Fighting Biases: Avoiding The Trap (Pt. 2)
Our second dive into best ways to counteract easy traps to fall into
Last week, we dove into a Business Review graphic making the rounds on NBA Draft Twitter about biases that are inherent in draft evaluation and how to best combat them. We were able to make it through ten topics on the graphic, provide a few examples and find ways to be aware of them while we scout so that we don’t fall into the trap.
We continue that premise by looking at the other ten biases mentioned on the graphic. Many of these are stretches a bit, far less frequent concepts you’ll face than the first ten, but are important to discuss nonetheless.
Before diving in, please consider reading Part One if you haven’t already, and checking out the graphic in-depth that we’ve shared below:
Outcome Bias
The example used beneath the graphic is a great one: gambling. Not just because the logic of the point makes sense: winning in Vegas doesn’t mean gambling is a good idea. But also because drafting prospects (and for teams, eventually investing in them) is a gamble. There are external factors at play that are outside of your control.
Vegas is still in business for a reason: casinos make money because they win more than they lose. Quite frankly, drafting prospects is going to be the same: we’re going to miss just as much, if not more, than we hit. We’ll overestimate guys who turn into duds. We’ll pass on guys who wind up being studs.
That doesn’t mean we shouldn’t necessarily play the game (quite literally, NBA teams don’t really have that choice). Instead, it means a realization that each outcome is its own, unique one based on the situation and the individuals involved. Just because Zhaire Smith and Jarrett Culver from Texas Tech were busts doesn’t mean Terrence Shannon Jr. will be. Smith had health issues derail his career and Culver’s shooting in college should have been seen as a flaw.
View Shannon as the next in line of Red Raiders to fail and you’ll do yourself a disservice. View him as his own prospect who will develop independently of the two before him and you return your focus to the process, not the result.
Overconfidence
Confidence comes by doing over and over again. Overconfidence is built by espousing the result over the process; that the person in question will achieve the result no matter what the process is like. As experts, we tend to think there are very few problems within our field we cannot solve.
To me, overconfidence comes in thinking that you can outsmart/ manipulate the process. In draft night strategy, it’s in how you construct trades to get the players you want. “We’ll trade back from 16 to 28 and 34 so we can get the guy we really like in the early second round”, but then the player you wanted gets picked at 22.
Placebo Effect
A placebo is a trick played on the brain, where you make it believe you’re getting the solution to problem when, in reality, the fix doesn’t actually solve anything but the issue goes away. Frankly, this shouldn’t be taking place in draft circles or front offices. If anything you have or do lacks substance and is solely about scoring some mental brownie points, it’s probably not an endeavor worth undertaking.
Pro-Innovation Bias
Hello, stats friends. Thanks for fighting the good fight.
As statistical models evolve, many have tried their hand at creating their own metrics to measure success and predict outcomes on the basketball court. Some of these methods have been productive — and many are proprietary so they never see the light of day on the internet.
Many of them are also inherently flawed and try to use data to describe parts of the game that are, by their very nature, immeasurable. Help defense and shot deterrence, making timely rotations… these are concepts that are tough to assign success or failure to and then even more difficult to properly value numerically. Be cautious of those models which are definitive in trying to offer all-encompassing defensive metrics for an individual, who is only one cog in the wheel of what goes into defense.
Front offices spend hundreds of thousands of dollars each year on analytics departments, staffers, technology, player tracking data and sports science initiatives. The goal is to try and gain an upper hand, and it’s a worthwhile endeavor — especially since there’s so much money going into sports nowadays. But a failure to identify the shortcomings of these technologies, their own blind spots and where they could easily fail, can lead to poor decisions that eventually pass the buck down to “well, the numbers suggested success here” when those numbers weren’t trustworthy in the first place.
Recency Bias
Perhaps none of the twenty biases listed above are more prevalent in the sports world than recency bias. In a ‘what have you done for me lately’ landscape, media coverage around the clock has expidited the process of critiquing athletes, general managers and coaches. The oversaturation of information available, and the frequency it’s debated, create a blur that makes events from two months ago feel like nine months in the past. Last week’s games have been so overdigested that, even if only one game has taken place since, it feels like a completed chapter in the past.
Fighting recency bias requires context and consistent work from all voices at the table. When I think about the negative impacts of recency bias, I think of guys like Malachi Richardson from Syracuse, who shot 37% from the field with a negative A:TO on the season and became the 22nd pick in the 2016 NBA Draft. Why did Richardson rise to those levels? 11-seed Syracuse made a surprise run to the Final Four, with Malachi spearheading the group through the tournament. The shock factor, and the fact he was one of the last prospects standing, made him appear as a trendy pick who was toolsy.
Here’s the catch: it wasn’t like Richardson was much better during this stretch than the season. In the NCAA Tournament, Malachi had a worse A:TO ratio (1.0 to 2.8) than in the regular season, shot 33.3% from two-point range and put up sizable numbers because he played the entire game, something players can do in the NCAAs thanks to the breadth of television timeouts.
Late-risers from the NCAA Tournament are always going to scare me. Sure, there’s value in having the conversation about players who perform best when the moment is biggest, an intangible feature nobody would deny on its face. But the risk in overvaluing that after the postseason is in falling to the recency bias trap. My suggestion: play a blind numbers game, removing the name from a prospect and looking at only their numbers and the differences between postseason and regular season. If the statistics match the playoff run, you might be onto something. If not, steer clear…
Salience
A tough one to grasp in draft circles, but it’s basically how noticable something is. If something stands out and hits you in the face, it’s likely to drive perception before the other factors can arrive. One example might be with shooting form. It’s an area I fell victim within the last two years on a guy like Tyrese Haliburton.
