Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

grantland.com

Post reply on HN
Print view View original

11–20 of 31 posts

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#11

Post by mmc »

balakk wrote: post #4

The camera presumably is overhead. From that perspective, what could be a useful feature for person tracking? I am assuming it wont be that easy to pickup just numbers of their jerseys. Faces are also not easy to get from that angle. Or do they have multiple cameras in place?

From this article: http://www.sportvu.com/PDFs/ESPN_62011.pdf

"Three cameras on each side of the court..."

This reminds me of a talk I heard a long time ago from Takeo Kanade at CMU, who was working on reconstructing 3D scenes from multiple cameras shooting different angles of the same action. One example was basketball, and he had a demo showing a play reconstructed as if the camera was on the ball. Pretty cool stuff. It needed lots of cameras, though, and I remember asking him if they had explored how well you could do with a minimum # of cameras. I remember his reply being essentially that they were less interested in bad results with few cameras :)

Here's a link to some surprisingly old press about his work being used for Super Bowl 35 in 2001: http://www.ri.cmu.edu/events/sb35/tksuperbowl.html They used 30 cameras.

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#12

Post by nickalewis »

the_watcher wrote: post #6
mmc wrote: post #2

The "ghost" players in these animated replays that show where the system thinks a player should have been are fascinating to see. I wonder how automatic it is - if games could be analyzed like that without much expert input, this would be a really cool thing to have for recreational teams who want to get better but can't afford expert coaching. I'm thinking my late-night adult league hockey team would probably benefi…

Unfortunately the cameras are pricey enough that only 15 of 30 NBA teams have purchased them, not to mention the half decade of work it took the Raptors to build the code to actually generate the visualizations. Maybe someday it will be commercially available and affordable, but that day seems a long way off to me.

I don't think it's the cost that's keeping other NBA Teams from purchasing ($100K/year is pennies to an NBA franchise). I suspect its more of a front-office vs coaching thing. Adopting this as part of a teams core prep takes education; Effectively learning how to understand this type of data and apply it in practice over the course of a season.

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#13

Post by ssharp »

the_watcher wrote: post #6
mmc wrote: post #2

The "ghost" players in these animated replays that show where the system thinks a player should have been are fascinating to see. I wonder how automatic it is - if games could be analyzed like that without much expert input, this would be a really cool thing to have for recreational teams who want to get better but can't afford expert coaching. I'm thinking my late-night adult league hockey team would probably benefi…

Unfortunately the cameras are pricey enough that only 15 of 30 NBA teams have purchased them, not to mention the half decade of work it took the Raptors to build the code to actually generate the visualizations. Maybe someday it will be commercially available and affordable, but that day seems a long way off to me.

I'm wondering why the 15 other teams aren't on board yet. If the price is only 100k, it's nothing to an NBA team. If those other 15 teams don't already have an advanced analytics team in place, then it makes sense not to have the camera until you have a structure in place to make sure of the camera data.

It will be interesting with this huge push in basketball over the past couple years to see where this type of analysis leads. The NBA model right now is pretty simple: get two or three superstars, align them with solid role players, and play in championships. It's hard to imagine analytics pushing a team like Houston to be able to beat a team like Miami or Oklahoma City in a seven-game series because the talent gap is so substantial. Houston can play a smarter game but the other teams still function better. What I could see happening, however, if a team like Oklahoma City, while slightly less talented than Miami, can implement insights from this analysis to raise their EV, it might overcome the slightly smaller talent gap. Even in that small sense, it makes more to spend a couple million a year on analytics than it does to pay an average NBA player several million. Since the NBA has both a salary cap and maximum contracts, it pushes the price of an average play up quite a bit, so better value may be had by raising the effectiveness of your current talent than trying to bring in new players.

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#14

Post by the_watcher »

ssharp wrote: post #13
the_watcher wrote: post #6

Earlier quoted context omitted.

Unfortunately the cameras are pricey enough that only 15 of 30 NBA teams have purchased them, not to mention the half decade of work it took the Raptors to build the code to actually generate the visualizations. Maybe someday it will be commercially available and affordable, but that day seems a long way off to me.

