Do you see this behaviour on board editor? One physical piece lights two squares!

Hi All!

New user here, having backed the the ChessUp 2 Kickstarter. I’ve asked this question on the Kickstarter Comment section but it is not very active!

Summary: On the board editor, by careful physical piece placement (with the piece not well centred in a square) it may be recognised as being present on two squares. This appears to impact some pieces (rooks and knights) more than others. It can be reproduced by all pieces of that type (and is irrespective of piece colour). It is easier to reproduce with some squares more than others (b7/c7 in particular for me).

Full Post: Having read all of the reports of faulty pieces my first (self-imposed!) activity was a rather tedious task of checking every piece was recognised in every square (32 pieces + 2 spare queens + 8 piece promotion pack = 42 pieces, x64 squares = 2688 combinations!). I did this by using the Board Editor and positioning one piece on one square at time, lifting the piece away each time (confirming recognition and removal successfully registered). Thankfully I did not identify any faulty pieces. What I did recognise, as I became sloppier at piece position while trying to speed up this activity, was that some positions appear to confuse the board, showing a single piece recognised as being active on two squares, when overlapped over two squares. In most cases a piece in this position simply did not recognise at all, a behaviour I would expect. But in some cases a piece straddling two squares was recognised as being present on both. I spent some time exploring this behaviour. It happens irrespective of piece colour and happens more regularly with some pieces (rooks and knights in particular) than others. The fact all instances of a given piece (irrespective of colour) behave the same suggests the behaviour is caused by the board more so than the piece (a replacement piece would not stop this behaviour). This appears to be much easier to replicate on some squares than others (b7/c7 in particular). Has anybody else noticed this issue? Is this problem unique to my board? On the one hand, I’d rather avoid the overhead of a board replacement but on the other I do not want a faulty board if this is an issue specific to my board.

Noting the accuracy of positioning that is required to reproduce this, it does not happen all the time but it is reproducible with effort (slightly less frequently by chance). Quite naturally I’m keen to ensure that there is definitely not a fault with my board. It’s a large investment and if there is something wrong then now is definitely the right time to ask the question.

Note: The absolutely on-form @Jeff did respond on Kickstarter to say this was ‘normal behaviour’. Curiosity in me wanted to know more and asked if this was understood (why only some pieces? why some squares more than others?). Jeff did not reply; completely understandably a) being so busy and b) Kickstarter’s comments to not manage notification of replies to replies very well. I also did not receive feedback from any other users. The question has now fallen far down the list and I suspect will not get a response. Rather than asking the question on Kickstarter again I thought I’d join the somewhat more activity community here!

Thank you for your help and support!

HI,

I did the exact same thing as you did on my board, testing each pieces on each square and obviously I did notice the same. However I can say that it is not something that break something in real game as the board will recognize a misplaced pieces as an “illegal move”, of course you will have to re enter your piece but honestly l, it never happened to me.

The only thing more strange that happened to me while testing the pieces and may have happened to you was that at some point, a misplaced (not center) piece has appeared on the editor as a King. I can’t remember which piece and I was not able to reproduce this even after trying hard. Also, it is probably not something that would break the game as it would be recognized as an illegal move and re centering the piece would fix this.

The reason there is not perfect square discretion is several things:

  1. It is a wireless connection from board to piece - there is not a strict boundary “contact” so to speak.

It is, however, quite accurate with square discretion.

Perfect discretion is actually (a) not desired and (b) not required.

(a) if we rejected a piece slightly out of the square, it would interrupt most games

Watch this sequence:

Observe how many pieces land outside the strict definition of a square.

In order to accept those moves, there is an overlap region when a piece can exist in both squares.

(b) It is not at all needed - the board recognizes moves and setup perfectly with natural play.

While you can force it to show two pieces, etc - that does not matter in normal use. As long as you get the piece mostly in one square, all works perfect.

Also - there is not correct behavior in this situation. Is half of a knight in a square 1 knight or 0 knights? It is undefined. We decide for the user when it matters. But the user has to intentionally place a piece that poorly, so again - this is not a concern for regular use cases.

As to why each piece and square combo behaves a bit different - these pieces all respond to a different radio signature (a different frequency). Each frequency has different beam patterns and different attenuation. Radio energy is also influenced by everything… literally everything… in the board and the environment. So a square in the middle is going to have a different radio pattern and response than a square at the edge. And not all edges are the same either. The product has displays, batteries, etc. Everything, including the shape of a particular piece, will influence the signal.

Your board is perfectly fine. If you want one that does not show a piece in two squares, we would need to ship you a broken one. However, caveat being that the broken one would not be able to play chess :sweat_smile:

It is an analog to digital conversion (radio energy to a chess piece). You just get to peek into that analog world playing around on the board editor. The only fix we can give you is to remove that feature - so you are not exposed to the analog world. But the board editor is very useful in upcoming updates, so we will leave it.

I admire the effort to test all scenarios. Be confident you have the most thoroughly tested unit and should hold onto it.

3 Likes

Thank you @TheGrinChess and @Jeff for your replies. It is reassuring to know that the behaviour observed is not only normal but also why. The explanation is very clear and when written in those terms it is actually impressive how well square discrimination is implemented to balance flexibility and accuracy, requiring a purposeful action to create a scenario where the piece could credibly be considered to be in either square. Of course the same explanation has also highlighted how even in these limited cases, the ‘correct’ answer is almost always determined by move legality. Many thanks for the reply – and I’m glad the Board Editor is staying as it has proved a very valuable diagnostic tool.

1 Like

And with the boardeditor flow that’s comming it will be really cool :wink: