themselves, which leaves touches as a glossed over stat, which in turn leaves it as a great way to gain an edge over the field in NBA DFS.
That article is meant to be something you can reference whenever you need to throughout the week. Whether it is for help when building DFS lineups or making season-long decisions, this can be your one-stop source for relevant advanced statistical information on every team. The plan is to update it any day there is a "small slate" of four or fewer games (usually Tuesdays and Thursdays, plus the weekend). You can check the timestamp on each team to make sure you are looking at up-to-date info. If you have any questions about any of the info here, feel free to hit me up on Twitter.
Players Underperforming (The average rate was 0.5552 last year) in DraftKings points per touch:
Jeff Teague, Minnesota Timberwolves (0.396 DK points per touch)
Teague's number was low (.432) last year too, but this is just ridiculous. One problem is that Minnesota has shot poorly to start the year, so his potential assists are being converted at a low rate (49%). Aside from that, it looks like it has just been poor luck on Teague's part. His rebounding rate, assist rate, and three point percentage are all slightly down. His per-touch numbers should normalize over a larger sample size.
Chris Paul, Oklahoma City Thunder (71.1 touches per game -- 0.447 DK points per touch)
Paul averaged 0.546 DK points per touch while playing alongside James Harden last year, but he has looked mega-washed to start this season. His touches per game are almost identical to last year, but he's not contributing nearly as much as a rebounder or distributor. I'd expect him to settle in around the league average over a larger sample, but it is clear that Shai Gilgeous- Alexander is the guard to own in this backcourt.
Dennis Schroder, Oklahoma City Thunder (73 touches per game -- 0.416 DK points per touch)
Schroder actually leads OKC in touches per game. But, because his shooting hasn't been there to start the year, his DK points per touch is down from 0.511 last year to 0.416 this season. His touch-implied DK average is closer to 40 points than the 30 he's averaging, so be sure to target him at his diminished price tag.
Lou Williams, Los Angeles Clippers (7.8 touches per game -- 0.496 DK points per touch)
Sweet Lou leads the Clippers in touches per game, but he hasn't been his usual efficient self. Williams averaged 0.571 DK points per touch last year, but poor shooting and a diminished assist rate have dropped that number quite a bit to start the 2019-20 campaign. His usage rate is still right around 30 percent, and I'd expect the other two numbers to bounce back as the season goes on. His recent struggles make him a great target in DFS. He should be averaging 40.9 DK points per game if his point per touch number was at the league average, but he's instead averaged just 36.5.
Fred VanVleet, Toronto Raptors (84 touches per game -- 0.436 DK points per touch)
VanVleet averaged just 0.429 points per touch last year, so it's possible that he's just an outlier when it comes to touch-to-point ratio. Still, he is shooting poorly and underachieving in the assist-to-potential assist category, so I would expect some positive regression with a much improved role. Last year's 0.429 mark came while playing alongside Kawhi Leonard and filling a bench role. VanVleet now is a solidified starter and one of Toronto's best offensive initiators. Now with Kyle Lowry sidelined for multiple weeks, VanVleet's role should yield far more than 0.436 DK points per touch.
VanVleet has averaged 102.5 touches in the past two games with Lowry sidelined. Even at 0.436 DK points per touch, that would put him at close to a 50 DrafKings point projection on a nightly basis. His salary is still floating around the $7k range, but the touch data suggests that it should be closer to $9k while Lowry is out. Be aggressive in targeting him until the price tag catches up.
Pascal Siakam (75.9 touches per game -- 0.625 DK points per touch) has actually been outproducing his touch-implied point totals. I bring him up just because he also will see a huge boost in touches with Lowry out. He's averaged 91 touches in the past two games, along with 9.5 potential assists -- up from his season average of 6.7 per game. Siakam projects as a $10k player with his current volume of touches.
Ricky Rubio, Phoenix Suns (77.4 touches per game -- 0.499 DK points per touch)
Rubio is going to be a perennial member of this list, just because he dominates the ball. His touches are always going to be high for a non-superstar level player. But there are definitely some trends worth noting with his early season usage.
