Possible Checker Placements

On a 40 by 40 square grid, Victor covers a 1 by 1 square with a checker piece. He is then able to cover the rest of the squares with 533 1 × 3 1 \times 3 triminos.

What is the total number of squares on which the checker piece could be placed?

Details and assumptions

The 1 × 3 1 \times 3 trimino consists of 3 squares arranged in a straight row (or column).


The answer is 196.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

8 solutions

Mark Hennings
Nov 4, 2013

Number the squares of the grid with the numbers 1 , 2 , 3 1,2,3 in the following pattern: 1 2 3 1 2 3 1 2 3 2 3 1 2 3 1 2 3 1 3 1 2 3 1 2 3 1 2 1 2 3 1 2 3 1 2 3 2 3 1 2 3 1 2 3 1 3 1 2 3 1 2 3 1 2 1 2 3 1 2 3 1 2 3 2 3 1 2 3 1 2 3 1 3 1 2 3 1 2 3 1 2 \begin{array}{|c|c|c|c|c|c|c|c|c|c} \hline 1 & 2 & 3 & 1 & 2 & 3 & 1 & 2 & 3 & \cdots \\ \hline 2 & 3 & 1 & 2 & 3 & 1 & 2 & 3 & 1 & \cdots \\ \hline 3 & 1 & 2 & 3 & 1 & 2 & 3 & 1 & 2 & \cdots \\ \hline 1 & 2 & 3 & 1 & 2 & 3 & 1 & 2 & 3 & \cdots \\ \hline 2 & 3 & 1 & 2 & 3 & 1 & 2 & 3 & 1 & \cdots \\ \hline 3 & 1 & 2 & 3 & 1 & 2 & 3 & 1 & 2 & \cdots \\ \hline 1 & 2 & 3 & 1 & 2 & 3 & 1 & 2 & 3 & \cdots \\ \hline 2 & 3 & 1 & 2 & 3 & 1 & 2 & 3 & 1 & \cdots \\ \hline 3 & 1 & 2 & 3 & 1 & 2 & 3 & 1 & 2 & \cdots \\ \hline \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \ddots \\ \end{array} There are 534 534 squares labelled 1 1 and 533 533 labelled 2 2 and 3 3 each. Thus the checker piece must be placed on a square labelled 1 1 , since every trimonoe covers one square labelled 1 1 , one square labelled 2 2 and one square labelled 3 3 .

However, if the board is rotated through 9 0 90^\circ , then the squares originally labelled 1 1 in rows 2 , 3 , 5 , 6 , 8 , 9 , 11 , 12 , 2,3,5,6,8,9,11,12,\ldots are now labelled 2 2 or 3 3 - the only squares that are labelled with 1 1 in both the original and the rotated position are the squares labelled 1 1 in rows 1 , 4 , 7 , 10 , , 40 1,4,7,10,\ldots,40 . There are 14 14 such squares in each of these 14 14 rows, and hence there are 196 196 squares that interest us.

It is each to show that a tiling with trimonoes is possible if the checker piece is placed on any of these 196 196 squares: the remaining squares in the same row as the checker piece can be tiled with trimonoes, and the remaining 39 39 rows come neatly in groups of three, and each group of three can readily be tiled by 40 40 vertically oriented trimonoes. Thus the answer is indeed 196 196 .

Well, I began it the same way. Can you please clarify, why only those 1 1 -squares are favorable which don't lose 1 1 after rotation. Why don't the entire set of 534 534 numbers satisfy?

A Brilliant Member - 7 years, 7 months ago

Log in to reply

The other way of thinking about this is to consider the square numbered with the 1 1 s, 2 2 s and 3 3 s forming diagonal lines in a NW/SE direction, as opposed to the SW/NE direction that they currently lie. A "good" square has to be numbered 1 1 for both of these numberings, and so the "good" squares are only those on rows 1,4,7,... This is Erick's argument.

Rotating the square just turns the first numbering system into the second.

Mark Hennings - 7 years, 7 months ago
Erick Wong
Nov 5, 2013

Label the x and y coordinates from 0 to 39. We claim that the valid checker positions are precisely those with x y 0 ( m o d 3 ) x \equiv y \equiv 0 \pmod 3 . First we show that the condition is necessary: 3-colour the squares of the grid in a checkerboard pattern according to the value of x + y ( m o d 3 ) x + y \pmod 3 . Then of the 1600 squares, exactly 534 will be coloured 0, while colours 1 and 2 occur 533 times each. Since any triomino covers exactly one square of each colour, the checker must lie on a square coloured 0: if the coordinates are ( x 0 , y 0 ) (x_0,y_0) then x 0 + y 0 0 ( m o d 3 ) x_0+y_0 \equiv 0 \pmod 3 . An identical argument with the colouring x y ( m o d 3 ) x - y \pmod 3 shows that also x 0 y 0 0 ( m o d 3 ) x_0-y_0 \equiv 0 \pmod 3 , which gives x 0 y 0 0 ( m o d 3 ) x_0 \equiv y_0 \equiv 0 \pmod 3 as claimed.

