Extraordinary chessboard or Kickstarter scam
February 24, 2020 11:24 AM Subscribe
"If you follow the Internet chess scene closely you will no doubt have noticed the new kid on the block: Regium. Purportedly an e-board that can connect to all the major chess websites and play seamlessly with a real physical board, over the Internet! [...] While we want to believe as much as any chess lover, we have a few doubts about these claims as well as the conduct of Regium. Here are some reasons for caution." Lichess investigates the Regium chessboard
See also Chess.com's post on their sponsorship relationship with Regium and their own findings, including: "The boards shown in promotional videos are manufactured by Spanish company Rechapados Ferrer. A representative of the company has indicated that the boards pictured in the videos have not been modified by Rechapados Ferrer with any components that would allow for the technology presented in those videos. They also confirm that they have not collaborated with REGIUM on the modification process of these boards."
See also Chess.com's post on their sponsorship relationship with Regium and their own findings, including: "The boards shown in promotional videos are manufactured by Spanish company Rechapados Ferrer. A representative of the company has indicated that the boards pictured in the videos have not been modified by Rechapados Ferrer with any components that would allow for the technology presented in those videos. They also confirm that they have not collaborated with REGIUM on the modification process of these boards."
I hope this inspires someone else to actually create the magic that appears here; that's some incredible (potentially fictional) technology.
I'm curious what, if anything, Kickstarter will do with this evidence.
posted by el io at 12:04 PM on February 24, 2020
I'm curious what, if anything, Kickstarter will do with this evidence.
posted by el io at 12:04 PM on February 24, 2020
This is a relatively niche market for chess nerds that talk A LOT about chess . . . it seems like such a low-reward scam to me with a high likelihood of getting called out immediately. Which is I guess what happened, on the central hub of chess nerd-dom.
The technology seems achievable at this point though, right?
posted by Think_Long at 12:21 PM on February 24, 2020 [3 favorites]
The technology seems achievable at this point though, right?
posted by Think_Long at 12:21 PM on February 24, 2020 [3 favorites]
The first article notes it's not only achievable, it already exists and is commercially available: Square Off. The Regium claim is that they've built something quieter and sleeker, but if you want a real physical chessboard that you can move your own pieces on and that will then move the other side's pieces either based on an AI engine or an actual human opponent's responses connected via chess.com, you can order that today (in the US, UK, Austria, Belgium, Canada, Croatia, and Cyprus at least), no need to kickstart.
posted by solotoro at 12:32 PM on February 24, 2020 [9 favorites]
posted by solotoro at 12:32 PM on February 24, 2020 [9 favorites]
Leaving aside the claims of fakery, I'm not sure how you could achieve this effect with fixed-point electromagnets on a single plane. I guess if they carefully control the gradient of magnetism between fixed electromagnets they could probably achieve it for a single piece on the board, but I don't see a way for this to work with a board full of pieces where it's able to move a single piece around without disturbing other pieces. Maybe with enough electromagnets you'd have enough "resolution" to do that, but would the magnetic field be powerful enough to drive motion at that point?
posted by Aleyn at 12:47 PM on February 24, 2020 [1 favorite]
posted by Aleyn at 12:47 PM on February 24, 2020 [1 favorite]
Techmoan did a sort of rambling review of the Square Off last fall. Read the video description for a brief overview of how it works, nothing too magical at the end of the day - just a cartesian robot and an electromagnet.
posted by Kyol at 12:51 PM on February 24, 2020 [3 favorites]
posted by Kyol at 12:51 PM on February 24, 2020 [3 favorites]
The video is worth it for the "Chess Player" actor at the end. Very believable.
posted by Kafkaesque at 1:04 PM on February 24, 2020 [1 favorite]
posted by Kafkaesque at 1:04 PM on February 24, 2020 [1 favorite]
That Square Off board — can you get two AIs (or two remotely-connected human players, or one of each) to play against each other? With no input from me, the board's owner and local proprietor? I feel like it might be worth the price just for the conversation piece. Just to have a game board, somewhere in my house, playing itself.
posted by penduluum at 1:21 PM on February 24, 2020 [14 favorites]
posted by penduluum at 1:21 PM on February 24, 2020 [14 favorites]
Watch the Techmoan video, there's a timelapse of the board replaying a game by itself.
