Digitally Assisted Billiards makes everyone a pool shark

There's cheating, and then there's out-thinking the room. Georgia Tech's own Justin Needham and Matthew Straub are clearly headed for great things, as their Digitally Assisted Billiards is reason enough to give these guys a degree in our eyes. Using an array of low-end kit -- just an eBox 2300 embedded computer, Logitech webcam, 4- x 3-foot mirror, a VGA projector, pool table and a few extenders -- the two were able to concoct a system for visually displaying the expected trajectory of all possible shots based on the location of the balls still on the slate. For a much better idea of what we're hinting at, check the demonstration video after the break.

[Via Hack-A-Day]

[sabih @ Dec 11th 2008 9:20AM]

Its just like yahoo pool now

[Nerd Ferguson @ Dec 11th 2008 9:30AM]

haha, i was just about to say that.

[Konstantin @ Dec 11th 2008 12:09PM]

Too bad that even with the help of that thing those guys suck at pool. Honestly, they didn't even make one shot!

[OneLove @ Dec 11th 2008 4:56PM]

Run a "level 10" diagnostic on that bitch and calibrate it. (these guys gonna be rich)

[Vianon @ Dec 11th 2008 9:20AM]

Code Quantum?

[Vianon @ Dec 11th 2008 9:29AM]

Interestingly enough "Code Quantum" in France is "Quantum Leap" in the US, my bad ;)

[Pumapayam @ Dec 11th 2008 3:09PM]

I was thinking about the same episode regarding that show. Funny how some things that seem futuristic eventually come true.

[Speney G @ Dec 11th 2008 9:23AM]

This looks like the episode of quantum leap where Captain Archer has to play pool and Ziggy lays out how to take each shot.

[chefgon_ign @ Dec 11th 2008 9:44AM]

Sam Beckett

[pavlindrom @ Dec 11th 2008 9:48AM]

Nice to meet you. I'm Pavlo...

[Pumapayam @ Dec 11th 2008 3:10PM]

Same comment as above, great concept, and even crazier that a Quantum Leap "invented" this idea.

[Ryback @ Dec 11th 2008 9:25AM]

This reminds me of the "Step by Step" episode where Mark Foster won a pool game once he understood it was all about angles and physics.

[Wwhat @ Dec 11th 2008 9:31AM]

The cameraman might needs some chemically assisted relaxing, or a camerastand.

[NickNick @ Dec 11th 2008 9:31AM]

As my science teacher said "Physicians are masters at pool"

[tom @ Dec 11th 2008 10:22AM]

Did you mean "physicist"?

Physician is a doctor. You really should correct your teacher about that

[NickNick @ Dec 11th 2008 10:24AM]

No, that would be my english teacher, he didnt teach me well in spelling.

[loosely_coupled @ Dec 11th 2008 11:23PM]

Forget spelling, you didn't even know the proper word... How could you possibly think "p-h-y-s-i-c-i-a-n" could spell "p-h-y-s-i-c-i-s-t"

[NickNick @ Dec 11th 2008 9:34AM]

...however that guy isn't a physician, he sucks!

[siman @ Dec 11th 2008 9:40AM]

I can see why they need this!! They did not pot one ball!!

[Patrick @ Dec 11th 2008 10:03AM]

In their defense, that pool table is seriously warped from being moved once without dismantling and being over 100 years old. Balls will literally make u-turns if they're hit at the right speed in some places.

[Andir3.0 @ Dec 11th 2008 10:51AM]

Not to mention the system seems to be very slow, dependent on a specific cue which the switch out to make the shot... Why not mount the camera directly above the table along with the projector and have it look for the cue that's in your hand (straight line, white tip) and draw real time representations? You simply trace down the cue, to the ball, determine it's trajectory, and recursively track the path up to X iterations.

[Imperor @ Dec 15th 2008 2:07AM]

So do it Andir! Of course you can do it better but buying and mounting the equipment along with coding a more advanced program might be a bit above hobby level for a lot of people. But sure, a high res camera and a widescreen projector mounted directly above a new table and powered by a real computer would be better and faster...
I'd say this is more of a "concept-build". Pretty cool though even if I wonder what kind of speed they've made the calculations from as this will affect trajectory and distance as well.

[Ethan @ Dec 11th 2008 9:49AM]

It's like the opposite of those pool tricks videos - you know what's coming, and they miss the shot. But so would I. And refined a bit they could sell that for quite a bundle.

[cwj @ Dec 11th 2008 9:54AM]

Finally, pool hall's can compete with computerized/Friday nite galactic bowling.

[Tony @ Dec 11th 2008 10:05AM]

Knowing the angles in pool is like knowing you need to run the fastest in a foot race. The point: knowing is not nearly enough. I can figure out most angles without all the digital help, I just need a stick that will actually shoot the ball for me! Now I would pay for that.