Conversely, if we place the checker at a position with both x and y divisible by 3, it's easy to see that the row it is in can be tiled with triominos, and that the subrectangles above and below that row each have height divisible by 3, and thus can be likewise tiled. Thus the condition is also sufficient, and the number of squares satisfying it is 1 4 2 = 196 14^2 = 196 ( x / 3 x/3 lies between 0 and 13 so there are 14 choices for x and similarly for y).

Very nice explanation. Good Job! :)

A Brilliant Member - 7 years, 7 months ago

Really an amazing way to solve. :)

sharvik mital - 7 years, 7 months ago
Peter Byers
Nov 8, 2013

Assign point values of 4 , 2 , 4, -2, and 1 1 to the squares according to this pattern:

4 , 2 , 2 , . . . 4, -2, -2, ...

2 , 1 , 1 , . . . -2, 1, 1,...

2 , 1 , 1 , . . . -2, 1, 1,...

where each row or column keeps repeating the same three numbers (either 4 , 2 , 2 4, -2, -2 or 2 , 1 , 1 -2, 1, 1 ).

If the checker piece is on a square with value 4 4 then the rest can be covered with triminos. (For example, first cover the rest of the checker's row and column. This leaves four (or fewer) rectangles whose dimensions are all multiples of three, which can obviously be covered with triminos.)

But on the other hand, the triminos must cover squares with point-value total of exactly zero (each trimino covers either 4 , 2 , 2 4, -2, -2 or 2 , 1 , 1 -2, 1, 1 ), so all possible locations of the checker must have the same point value.

And since we already showed that every square with point value 4 4 is a possible location for the checker piece, the answer will be the number of such squares or 1 4 2 = 196 14^2=196 .

This is quite cool, I like how you get the necessary condition in one shot by tensoring together the two orientations!

Erick Wong - 7 years, 7 months ago
Hadia Qadir
Sep 7, 2015

As the values ofK's are increasing, we take the grids by trimming the previous grid along all the 4 boundaries. So the required answer is .

We can cover any 3 × 40 3 \times 40 rectangle by filling first the 3 × 39 3 \times 39 part, and then filling the last 1 × 3 1 \times 3 part with a vertical tetromino (or horizontal, if you work from side to side). So we can always fill the 39 × 40 39 \times 40 part of the board. For the row or column we have at this moment, we put 13 13 tetrominos, and we can move the tetrominos from 0 0 to 13 13 times. This means we have in that row 13 0 + 1 = 14 13 - 0 + 1 = 14 spaces for the checkers piece. But this can't be repeated in all the rows. We can repeat it whenever we can move one or more 3 × 40 3 \times 40 rectangle (because we cant break the tetrominos). So again we have 13 13 of those rectangles, which gives 14 14 rows in which we can make the 14 14 other operations. Important: Don't repeat this from side to side if you did it vertically, because you'll repeat the same spaces. So the answer is 14 14 = 196 14 \cdot 14 = \boxed {196} .

Bitan Basu
Nov 6, 2013

You can think the square grid as a collection of cells with 0 1 or 2 written on it. 0 1 2 0 1 2 0 ...... 1 2 0 1 2 0 1 ...... 2 0 1 2 0 1 2 ....... ........................

Now each trimino can be placed on 012 thus the total 3

As the sum of all elements is divisible by 3, so the covered cell must be a 0 cell. There are 14*14=196 0 cells. so ans is 196

There are 534 534 squares labelled 0 0 ...

Mark Hennings - 7 years, 7 months ago

Log in to reply

This is a good question.
Indeed there are 534 zeroes. But the zeroes will not coincide if you rotate it thus eliminating such chances.

Bitan Basu - 7 years, 7 months ago

Log in to reply

Indeed, but your original post mentioned nothing about rotation.

Mark Hennings - 7 years, 7 months ago

You didn't show that these could indeed be covered. All you have stated is that number of remaining elements is divisible by 3.