I think that a board with fixed electromagnets is not impossible, as the Square Off board proves you can move pieces past each other running along the gaps in-between pieces, though the question is why you would want to fiddle with all those inductors. Instead, you could put a number of free-moving robots each with an electromagnet underneath a thin chess board, such that they can move enough pieces at once that it would be difficult to distinguish from a grid of electromagnets. Add some RFID tags in the pieces to identify them and you'll only need to put the pieces on the board for the robots to sort them for you.
posted by Eleven at 1:35 PM on February 24, 2020 [3 favorites]
I think that a board with fixed electromagnets is not impossible, as the Square Off board proves you can move pieces past each other running along the gaps in-between pieces, though the question is why you would want to fiddle with all those inductors. Instead, you could put a number of free-moving robots each with an electromagnet underneath a thin chess board, such that they can move enough pieces at once that it would be difficult to distinguish from a grid of electromagnets. Add some RFID tags in the pieces to identify them and you'll only need to put the pieces on the board for the robots to sort them for you.
posted by Eleven at 1:35 PM on February 24, 2020 [3 favorites]
I guess if they carefully control the gradient of magnetism between fixed electromagnets they could probably achieve it for a single piece on the board, but I don't see a way for this to work with a board full of pieces where it's able to move a single piece around without disturbing other pieces.
I would bet a decent amount of money that this is exactly what the designers did, but the results of a resolution of say 16x16 electromagnets (4 per square) was less impressive than they were hoping for. This would normally result in a prototype designer saying "Oh well better luck next time" and giving up, but this is a world where there's a service specifically geared towards providing crowdsourced money to inventors with no guarantees of success.
Instead, you could put a number of free-moving robots each with an electromagnet underneath a thin chess board, such that they can move enough pieces at once that it would be difficult to distinguish from a grid of electromagnets.
The problem with this and the Square Off board is both of them require moving parts and thus more expensive while introducing points of failure. A zero-moving-part grid of electromagnets that move the pieces like a linear motor would be extremely reliable and cheaper to produce, assuming the electromagnets can be made small and powerful enough.
posted by Mr.Encyclopedia at 1:40 PM on February 24, 2020 [1 favorite]
I would bet a decent amount of money that this is exactly what the designers did, but the results of a resolution of say 16x16 electromagnets (4 per square) was less impressive than they were hoping for. This would normally result in a prototype designer saying "Oh well better luck next time" and giving up, but this is a world where there's a service specifically geared towards providing crowdsourced money to inventors with no guarantees of success.
Instead, you could put a number of free-moving robots each with an electromagnet underneath a thin chess board, such that they can move enough pieces at once that it would be difficult to distinguish from a grid of electromagnets.
The problem with this and the Square Off board is both of them require moving parts and thus more expensive while introducing points of failure. A zero-moving-part grid of electromagnets that move the pieces like a linear motor would be extremely reliable and cheaper to produce, assuming the electromagnets can be made small and powerful enough.
posted by Mr.Encyclopedia at 1:40 PM on February 24, 2020 [1 favorite]
I don't know, an electromagnet grid with individually addressed inductors each with analogue control seems quite expensive too!
Making a cheap automatic chessboard sounds fun! It almost suggests itself as a neat Raspberry Pi project using a camera and a robot arm. Recognise the pieces with visual recognition AI, work out the next move, then guide the robot arm to pick up the piece. Someone else can waste their time on that though!
posted by Eleven at 1:55 PM on February 24, 2020
Making a cheap automatic chessboard sounds fun! It almost suggests itself as a neat Raspberry Pi project using a camera and a robot arm. Recognise the pieces with visual recognition AI, work out the next move, then guide the robot arm to pick up the piece. Someone else can waste their time on that though!
posted by Eleven at 1:55 PM on February 24, 2020
Apparently, 4 of the 6 team members of Regium do not exist. Their identities may have been forged from thin air, using common names that yield no relevant results on the Internet. The only pictures we have of them seem to have been generated by https://www.thispersondoesnotexist.com/. -the lichess article
*endless giggling at this dumb future*
posted by egypturnash at 2:10 PM on February 24, 2020 [13 favorites]
*endless giggling at this dumb future*
posted by egypturnash at 2:10 PM on February 24, 2020 [13 favorites]
Truly bizarre, if this is the scam that it appears to be. After evidence of video retouching, they used a sock puppet to post "a second video, where a trick, CGI or Stop Motion is impossible."
And the video does look entirely convincing, except for a bizarre bubble that pops up to cover half the screen for one second at 1:12, complete with sound effect.
If this is CGI, it's well done—he interacts with the pieces very convincingly. But why go through all that trouble and then throw in a giant red flag that something's being covered up? If it's not CGI, then why does everything else about the project make it seem like a scam?
…Ah ha, just realized that the camera makes a weird move that takes the board completely out of frame for a moment at 2:19. All the automatic piece movement happens before that transition, and most of the picking pieces up and moving them off the board happens after that transition.
posted by designbot at 3:11 PM on February 24, 2020 [2 favorites]
And the video does look entirely convincing, except for a bizarre bubble that pops up to cover half the screen for one second at 1:12, complete with sound effect.
If this is CGI, it's well done—he interacts with the pieces very convincingly. But why go through all that trouble and then throw in a giant red flag that something's being covered up? If it's not CGI, then why does everything else about the project make it seem like a scam?
…Ah ha, just realized that the camera makes a weird move that takes the board completely out of frame for a moment at 2:19. All the automatic piece movement happens before that transition, and most of the picking pieces up and moving them off the board happens after that transition.
posted by designbot at 3:11 PM on February 24, 2020 [2 favorites]
I am just waiting for it to come out that Angel Delgado - CTO, Engineer, Mensa Member - is not actually a member of Mensa. Looks like the only way to check their rolls is to be a member though.
posted by egypturnash at 3:42 PM on February 24, 2020 [1 favorite]
posted by egypturnash at 3:42 PM on February 24, 2020 [1 favorite]
designbot: "Truly bizarre, if this is the scam that it appears to be."
Yeah I mean come on, Angel Delgado is a MENSA member!
posted by chavenet at 3:42 PM on February 24, 2020
Yeah I mean come on, Angel Delgado is a MENSA member!
posted by chavenet at 3:42 PM on February 24, 2020
That second video seems like the sort of thing that Captain Disillusion would have a field day with.
posted by Kyol at 4:40 PM on February 24, 2020 [5 favorites]
posted by Kyol at 4:40 PM on February 24, 2020 [5 favorites]
Huh. I think the most suspicious thing about the second video is that it starts out on a tripod and switches to handheld. It's a whole lot easier to mask out part of a scene if the camera doesn't move, and the only time we see a piece move while we can see under the table is at 0:42, before it goes to handheld.
Actually, I might have found something. If you watch the lower right corner of the video at 1:17, it looks like something's moving under the glass. My guess is that there's someone across the board from the player moving the pieces with magnets, and they're just masked out for the first pawn move.
posted by ectabo at 4:53 PM on February 24, 2020
Actually, I might have found something. If you watch the lower right corner of the video at 1:17, it looks like something's moving under the glass. My guess is that there's someone across the board from the player moving the pieces with magnets, and they're just masked out for the first pawn move.
posted by ectabo at 4:53 PM on February 24, 2020
New Video where they Zapruder their own original video rather than filming anything new.
They also claim that it's only 65% working, which is the first thing I've seen that makes me think it could be real.
posted by mmoncur at 12:08 AM on February 25, 2020
They also claim that it's only 65% working, which is the first thing I've seen that makes me think it could be real.
posted by mmoncur at 12:08 AM on February 25, 2020
> I don't know, an electromagnet grid with individually addressed inductors each with analogue control seems quite expensive too!
A $7 Arduino i2c has 111 free addresses (out of 128; the remainder are reserved). A 16 x 16 grid mentioned above would use 32. Double the resolution (for even finer control) and you would still only need 64 addresses.
I'm not sure how you could make this work with any kind of smooth movement though. Wiring would be complex, a circuit board would be huge and expensive, and a grid of hundreds of inductors and electromagnets sounds complicated and fiddly and failure-prone.
To my mind, idealizing a machine that silently moves things around is the wrong way to do it. What they should do is have a bigass electromagnet under each square, with smaller elements in each corner. A move is made by suddenly charging a magnet under a piece, using the smaller elements to aim, and launching the piece into the air to be caught by the magnet at the destination square.
posted by ardgedee at 3:56 AM on February 25, 2020 [7 favorites]
A $7 Arduino i2c has 111 free addresses (out of 128; the remainder are reserved). A 16 x 16 grid mentioned above would use 32. Double the resolution (for even finer control) and you would still only need 64 addresses.