Not only does Rubio lead the Suns in touches per game, but he's 17th in the entire NBA. He's eighth in the NBA in potential assists per game and is taking the ball out of Devin Booker's hands a lot more than we've grown accustomed to over the past few years. Booker is averaging just 53.2 touches per game this season, after leading the team with 75.7 last year.
Rubio's touches are trending up, too. He's averaging 85.5 touches over his past four games, as opposed to Booker's 51. He's beginning to look like an $8k player on DraftKings unless he sees a role change.
Players Overachieving (The average rate was 0.5552 last year) in DraftKings points per touch:
As a general rule, big men are going to overachieve when it comes to DraftKings points per touch. Guys who get a lot of their DK production from rebounds and blocks don't need as many touches to put up big Fantasy totals. So, some of the most extreme outliers (Karl-Anthony Towns/Anthony Davis/Andre Drummond/Joel Embiid) won't make this list because they are likely to continue being an outlier.
Kawhi Leonard, Los Angeles Clippers (65.4 touches per game -- 0.819 DK points per touch)
Leonard averaged 0.793 DK points per touch last season, and his usage rate is significantly higher this season. So, it isn't a huge shock that he's averaging even more DK points per touch this year. However, if the usage rate normalizes, Kawhi could be someone who is overvalued as the season goes on. Lou Williams actually is averaging more touches than Kawhi, and Paul George will be sure to cut into that number some when he returns.
Damian Lillard, Portland Trail Blazers (82.1 touches per game -- 0.659 DK points per touch)
Lillard's rate (0.559) was much more in line with the league average last year. His touches are down slightly from last season, but an inflated usage rate, true shooting percentage, and assist rate have caused him to be more efficient with his touches than we are accustomed to. This means one of two things. Lillard's touches could return to the levels we saw over the past two seasons, in which case he would become a fringe $10k type of player. There's no real reason for his touches to be down, especially with Portland ranking in the top-10 in pace. The other possibility is that Lillard's touches remain where they are and his efficiency rates normalize, which would mean that he is overvalued at the moment.
Russell Westbrook, Houston Rockets (75.1 touches per game -- 0.654 DK points per touch)
Westbrook is always going to be an outlier. He averaged 0.646 DK points per touch last season, so he's not overperforming much by his standards. I featured him to point out that his touches are down to 75.1 per game while playing with Harden. Russ has been top-two in touches in each of the past three seasons. His lowest per-game average during that time was 91.2. If his efficiency comes down at all, Russ could be drastically overvalued in DFS given his dip in touches.
Kyrie Irving, Brooklyn Nets (81.4 touches per game -- 0.64 DK points per touch)
Irving's touches are up from 76.8 per game last year in Boston to 84.6 as a member of the Nets. The increase in pace definitely is a contributing factor. I believe he can sustain the increase in total touches, but I am more skeptical of his DK point per touch rate. He averaged 0.587 last year, and if his career-high 33.2 percent usage rate and 38 percent assist rate drop at all, his 0.64 DK point per touch mark is going to see a definite decline.
Potential Assists
Another advanced stat that will be referenced frequently in that article is 'potential assists,' which is "a situation where a ball-handler make a pass to a player who takes a field goal attempt within the determined amount of time an effort required of earning an assist, called an assist window." Potential assists had a 0.9897 correlation to actual assists last year, which should come as no surprise. With a year's worth of data, a player's assist-to-potential assist ratio usually is going to balance out.
Through the first couple weeks of this season, the correlation between potential assists and actual assists is a bit lower (0.9546), but is still extremely high. The correlation is so strong between the two stats that the number has normalized already, even with just a few games worth of data for most teams. But, the standard deviation is significantly higher. The standard deviation of potential assists to assists was just 0.0193 last year, meaning that most players' ratio normalized over a season's worth of data. So far this season, the standard deviation of potential assists to assists was 0.1867. That is a massive difference, and it should come as no surprise, since we only have a handful of games to collect data from on a player-by-player basis.
This is important, because it means we can still uncover value in players who are underperforming in the assist-to-potential assist ratio department.
Be sure to check the 'Early-Season NBA DFS Guide: Advanced Stats for Every NBA Team' before setting your daily Fantasy basketball lineups. Also, be sure to follow me on Twitter, where I'll post the Advanced Stats Guide any time it gets updated.
Share This Story