Haliburton sported an unorthodox shooting form in college. It went in a ton — 42.6% from deep over two years, and ultimately that is what matters. Because the form wasn’t standard and had its own hitches, our evaluation centered on Haliburton not being able to get away with the same shot on an NBA floor. It was slower off the dribble, his first step in attacking closeouts would be locked by how bow-legged his stance was, all the more.
Missing from my evaluation: the prerequisite touch that goes into being over 40% from deep no matter what the mechanics are. If the mechanics could be sped up, cleaned up and altered slightly to be a tad more natural, Haliburton’s feel would allow him to become a good shooter. And that’s exactly what happened.
Haliburton’s mechanics were the most noticable trait to his jumper, and they slapped me in the face as a negative. But in focusing too much on what was wrong with the shot, we did little to diagnose how to it could be right with a minor tweak or two. That’s why he’s shooting over 40% from 3 for his brief NBA career: he’s made those tweaks, and now our whole evaluation on him changes from a mid-first rounder to a lottery talent.
Selective Perception
In order to understand this phenomenon, I had to do a little more digging and research. Here’s what I came away with: this is how individuals perceive what they want to in media messages while ignoring opposing viewpoints. People tend to forget or ignore pieces of information (conveniently) that contradict their beliefs while clinging onto the data that reinforces their beliefs.
We’ve all seen prospects before who are wildly inconsistent. They’ll have great games with statistical outputs and then horrid performances where they barely see the floor. Those who championed Brandon Boston Jr. as being a first-round-caliber pick likely spent more time discussing his best performances (21 points on 7-13 shooting against South Carolina on March 6th) and ignoring his worst (0 points in 23 minutes vs. Mississippi State on March 11th). Boston had eight games with 15 points or more during his lone season at Kentucky, and seven games with six points or fewer. That volatility needs to be looked at on the whole. Choosing just the good or just the bad to try and reinforce your position or evaluation is the definition of selective perception.
Stereotyping
There’s a reason why one of our ten commandments of scouting basketball is to avoid player comparisons. Holding up a prospect next to a former one sets up an unrealistic expectation for that player to match up to and often looks only at one or two shared traits (because there’s really very few guys who share more than a handful).
We stereotype with generalities a ton beyond simple Player A to Player B comparisons. We assume European prospects are great shooters or have high levels of skill. That perception is rightfully being challenged more, especially as the model of European development is changing to include players who start at an older age or move to Europe in their formative years.
Even smaller words, such as adjectives, can be stereotypes in their own way. I think back to this scene from The League several years ago that pokes direct fun at the labels that get tossed around in draft circles:
Be careful for those convenient cultural comparisons that show stereotypes in action. The best way forward is to train yourself not to use them — and watching some satire like The League a little more often doesn’t hurt.
Survivorship Bias
History is written by the victor. Often times, we hear only the perspective of those who have made it to the end simply because they’re the ones who survive. As a result, we might be more inclined to think something is easier than it actually is because, well, the survivor was able to conquer the task and they’re who we hear from.
For this, I think about making it to be a video coordinator. We here these success stories about coaches like Erik Spoelstra, Frank Vogel and others. They scratched and clawed from hard work on the video room, starting as interns and the lowest guys on the totem pole, all the way to the top. It’s a romanticized tale, but it does little to illuminate just how many in those positions burn out and disappear in the pressure cooker of the NBA, or even those who fail to get those jobs because they can’t afford the unpaid labor often found with entry-level internships.
I think I’m guilty of doing this in one way with NBA draft scouting: teaching shooting. There are a few guys who have developed such a fantastic reputation (Chip Engelland, Jeff Hornacek, Dave Hopla, for example) in rebuilding jump shots that I tend to take the easy way out and say “man, I hope this non-shooting prospect winds up where these guys are on staff. They’ll get him to be a great shooter.”
Yes, those guys are some of the best at fixing shooting woes. But we only hear about them because of the public nature of their success stories. For every Kawhi Leonard, Tony Parker or Shane Battier that Engelland has built, there’s a Dejounte Murray who hasn’t found success. Look at all the ones who fit the mold, not just the triumphs that back up a certain narrative.
Zero-Risk Bias
Alright, so no prospect is truly a zero risk draft selection. But there are certainly those whose outcomes are more volatile than others. That can be due to myriad of reasons: high variance of outcomes from the player, positional versatility isn’t there, one swing factor to their production, off-court or character question marks.
From what we’ve experienced, the draft pick you’re trying to fill often times can indicate what level of risk you’re looking to take on. Those who have high picks are looking for a different balance of risk vs. upside than those who have second rounders; the consequences for missing out on your second rounder aren’t nearly as dire.
As a result, I’ve noticed a movement in online draft analysis circles over the last few years: the high-risk prospect wins out. It’s easy to sit behind our computer screens, say that JT Thor should be a first round pick because hey, those are the types of players you want to swing on. In reality, the consequences of swinging and missing are actually apparent — it’s just easy for us to dismiss because they don’t fall upon our shoulders.
What we’ll see is that sometimes back-against-the-wall front offices play the safer course, opting for instant impact from a rookie or a safer, low floor option who won’t someday get lumped into a hit piece on SB Nation calling for their firing. People’s jobs are on the line here, and naturally they’ll want to keep them. Often that means mitigating risk and treading the safer route. As the zero risk bias points out, such a pathway forward is not at all a guarantee of outcome.
This was part two in a two-part series looking at biases on the Business Insider infographic and their relation to NBA Draft Scouting.