I'm not sure how you could make this work with any kind of smooth movement though. Wiring would be complex, a circuit board would be huge and expensive, and a grid of hundreds of inductors and electromagnets sounds complicated and fiddly and failure-prone.
To my mind, idealizing a machine that silently moves things around is the wrong way to do it. What they should do is have a bigass electromagnet under each square, with smaller elements in each corner. A move is made by suddenly charging a magnet under a piece, using the smaller elements to aim, and launching the piece into the air to be caught by the magnet at the destination square.
posted by ardgedee at 3:56 AM on February 25, 2020 [7 favorites]
launching the piece into the air to be caught by the magnet at the destination square.
You’ve just invented Regium Checkers!
posted by valkane at 5:13 AM on February 25, 2020 [2 favorites]
You’ve just invented Regium Checkers!
posted by valkane at 5:13 AM on February 25, 2020 [2 favorites]
I mean, here goes some stereotyping, and not a bit self directed at that, but messing with nerds with time on their hands is a 'do at your own risk' sort of venture. And, well, I'm going to go out on my previous-chess-playing experience here and say that a decent percentage of chess players of a certain type are nerds that have time on their hands.
I'm not surprised this is being blasted and that the fake team members are being pegged. Pardon my while I get some popcorn (and clean my hands before I move my piece...).
posted by RolandOfEld at 6:48 AM on February 25, 2020
I'm not surprised this is being blasted and that the fake team members are being pegged. Pardon my while I get some popcorn (and clean my hands before I move my piece...).
posted by RolandOfEld at 6:48 AM on February 25, 2020
The Kickstarter is live. Money wise its going well for them but the comments sections is currently a war between 2 (probably 1 with 2 accounts) people and the creator posting "its a scam" and "no it's not" copy and pastes to the tune of like 200 comments. Even so, so far they have raised 33,000 out of a requested 50,000 so I guess they have that going for them. Interesting to see if it survives the full 15 days.
posted by Captain_Science at 7:25 AM on February 25, 2020
posted by Captain_Science at 7:25 AM on February 25, 2020
I hadn't seen the Square Off before. Looks nice, but it seems weird that it took 30 years to connect the Milton Bradley Grandmaster to the internet. (And get a bit quieter as well).
The Regium would be much nicer (if it's real)- autosensing pieces, and the speed. Love the 'lock' feature.
Interesting that in their video, there's a clip of Carlsen-Morozevich at the World Blitz championship. They're playing on a DGT board, and as far as I can tell, DGT and Rechapados are not related.
posted by MtDewd at 8:00 AM on February 25, 2020
The Regium would be much nicer (if it's real)- autosensing pieces, and the speed. Love the 'lock' feature.
Interesting that in their video, there's a clip of Carlsen-Morozevich at the World Blitz championship. They're playing on a DGT board, and as far as I can tell, DGT and Rechapados are not related.
posted by MtDewd at 8:00 AM on February 25, 2020
There's a significant technical issue with the idea of a grid of electromagnets, in that simply due to the nature of magnetic fields there aren't stable points of magnetic attraction between multiple magnets (as far as I know) - a given point in space will always have a net pull towards one of the magnets in particular. Furthermore, the field strength of magnets falls off extremely quickly (inverse-cube, rather than the familiar inverse-square for things like gravity and electrical charge). The very smooth movement depicted in the video seems pretty implausible to me, given those factors.
Intuitively I would think that the most common "scam" failure mode for Kickstarter is "we learned after taking everyone's money that the problem is a lot harder in practice than we thought", and that definitely sounds like a thing that would happen with a naive approach to "let's make a chess board with a grid of electromagnets". (It turns out that "slick CGI mockup" + "funding" does not turn out to equal "working product that looks like the slick CGI mockup". Go figure.) A fundamental issue in crowdfunding is that a bit of dishonesty in presentation can be extremely tempting (e.g., nobody's gonna give their money to two guys in a garage, let's say we're a company) -- but any lack of transparency will tend to spiral out of control as you need to lie more to cover up your lies.
