There's an app for that!

[9BallPaul]

I'm no techie, but I'm guessing that there's a cell phone out there that -- with the right software -- could take a picture of a nine-ball layout, apply some software, and devise the smartest way to run the rack.

Seems all it would require is the right programmer and good consultant. Of course, the player would still be required to execute the shots, which means there's no guarantee here, just an advantage.

Question is this: Would it be legal? Should it be?

 

[pooladdiction]

To answer your question: I don't think that would be legal nor do I think it should be, during match play. Having said that, I would definitely purchase this app if it was available. I think it would be a great training aide in recognizing patterns.

 

[LAlouie]

I would think the algorithm for that would be almost impossible because there's no linear solution to a runout unless every ball was a stop shot. ie,,,it's one thing to program a sequence of "shoot-stop-shoot-stop", but a totally different program to say "shoot with 1o'clock english(where is that on the ball)/ medium speed(what is that)-cueball banks 10" off the rail- etc etc"...and that's on a baby layout with the simplest solution.

All in all, it would be too much of a hassle, and I think no one would use it anyway.

 

[CarlB]

I would think the algorithm for that would be almost impossible because there's no linear solution to a runout unless every ball was a stop shot. ie,,,it's one thing to program a sequence of "shoot-stop-shoot-stop", but a totally different program to say "shoot with 1o'clock english(where is that on the ball)/ medium speed(what is that)-cueball banks 10" off the rail- etc etc"...and that's on a baby layout with the simplest solution.

All in all, it would be too much of a hassle, and I think no one would use it anyway.

I agree...Just the algorithm for english/cue ball reaction, collision induced throw, minute angle difference would be more than the phone processor could handle.

Now an app that could show you optimal breaking spot for the different games, or maybe an app that you take a picture of the rack from overhead and it analyzes gaps therefore break patterns and where the balls SHOULD go with a soft/medium/hard break shot from a certain cue-ball location should be possible.

Carl.

 

[9BallPaul]

I would think the algorithm for that would be almost impossible because there's no linear solution to a runout unless every ball was a stop shot. ie,,,it's one thing to program a sequence of "shoot-stop-shoot-stop", but a totally different program to say "shoot with 1o'clock english(where is that on the ball)/ medium speed(what is that)-cueball banks 10" off the rail- etc etc"...and that's on a baby layout with the simplest solution.

All in all, it would be too much of a hassle, and I think no one would use it anyway.

Yeah, it would require a fresh photograph before each shot to be effective. Position would have to be re-analyzed shot-by-shot.

Still, seems to be within the realm of what's happening.

 

[dgem]

This is what you can do..take the layout of the table then plot it on a virtual pool 3 video game and let George breedlove or Jeannete Lee play it out for you.

:thumbup:

 

[ctyhntr]

This would probably be the most practical approach. Virtual Pool 3 already has a computer player mode. All they need to do is provide a mode where you can set up shots, and show the computer's shot selection.

This is what you can do..take the layout of the table then plot it on a virtual pool 3 video game and let George breedlove or Jeannete Lee play it out for you.

:thumbup:

 

[metallicane]

I wouldn't be surprised if this came out in a few years. The technological advances these days are truly amazing.

 

[9BallPaul]

I wouldn't be surprised if this came out in a few years. The technological advances these days are truly amazing.

I'm thinking the same. But would it be legal? Should it be legal?

 

[ctyhntr]

If there are no rules against it, then its legal. The question is using it during a match, is it ethical? I see it as being no different as coaching during a match, and should adhere to current rules and limits.

I'm thinking the same. But would it be legal? Should it be legal?

 

[donuteric]

Actually I think with the advancement in artificial intelligence over the years, I wouldn't be surprised if somebody came out with the algorithm to do it. Very much like chess, decision theory is implied to make the most educated move at each point of the game. Given that we can calculate the CB action (with or without English) off the rail, and know the area for the table to leave the CB after the shot, I really think it's possible. Although it can probably only predict the game based on the most ideal table and situation.

 

[youngstownkid]

The Environment:
First, you are talking about duplicating the physics that was done in a virtual pool type application. or hypothetically maybe it would be possible to get an API from the makers of the mobile version of Virtual Pool.

The Algorithm:
Coming up with an algorithm would be similarly difficult to dijkstra's shortest path algorithm (which i believe is still what is used on most mapquest type direction websites), which is no easy task (to sarcastically understate it). In some ways it will be more difficult since we don't actually have roads to choose from, rather the developer would be working with choosing paths within several "windows".

Real world variables that wouldn't be taken into account:
the cue ball - ease to spin (this would affect post collision movement)
the felt - how fast (this would affect speed to use)
the felt on the rails - how fast, how clean (would affect how the cue ball reacts to the rail when spinning)
size of pockets - affect ability to cheat the pocket on close shots

Add this all up and this makes for a very difficult task, I really don't see this ever being even considered let alone accomplished. This may not be feasible for the pro bono type of developers that love pool so much that they'd do it for free. the ones that would be qualified to take on the task development-wise would want PAID for it, and would also have to work with someone very knowledgeable about pattern play to help teach them so they could develop the algorithm, and this would amount to a very large bill.

 

[abs]

I think a fair set of algorithms exist in the current and last few generations of pool games; they would just need to be tuned-up a bit, and also modified to work in this specific proposed application. But, as good as they are, they are still crude and unrefined compared to what an “Open” level or better, or $ player, would be looking for.

