Truth is often stranger than fiction, and basketball is no exception, especially when it comes to predicting performances. Can NBA 2K predict NBA athlete performances, even for our own Australian All-Star, Ben Simmons?
Predicting stats is nothing new. For example, FiveThirtyEight has been in the business of NBA predictive modelling since 2015, and its latest system, RAPTOR (Robust Algorithm using Player Tracking and On/Off Ratings) attempts to measure impact players have on both offence and defence, by points scored (or limited). This season, the team at NBA 2K made its own projections on how the 2019/20 season could fare, by using their MyLEAGUE mode to get the official NBA 2K season sim right before season tip-off, to ensure roster accuracy.
So far, one prediction has been tracking well: a potential Los Angeles Lakers-Milwaukee Bucks Finals matchup. Some other thoughts include a scorching hot take of a 4-0 Lakers series win, LeBron James as Finals MVP, and Giannis Antetekounmpo as repeating MVP.
"While we are just a few weeks into the NBA season, the Lakers and Bucks are among the best teams in their respective conferences and could well be on track to play in the finals, like we predicted," an NBA 2K spokesperson said. "The duo of LeBron James and Anthony Davis are definitely living up to the hype and Giannis [Antetokounmpo] is certainly in the MVP consideration at the moment as well."
Team 2K also had an attempt at predicting Ben Simmons' numbers this season - were they right?
"We put a lot of work in the AI of NBA 2K each year to ensure the teams play like the real life versions," NBA 2K also said. "As a result, we feel our simulation are an accurate representation of what could happen. We can’t predict injuries, trades and suspensions which could impact a team or player’s performance but it will be interesting to see if the Lakers do win it all this year and to see how accurate our other predictions turn out to be."
Here's the moment of truth.
NBA 2K prediction: Simmons doesn't make a three this season
Here's a Simmons take that didn't age well. The team at 2K probably expected the point guard to not take (or rather, make) any three-pointers this season, based off his reticence on shooting in earlier seasons. The 76ers guard has at minimum, broken that take with his first regular season three-pointer in November. He's 2/5 for the season so far, 26 games in.
Simmons isn't quite living up to Sixers head coach Brett Brown's challenge of taking at least one three-pointer per game yet, but it's a start.
NBA 2K prediction: Simmons averages 23.87 points, 7.57 assists, 9.13 rebounds
26 games in, Simmons is averaging 14.3 points, 8.2 assists, 6.8 rebounds in real life. The assists are close enough to NBA 2K's prediction, and rebounding numbers dipped because teammates Joel Embiid, Tobias Harris and Al Horford soak up a goodly chunk. The gap in scoring projection however sticks out.
Why isn't Simmons scoring more?
Here's the thing: Simmons is an efficient scorer. If anything, he became even better (56.3% last season, to 57.1% now). Efficiency doesn't mean anything if you're not taking more shots, and his shot attempts are slightly down (12.2 to 10.2), a logical outcome from being the playmaker on a championship contender.
Simmons is effectively Philadelphia's fourth option in the scorer's pecking order right now. Embiid leads both usage rate (31.1%) and scoring (22.8 points), followed by Harris (24%, 19.7 points), then Josh Richardson (21.5%, 14.7 points). Al Horford is slightly behind Simmons right now, and will likely hover between fourth to fifth - the veteran big man has always been an underrated do-it-all player who seldom posts eye-popping numbers.
To average nearly 24 points, Simmons would have to either be the first option on the Sixers' offence (that's a solid yeah nah, because Embiid), Or, he could somehow morph into the Australian Greek Freak (the Aussie Freak which doesn't have quite the same ring), and become a ridiculously efficient scorer who gets points every time he tries to score. Either he bulldozes for points in the paint, or soaks fouls and free throws at an insane level while using the same 10-12 possessions he's given right now (nope). Or, he could just splash tons of three-pointers (nope).
The scoring hasn't taken a jump, but Simmons' defensive efforts have most definitely been noticed. His length and instincts have allowed him to disrupt passing lanes, get a hand in for steals, and trigger his favourite transition offence.
Philadelphia is still fine-tuning its rotations and offence as the season progresses, but the situation isn't likely to change too much. Simmons had a season-high 34 point show against Cleveland in early December, along with a couple of 24-point nights (Boston in October, New Orleans in December). He might work his way up to a few more three attempts, but it's unlikely he will be a 20+ point scorer most games, especially with a healthy Embiid. Or, is increased scoring something we could actually look forward to this season from Ben Simmons?
The NBA 2K team possibly missed the boat on Simmons' scoring, but they could be right when it comes to the big picture predictions. Data analytics and predictions are only getting better with every successive NBA season, but if history's taught us anything, it's that basketball, especially those big moments are unpredictable - and that's why we love the sport.