My personal heuristic that I've learned is "don't crowdfund anything with unsolved technical challenges". Crowdfund a board game that's already designed and playtested, and needs funding for a print and production run? Sure, knowing how to get a board game printed and how much that costs is a tractable problem. Crowdfund a really cool-sounding multiplayer online computer game that's doing a thing you've never seen in a game before? No matter how sincere and passionate the people behind that project are, there's a good chance the promised game never materializes.
posted by NMcCoy at 11:07 AM on February 25, 2020 [5 favorites]
Intuitively I would think that the most common "scam" failure mode for Kickstarter is "we learned after taking everyone's money that the problem is a lot harder in practice than we thought", and that definitely sounds like a thing that would happen with a naive approach to "let's make a chess board with a grid of electromagnets". (It turns out that "slick CGI mockup" + "funding" does not turn out to equal "working product that looks like the slick CGI mockup". Go figure.) A fundamental issue in crowdfunding is that a bit of dishonesty in presentation can be extremely tempting (e.g., nobody's gonna give their money to two guys in a garage, let's say we're a company) -- but any lack of transparency will tend to spiral out of control as you need to lie more to cover up your lies.
My personal heuristic that I've learned is "don't crowdfund anything with unsolved technical challenges". Crowdfund a board game that's already designed and playtested, and needs funding for a print and production run? Sure, knowing how to get a board game printed and how much that costs is a tractable problem. Crowdfund a really cool-sounding multiplayer online computer game that's doing a thing you've never seen in a game before? No matter how sincere and passionate the people behind that project are, there's a good chance the promised game never materializes.
posted by NMcCoy at 11:07 AM on February 25, 2020 [5 favorites]
They also claim that it's only 65% working, which is the first thing I've seen that makes me think it could be real.
Ah, I see. "Pawns, kings, and rooks work just fine, it's just the 35% of pieces that move more than one space diagonally that we haven't figured out." Or possibly "13 times out of 20, when we try to move a piece, it ends up where we tried to put it and not knocked over or flung off of the board."
65% working does not mean "with 35% more work than we've already done, we'll have a perfectly-functional product".
Per the heuristic in my pervious post, you shouldn't touch any "65% working" crowdfunding project with a 10-foot pole.
posted by NMcCoy at 11:28 AM on February 25, 2020 [4 favorites]
Ah, I see. "Pawns, kings, and rooks work just fine, it's just the 35% of pieces that move more than one space diagonally that we haven't figured out." Or possibly "13 times out of 20, when we try to move a piece, it ends up where we tried to put it and not knocked over or flung off of the board."
65% working does not mean "with 35% more work than we've already done, we'll have a perfectly-functional product".
Per the heuristic in my pervious post, you shouldn't touch any "65% working" crowdfunding project with a 10-foot pole.
posted by NMcCoy at 11:28 AM on February 25, 2020 [4 favorites]
Tom Standage's book about the Mechanical Turk is pretty good, by the by.
posted by Chrysostom at 2:08 PM on February 25, 2020
posted by Chrysostom at 2:08 PM on February 25, 2020
I liked that book as well.
Don't suppose there's a tiny man inside the Regium board....
posted by MtDewd at 6:58 AM on February 26, 2020
Don't suppose there's a tiny man inside the Regium board....
posted by MtDewd at 6:58 AM on February 26, 2020
And the kickstarter has been suspended.
posted by Captain_Science at 10:00 AM on March 2, 2020 [5 favorites]
posted by Captain_Science at 10:00 AM on March 2, 2020 [5 favorites]
> Don't suppose there's a tiny man inside the Regium board....
I'm sure they could secure one for the prototype, but that approach might not scale to production.
posted by ardgedee at 12:38 PM on March 2, 2020
I'm sure they could secure one for the prototype, but that approach might not scale to production.
posted by ardgedee at 12:38 PM on March 2, 2020
« Older Katherine Johnson has passed at the age of 101. | wrecks it after a riotous all-animal party Newer »
This thread has been archived and is closed to new comments
A legitimate company would, if there product were real and had all the features they advertised, simply respond to the accusations with proof that their stuff is real. Sounds like they went the other way...
posted by Captain_Science at 11:52 AM on February 24, 2020 [11 favorites]