As was pointed out, the environmental variables are not accurately accounted for, the users skill is not well accounted for, the dynamic changes with each shot needs to be assessed with each shot (assuming the leave isn’t precisely what was plotted).

Also, as was pointed out, the computing power of a phone isn’t nearly close enough to handle this. Static images, or even video streams, could be sent to a server to have the analysis done, and then have the results sent back to the handset, but this would be slow because of the mobile network overhead… Maybe too slow for a good/useful user experience.

In the end, when all is said and done, most of us have the computing power and algorithms in our heads anyway; I submit that brains will out-do any application for pool. The machines finally took us down at chess, but in the skill game of pool the combination of environmental analysis, tactics, strategy, skill, and technique is something the human brain can cope with fairly easily and amazingly precisely.

Maybe in 500 years it will be different, but for now, we “meat-sacks” rule.

 

[hang-the-9]

Chess is not much different from Pool in the amount of moves you pick from per shot (move). Even if there are less variables in a single move in chess than in options you can do to a hit, in chess you also need to think of the 30 things the oponent can do in reaction to your move and plan each move based on that, which add a whole other level of thinking that needs to go to each move.

There are chess playing computers that are damn good. Get this over to IBM to work on, if they can make a computer to beat a top chess pro, and a computer to beat humans at trivia, they can make one to calculate the easiest way through a rack. In fact if we can get this thought into the right head in an AI field, I'm betting this will get programmed.

 

[youngstownkid]

couldn't disagree more with the chess comparison. chess is a pure science. there are only 64 positions and 6 different pieces. pieces can't be bumped. chess is 2 dimensional. pool is 4 dimensional (3 dimensions of space, 4th being that of time) chess doesn't involve physics.

ok, yes, in theory it ca-ca-could be done. but it won't. it would cost waaay too much money, and wouldn't be something that is marketable. i mean, we're talking about an extremely small target audience:

ppl that play good enough to run out a rack
+ techie enough to use an iphone
+ nerdy enough to put the two together
+ willing to pay a LOT for it
+ willing to get the $hit kicked out of them for taking so long at the table, fouls called on them for coaching, rules made against them.

 

[LAlouie]

Yeah, it would require a fresh photograph before each shot to be effective. Position would have to be re-analyzed shot-by-shot.

Still, seems to be within the realm of what's happening.

It doesn't take into account YOUR stroke. A shot with 1o'clock spin at the cueball's edge is different from one player to another. The most common sense solution would be for an app to map out the run, showing you the required position play, and then leaving it up to you to actually do it.

It would be impossible still, because the app cannot make judgments on whether to play a shot or safe. My guess is it would attempt to pocket all shots since the essential input of it's creator would be that the goal of pool is to pocket balls. No app can make a choice of "safety or not safety". That is a human trait.

They interviewed the guy behind the IBM computer "Watson", and the interviewer asked him all kinds of questions about AI. The guy basically said what we've heard about computers for eons. Watson was built to do singular thing, play jeopardy, and it did not have the wherewithall to think for itself, so I highly doubt that some shmoe building an android/iphone app has figured out what everyone else hasn't.

 

[supergreenman]

a buddy of mine and I talked about developing this very thing. He figured it was doable.

we dicussed using some sort of bluetooth glasses that had a camera mounted on them and delivered a Heads Up Display HUD on the lenses.

As far as the environment goes, that's not such a big deal, all pool tables have 1 common thing. They're all exactly half as wide as they are long, this would enable the app to identify ball locations, once they have that there there are multiple gaming engines already in existance that would allow the phone to plan a layout.

Would this be an advantage to an experienced player? Hell no, I doubt a computer can match human imagination when it comes to layout planning.

Would it help a banger. Absolutely.

 

[Andrew Manning]

Hell no, I doubt a computer can match human imagination when it comes to layout planning.

It doesn't even take much human imagination to thoroughly outdo a computer in this regard.

It would be interesting to find out what might come out of an AI-based system, fed with millions of racks of actual humans playing pool, to see if it could start to come up with rulesets for what shot to choose. But as someone who programs for a living, I can tell you that kind of AI is VERY hard to accomplish effectively, and that without AI (relying only on hard-coded logic defined by the programmer), it wouldn't be worth sh*t, even with great programmers and extremely knowledgeable pool-player consultants.

-Andrew

 

[LAlouie]

a buddy of mine and I talked about developing this very thing. He figured it was doable.

we dicussed using some sort of bluetooth glasses that had a camera mounted on them and delivered a Heads Up Display HUD on the lenses.

As far as the environment goes, that's not such a big deal, all pool tables have 1 common thing. They're all exactly half as wide as they are long, this would enable the app to identify ball locations, once they have that there there are multiple gaming engines already in existance that would allow the phone to plan a layout.

Would this be an advantage to an experienced player? Hell no, I doubt a computer can match human imagination when it comes to layout planning.

Would it help a banger. Absolutely.

That doesn't help at all. All you're saying is you figured out a way for the camera to identify the balls and the layout, right? The player doesn't need a camera for that...all he needs are his peepers. :wink::wink::wink: To plan a layout, all you need to know is 1-2-3-4-5-6-7-8-9. It's HOW you get there that's the problem.