I'm wondering why the 15 other teams aren't on board yet. If the price is only 100k, it's nothing to an NBA team. If those other 15 teams don't already have an advanced analytics team in place, then it makes sense not to have the camera until you have a structure in place to make sure of the camera data. It will be interesting with this huge push in basketball over the past couple years to see where this type of anal…

I completely agree, but, as referenced in the article, there is still pretty strong resistance to analytics-driven decision making in the NBA. Many NBA coaches and GM's are former players, who generally resist advanced stats. Hence "it can't measure personality, chemistry, heart." That quote fails to mention neither can anything else, so it makes much more sense to focus on what can be measured than on what can't (I'll concede personality should in some cases be taken into account if there are clear clashes).

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#15

Post by JoelSutherland »

mmc wrote: post #2

The "ghost" players in these animated replays that show where the system thinks a player should have been are fascinating to see. I wonder how automatic it is - if games could be analyzed like that without much expert input, this would be a really cool thing to have for recreational teams who want to get better but can't afford expert coaching. I'm thinking my late-night adult league hockey team would probably benefi…

It sounds like there probably wouldn't be enough data to generate the ghost players. According to the article they depend on the skills of the individual covers. (i.e. Player X is a 35% three point shooter so don't leave him in the corner to help on a cut)

The 82-game NBA schedule really helps to generate large enough sample sizes about players.

I could see it being potentially possible to just use average data, but if your rec league is anything like mine, talent ranges from a guy can credibly fake a three and then dunk after one dribble to guys that can barely get up and down the court.

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#16

Post by kevinthew »

To the uninitiated, and those who like sports analytics (specifically the NBA), you should absolutely follow Zach Lowe's articles on Grantland and elsewhere. He's the best sports writer on the subject, and probably the only coherent one at that.

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#17

Post by the_watcher »

nickalewis wrote: post #12
the_watcher wrote: post #6

Earlier quoted context omitted.

Unfortunately the cameras are pricey enough that only 15 of 30 NBA teams have purchased them, not to mention the half decade of work it took the Raptors to build the code to actually generate the visualizations. Maybe someday it will be commercially available and affordable, but that day seems a long way off to me.

I don't think it's the cost that's keeping other NBA Teams from purchasing ($100K/year is pennies to an NBA franchise). I suspect its more of a front-office vs coaching thing. Adopting this as part of a teams core prep takes education; Effectively learning how to understand this type of data and apply it in practice over the course of a season.

It should be pennies to an NBA franchise (when they are spending millions on players who don't play). Unfortunately, owners like Robert Sarver and Donald Sterling are notorious for refusing to spend on marginal upgrades like a quality visitors locker room, or a a full support staff. Even Paul Allen has a reputation for refusing to pay for a high quality training staff in Portland - probably a bit undeserved since he seems to spend on other important things and I can't imagine why training staff would be something he ignores. Also, there is a floor for player salaries that must be met.

The analytics vs. coaching issue is a real one too, but don't discount owners being cheap.

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#19

Post by wazoox »

Wow, memories. In 1998 I worked on the FIFA World Cup graphics, and two guys on the team were developing a camera player tracking and stat acquisition system. Back then, raw CPU power was a real limitation; the fastest PCs were PII 333 Mhz and they just couldn't track automatically 22 players from 4 cameras (one atop each corner of the stadium).

In the end the system was almost working, however we needed two people watching the TV stream and clicking on players to identify them when the automated system lost track. However it provided already tremendously detailed stats, but unfortunately nobody really cared :)

Re: The Toronto Raptors, SportVU cameras, and the NBA's analytical revolution

#20

Post by eumenides1 »

swanson wrote: post #3
mmc wrote: post #2

The "ghost" players in these animated replays that show where the system thinks a player should have been are fascinating to see. I wonder how automatic it is - if games could be analyzed like that without much expert input, this would be a really cool thing to have for recreational teams who want to get better but can't afford expert coaching. I'm thinking my late-night adult league hockey team would probably benefi…

It would be interesting - though it will be some time before the price comes down. The cameras to do the tracking cost $100k/year at the moment.

While I agree 100K is alot, it's nothing compared to salary and cap space. To be able to spend money outside of cap restictions to imporove your team is a stragetic advantage.

I don't think the price will come down, the only people who need it are wealthy and the wealthy can't afford to not have it.

Post reply on HN