Mirza Baig - 7 years, 7 months ago

Log in to reply

I appreciate your question.
The basic idea to solve the problem is to visualize the structure.There are 40=1(mod3) rows here.So, if you keep it in a row which is 1(mod3) then it will divide the whole board into a rows upside and b rows downside.As a+b=40-1=39 we can divide 39 into a and b such that both are multiple of 3.Now you will place horizontal triminoes till 39th column. And for the 40th column we will use 3 rows at once.Now for the row where I have placed the checker,it should be placed at 0(mod 3)th column as it will partition it into two blocks which are divisible by 3.Thus it is possible to tile them.

Bitan Basu - 7 years, 7 months ago

Log in to reply

Thanks! Now it makes sense.

Mirza Baig - 7 years, 7 months ago
Lucy Gow
Nov 5, 2013

Colour the diagonals that go in the bottom-left to top-right direction in three alternating colours: red, white, and blue. Then the diagonals that are coloured red will be of lengths 1 , 4 , 7 , 10 , . . . , 37 , 40 , 37 , . . . , 4 , 1. 1, 4, 7, 10, ..., 37, 40, 37, ..., 4, 1. So the number of red squares is 40 + 2 × n = 1 12 ( 1 + 3 n ) = 40 + 2 ( 12 + 3 × 13 × 12 2 ) = 532. 40 + 2\times \sum_{n=1}^{12} (1+3n) = 40 + 2(12 + 3\times \frac{13 \times 12}{2}) = 532. The diagonals that are coloured white will be of lengths 2 , 5 , 8 , . . . , 38 , 39 , 36 , . . . , 6 , 3. 2, 5, 8, ..., 38, 39, 36, ..., 6, 3. So the number of white squares is n = 1 12 ( 2 + 3 n ) + n = 1 12 ( 3 + 3 n ) = 351. \sum_{n=1}^{12} (2+3n) + \sum_{n=1}^{12} (3+3n) = 351. The diagonals that are coloured blue will be of lengths 3 , 6 , 9 , . . . , 39 , 38 , 35 , . . . , 5 , 2 , 3, 6, 9, ..., 39, 38, 35, ..., 5, 2, so that the number of blue square is also 351.

Since with this colouring of the squares, any trimino covers a red, a white, and a blue square, a tiling by triminos will colour the same number of red, white, and blue squares. Therefore, Victor must place the checker piece on a red square.

Furthermore, if we colour the squares down the diagonals in the other direction so that the diagonals go in the top-left to bottom-right direction, we can apply the same argument. Then Victor can only place the checker piece on squares that will be coloured red in both colouring schemes. These squares form a 14 × 14 14 \times 14 grid, with a total of 196 squares.

Conversely, we can always tile the grid after Victor covers one of these 196 tiles: it is easy to see that one can tile a 4 × 4 4 \times 4 square grid with triminos if one of the corners is covered. When Victor places a checker piece over a square, draw a 4 × 4 4 \times 4 grid around it so that the covered square is in the corner. Then we can tile that 4 × 4 4 \times 4 grid, and we can also tile the remaining board because it can be broken up into rectangles where each side is a multiple of three squares wide.

Your counting has gone wrong. The count of red squares should be between n = 0 n=0 and n = 12 n=12 , giving 534 534 squares, and your count of white and blue squares should also be between n = 0 n=0 and n = 12 n=12 , giving 533 533 each time. As a check, 534 + 533 + 533 = 1600 534 + 533 + 533 = 1600 .

Mark Hennings - 7 years, 7 months ago

Log in to reply

Thanks! I should have checked. Oh well.

Lucy Gow - 7 years, 7 months ago
Sujoy Roy
Nov 4, 2013

For ( 40 3 2 k ) × ( 40 3 2 k ) (40-3*2*k) \times (40-3*2*k) grid we find ( 40 1 3 2 k ) / 3 × ( 40 1 3 2 k ) / 3 = 4 ( 40 1 3 2 k ) / 3 (40-1-3*2*k)/3 \times (40-1-3*2*k)/3 = 4*(40-1-3*2*k)/3 different places to put the checker piece and the positions are 3 i + 1 3*i+1 for first and last row where i = 0 , 1 , . . . , ( 40 1 3 2 k ) / 3 i = 0, 1, ..., (40-1-3*2*k)/3 and 3 j + 1 3*j+1 for first and last column where j = 1 , 2 , . . . , ( 40 1 3 2 k ) / 3 1 j = 1, 2, ..., (40-1-3*2*k)/3-1 and k = 0 , 1 , . . . , 6 k = 0, 1, ..., 6 .

As the values of k k 's are increasing, we take the grids by trimming the previous grid along all the 4 boundaries. So the required answer is 4 ( 13 + 11 + 9 + 7 + 5 + 3 + 1 ) = 196 4*(13+11+9+7+5+3+1)=196 .

sujoy roy - 7 years, 7 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...