The Reckoner - April 2018

Page 5

April 2018

P5

VOL. 6 NO. 4

EDITORIAL The Curry Effect

Analyzing the Analytica Affair

T

GRAVITY Continued from P1

OP-ED by EMMA WAI

he extent and nature of politically-motivatHowever hopeless the overall situation may ed online targeting has long been a topic of seem, there are solutions to smaller-scale isdiscussion, but as politics becomes increasing- sues that need to be addressed. Of the many ly divisive, the line to be walked on the use of problems uncovered in recently-released data, such tactics has drawn finer and finer. Now, as one of the more concerning was the revelation the investigation into Russian interference in that several agreements asked for permission the US election continues to build, with Britain to access friends’ data, and that no regulations continuing to reel two years after its decision to were in place to prevent them [2]. Circumstancexit the EU, the debate has seen sudden atten- es where one waives the privacy and rights of tion thanks to a whistleblower. The information another should not exist. Though such circumreleased details substantial leaks of private data stances were created through user negligence, in attempts—“attempts”, we are told, though in the law should serve to protect the average perreality, they were likely far more effective than son. mere attempts—to influence both. In an ideal world, games and applications That’s where Cambridge Analytica comes would provide specific explanations as to why into play. Concerning numbers revealed that 87 they need certain permissions, Facebook would million Facebook users had their data accessed limit the permissions that apps can demand, or shared with the data-based consulting firm and users would actually read the agreements [1]. Further controversy emerged with reports they sign. In reality, the average user is unlikeof gross misuse of this data; the data accessed ly to examine permission agreements letter by was used to create psychological profiles intend- letter, at least partly because they are unnecesed to change audience perceptions sarily lengthy. There needs to be laws ensuring and behaviours, eventually emsuch text is simple and understandable. ployed in support of now-PresSuch actions have precedent in the ident Donald Trump during forms of food labels—in Canathe 2016 US Presidential da, severe allergens must be Election and the winning declared in a manner that is ‘NO’ campaign during the easily understandable to the Brexit referendum. everyday shopper [2]. This It wasn’t shocking means that if a food contains when we found out Faceeggs, the package must say book collected our private the word “eggs”, rather than data, nor was it surprising when alternatives such as albumin, we learned it was sold, leaked, or livetin, or ovo [3]. given away to third parties, and Needless to say, food laeven less so when it was revealed belling and online agreedata was largely accessed thanks ments aren’t identical: to lacklustre security precauwhereas food labels are tions. In cases such as these, it physical, with a clear is our conscious choice to share point of transaction, the our data, or at least provide it internet’s global nature to companies like Facebook. prevents agreements from It becomes far more troubling, being overseen as local isand in fact severely problematsues. This again falls into ic, when it affects citizens who the issue of jurisdiction Illustration: Sheri Kim haven’t made this choice. Look when it comes to online inat Brexit, where the decisions made in their teractions, and at the current moment, there is country were reportedly influenced by foreign no clear solution. entities. So what is there to do? Little, except being Ideally, there would be laws or guidelines more careful with your personal data. But who protecting personal data as well as preventing knows? It might not even matter. For all you foreign influence in decisions of national sig- know, someone might have already signed your nificance. Given the severity of their offences, privacy away. ■ Cambridge Analytica should be penalized, as should Facebook for its negligence in prevent- Works Cited ing the breaches. Unfortunately, with a lack of [1] https://www.reuters.com/article/us-facebook-privainternational authority to regulate data securi- cy/facebook-says-data-leak-hits-87-million-users-widenty, we are forced to rely on the morals of private ing-privacy-scandal-idUSKCN1HB2CM companies to protect confidential information. [2] https://www.wired.com/story/facebook-privacyBut Cambridge Analytica is just one compa- apps-ads-friends-delete-account/ ny, and as wrong as their actions were, they are [3] http://foodallergycanada.ca/allergy-safety/food-laevidence of a larger problem. Even if users take belling/ necessary steps to protect their data from third [4] https://www.canada.ca/en/health-canada/services/ parties, Facebook still has access to it—and food-nutrition/reports-publications/food-safety/eggs-priwith it, the privilege to provide it to the high- ority-food-allergen.html est bidder. The nature of online data collection Op-eds are opinion articles that reflect the views of the and selling ensures that companies cannot be author, but not necessarily those of the Editorial Board or trusted with our data, and yet, for those that of The Reckoner as a whole. Please note this important diswish to use online services such as Facebook, tinction when reading this article. they are left with no other option but to continue without legal protection.

his teammates even marginally better on the court. Even in his highest scoring games, Westbrook’s Gravity Score trails Raptors star DeMar DeRozan which, admittedly, is only true for this year. The evolution of DeRozan is highlighted here. On his best nights, he creates shots for his teammates at an equal or greater rate than Rockets duo James Harden and Chris Paul, a vast improvement over last year, where his greatest Gravity Score was a mere 1.7. While it is true that the Gravity Score can easily be influenced by how good your teammates are, not even Westbrook can complain about having a bad team this year.

The graphs above highlight the performance of Curry and Westbrook’s teammates when each point guard is on or off the court. The usage rate conveys how often the play ends with a specific player, while the effective field goal percentage (eFG%) reveals how well each player shoots. A good facilitator should naturally make his teammates better when they are on the court. For both Westbrook and Curry, we expect each of their teammates’ eFG% to decrease when the two point guards are off the court. While this is true for Curry, the opposite occurs for Westbrook. Westbrook makes his teammates better by not playing. His four assist lead over Curry is essentially a moot point. Assists don’t tell the story of how a team is benefitting from a player’s presence. Open shots, effective field goal percentage, and close defender frequencies do. But what happens when Curry isn’t catching fire from beyond the arc? In games where he’s only scoring ten points, does his gravity disappear with his shooting stroke? Or is the threat of his three-pointers frightening enough to warrant the same attention? In a nutshell, it’s a bit of both. Gravity is a somewhat polarizing stat since there is so much variability in the defense’s movement. Is a defender’s movements really a product of gravity? The calculation algorithm is a bit finicky; it doesn’t accurately depict a player’s impact on his teammates, only that a defender is, say, a few centimetres closer than usual. The stat is generally riddled with uncertainty. A simpler approach would just be to create a score based on eFG%, wide open shots, and closest defender frequency. And that’s exactly what we did. Introducing The Reckoner’s Gravity Score.

Within the confines of the list, Westbrook’s Oklahoma City Thunder have a collective Gravity Score of 9.8, similar to the eastern powerhouse Boston Celtics, while also leading the Cleveland Cavaliers, Minnesota Timberwolves, Philadelphia 76ers, as well as the New Orleans Pelicans, each by a significant margin. Westbrook also contributes the lowest percentage of gravity to his team’s starting five, with a mere 7.2%. His teammate, Paul George, by comparison, contributes 29.5%. If Paul George and Ben Simmons can make their teammates better, Westbrook has no excuse. It’s simple to explain if you watch both Westbrook and Curry play. Westbrook is often reluctant to move off the ball. The Thunder’s offense revolves around a series of isolation plays for their three stars. When Paul George has the ball, for example, Westbrook often clears out of the way, opting to stand pat at the half-court line. That’s not to say the Westbrook doesn’t have gravity. He clearly does, as teams respect his scoring ability. But imagine Westbrook as a strong magnet. Now place that magnet two hundred kilometres away from the target. Will the magnet pull the target closer? No, it won’t. Curry, on a Durant isolation play, instead takes an active role in the offense. Every second, he’s calling for a different off-ball screen. Since the defense has to track Curry’s position at all times, it takes attention away from the actual play and directs the defense’s focus onto him. This way, both the defender and his teammates are distracted, making it a true isolation play. What also caught our eye with the Gravity Score is how we can compare scores during bad games. By utilizing an outlier-determining method, we could plot out an average gravity over the stretch of games where performance was below expectations. Note that some players (LeBron James, for example) aren’t listed simply because they are consistent enough that none of their point totals can be considered outliers.

A Fragile Balance by GRACE XIONG

The Gravity Score relies solely on how well the team is playing when each player is on the court. A high Gravity Score essentially means that a ton of open opportunities are being created when that player plays. We decided to plot both the NBA scoring and assist leaders to find out just how much impact they have on their teammates. It’s interesting to note that, for most players, the Gravity Score is heavily correlated with how much each player is scoring. The diamond-shaped marker, which denotes the Gravity Score for each player’s lowest scoring game, is mostly found in the lower half of the score range. This makes sense, since players on an off night will not require the same defensive attention from an opposing team. Surprisingly, Draymond Green trends in the complete opposite direction. His lowest scoring games constitute those where he has had the most impact. Clearly, a player can do more than just score to make his teammates better. In that regard, even Draymond Green is a better facilitator than Russell Westbrook. Notice how Westbrook lags behind most other players in terms of gravity. His presence does not make

www.thereckoner.ca

The results shouldn’t be surprising. In games where a star scorer doesn’t score (Kemba Walker, Blake Griffin), his gravity is generally low. In comparison to a facilitator (Simmons), a skilled passing big man (Nikola Jokic), or even a sharpshooter (Kevin Love), these players provide more value even when they’re not scoring because the threat of a pass and the threat of shot space is more threatening than the threat of isolation baskets. The only outlier, in this case, is probably DeRozan, who is making a conscious effort to draw his isolation defenders to one side while eying a pass to another. DeRozan is also benefiting from excellent teammates and his improved ability and greater tendency to pass this year has taken his impact to another level. Clearly, not all assists are equal, at least, in terms of impact. Facilitation of an offense involves more than just mindless passing. It’s about putting your teammates in space so that they can score, so that they have the time and shot space to succeed. ■ Charts: Varun Venkataramanan


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.