[LondonConsultant @ Dec 11th 2008 10:13AM]

Pool sharks? More like pool minnows...

[tom @ Dec 11th 2008 10:24AM]

Few years from now, there won't be Yahoo pool. It will be called Googhoo pool or Microhoo Pool

[dvsbstrd @ Dec 11th 2008 11:27AM]

wat?

[Grant H @ Dec 11th 2008 10:24AM]

You'd give them a degree for this horse shit?!

Where do I sign up? I've got a project of digitally assisted Ass-to-Mouth I wanna simulate with computers and glowy projector lines.

[StalematE @ Dec 11th 2008 12:16PM]

You never go ass to mouth!!!!

[Imperor @ Dec 15th 2008 2:11AM]

All depends on whether she (/he?) washed properly before...

[Frankenstein Black @ Dec 11th 2008 10:25AM]

Now if they could only make it a hidden infrared system with matching contact lenses for the “geek hustler” in all of us ;^)...

[SlappyWag @ Dec 11th 2008 11:53AM]

FTW: "Ah, shit, it's getting close to out of batteries."

[douglas @ Dec 11th 2008 12:31PM]

Is it me or do they still stink, even after their project?

[bangladeshiluv @ Dec 11th 2008 12:56PM]

i always though a pool shark was some dude who stayed underwater to check out girl's asses

[tbone @ Dec 11th 2008 1:08PM]

yeah, but they still have to hit the ball correctly, woops scratch!

[wickedpheonix @ Dec 11th 2008 1:10PM]

I wouldn't think that this would seriously allow an amateur to complete with a pro though since pros have topspin and all that - curved projectories, just another tool in their arsenal that would allow them to overcome a system like this one (even if this system was perfect).

[TIMMAH! @ Dec 11th 2008 2:04PM]

Yes, it'd be a good basic learning tool to learn the angles of a cue ball struck dead center. The trajectories of both the cue and target balls will change depending on the english applied to the cue ball. Maybe they could project these as alternative tracks in a different color.

[kev @ Dec 11th 2008 1:22PM]

There's already a laser assisted tool for this sort of thing, and it doesn't help you anyway, since this is only one thing you've got to know to be good at billiards... There's putting english and throw on the ball, having muscle memory for how hard you need to hit on which speed cloth, if you can't master those you're dead against even an amateur.

[Ghen @ Dec 11th 2008 2:08PM]

With all this BS I was expecting virtual balls too.

[StalematE @ Dec 11th 2008 2:41PM]

that's what sh...

[pagan696 @ Dec 11th 2008 2:10PM]

reminds me of Donald (Duck) in MathMagicland.

[MikeWard1701 @ Dec 11th 2008 2:17PM]

Interesting proof-of-concept. Would be interesting to see how it developes.

Given time and some extra tech, the accuracy could be improved drastically along with realtime trajectories, you could even factor in spins and stalls if you could measure the cues verticle angle and and the height of the tip from the table.

[Tom @ Dec 11th 2008 2:21PM]

This is a lot of nerdery just to play billiards like a 10-year-old.

Seriously, if you were going to make a demo video, don't think you'd maybe:
1. get someone shooting who is skilled
2. use the clips where the balls actually go in the direction predicted by the projection
3. invest in a webcam with higher than 640x480 resolution. HD, even... (and why not just mount it in the same location as the projector, facing the same mirror?)

The problem here is that you have a virtual assistant made by non-experts... it's a ton of technology that still can't elevate your play above the most basic level, because the linear path the ball follows is only the most basic component of outcome of the shot. Speed alone (and the rotation it creates or doesn't create) is going to mess this up on any hard shot.

(And, on a side note, that guy they borrowed the mirror from needs it put back on his bedroom ceiling.)

[Deep @ Dec 11th 2008 2:23PM]

but this system still doesnt incorporate spin placed on the cue ball, which can drasically change the outcome of direction of the ball that the cue ball hits...

[dudi @ Dec 11th 2008 2:28PM]

i think it can compute any spin put left/right wise, since it can jot a line with respect to the circle it sees as the cue and the position of the stick wrt to it..
But the top and bottom spins will be a problem i guess.
But these guys have mounted the webcam itself in a near isometric position and neither horizontal or vertical.. so i think it was quite an awesome coding which went into the program.

[dudi @ Dec 11th 2008 2:25PM]

Nice!

[Magallanes @ Dec 11th 2008 3:27PM]

my strategy is :"hit harder".

[sctrojanman @ Dec 11th 2008 4:53PM]

You know what would have made this really impressive? If they actually pocketed a ball.

[alexander_h @ Dec 11th 2008 6:36PM]

The program fails to take into account the speed of the balls, which plays a part in multiple-bounce collisions. Just looking at the OP picture tells me it's not going to work.