Loyal Lines

Given an N N -gon, we say that a line is loyal if it intersects the perimeter of the N N -gon along the entire length of exactly one edge. Determine the largest integer N N such that every non-degenerate N N -gon has a loyal line.

Details and assumptions

  • An N N -gon is non-degenerate if no three consecutive vertices are collinear, or equivalently, that no two consecutive edges are on the same line.
  • The above image is an example of a 20-gon that doesn't satisfy the conditions of the problem. The 10 dotted lines are all the lines that contain at least one edge of the 20-gon, however, because all of these lines contain 2 edges, none of them are loyal lines. This shows that the answer is not 20.


The answer is 13.

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.

3 solutions

Matt McNabb
Aug 28, 2013

If N 10 N \ge 10 and N N is even then we have a simple counterexample: the outside of the N 2 \frac{N}{2} -pointed star. For example N=10 .

The cases of N N being odd are a bit more problematic. If every line coincides with an even number of edges, then the total number of edges must be even. Therefore in this case there must be a line coinciding with an odd number of edges.

The minimal necessary case would be a line coinciding with 3 3 edges. In this case there must be at least 6 6 other lines (one through each endpoint of the 3 edges). Since these lines must each contain at least 2 2 edges, there must be at least 15 15 edges in the n-gon.

I was able to find a 15-gon counterexample; a modification of the pentacle star .

The same modification of the pentacle star that produced this, can be done on any of the other higher-pointed stars (adding 5 edges in total), so we have counterexamples for all N 14 N \ge 14 now, and have proven that N = 15 N=15 is the smallest odd counterexample.

Therefore N = 13 N=13 is the largest N N with no counterexamples.

There are a number of different stars, depending on how you try to join the vertices. If we fix the 2 n 2n -gon (for n 5 n \ge 5 ) to be the non-convex one obtained by trying to join vertex 1 to vertex 3, vertex 3 to vertex 5, and so on (so every attempted line misses out one vertex), then this n n -pointed star is indeed an example of a polygon with 2 n 2n sides for which no line exists containing just one edge.

Here is a systematic and relatively simple way of obtaining suitable examples of m m -gons with the desired property for all odd m 15 m \ge 15 . Start with the 2 n 2n -gon star where n 6 n\ge6 is even, and take a new point equidistant between two outer points of the star. Join that point to the two vertices (almost) diametrically opposite that point, and you obtain a 2 n + 3 2n+3 -gon with the desired properties:

alt text alt text

Alternatively, join that point to one of the two vertices (almost) diametrically opposite that point and to a point on the line just "inside" the other (almost) diametrically opposite point, and you obtain a 2 n + 5 2n+5 -gon with the desired properties:

alt text alt text

Between these two techniques, we get all m m -gons with m 15 m\ge15 and odd.

Mark Hennings - 7 years, 9 months ago

Log in to reply

Very nice! I was constructing pointed stars and snipping off their tops instead.

Calvin Lin Staff - 7 years, 9 months ago

Another way of generating the odd cases N = 17 N=17 and upwards is to draw the ( N 3 ) / 2 (N-3)/2 -pointed star, and then draw a line cutting off three of the points. Example with N=17

Matt McNabb - 7 years, 9 months ago
黎 李
May 20, 2014

connecting ith and (i+2)th side of a convex n-gon to obtain an n-star shows N=2n is not right(n>=5) draw a butterfly (4 lines) on an equilateral triangle(3 lines) shows N=15 is not right, likewise N=17,19…… is not right if N=13, there must be a line having at least 3 segments, but that requires at least 6 addtional lines each having 2 segments, 6*2+3=15>13

Hariram K
Aug 28, 2013

First we will show that if the condition is false is not possible for some N, then it is not possible for N+2 as well.

For the condition to be not satisfied by a polygon, it must not be convex. So we can always find a line which cuts the polygon in four points. If only two vertices of the polygon are in one side, we cut the polygon along this line (like cutting a piece of that polygon shaped paper. look at the linked image for clarity). This new polygon would still satisfy the conditions and has 2 more sides. (I assumed that we can always find a polygon that can be cut with only two vertices on one side)

The linked image shows example of 10-sided and 15-sided polygon for which the condition is not true. By what I proved earlier, the condition is not true for 10+2m and 15+2m sided polygons for integer m. So it is not possible for 10, 12, 14, 15, 16, 17, .... Hence it appears that 13 is the largest value of N for which the condition gets met. All we need to show that there are no 13-gon violating the condition.

Since 13 is an odd number there must be at least one line which contains odd number of edges of our 13-gon. If our polygon has to violate the condition, each such line must have more than 1 edge. So it must have a line which has 3 edges. This line will have 6 collinear vertices of the polygon. There must be 6 more edges in the polygon each passing through one of these vertices. All these edges must be mutually non-collinear. They all must lie on a line which has more than 1 edge. So our polygon must have 6 more edges. Hence I have counted total 3+6+6 distinct edges which our odd-sided polygon must have in order to violate the conditions. So a 13-gon will always satisfy the condition.

"I assumed that we can always find a polygon that can be cut with only two vertices on one side" - I think this needs more work , for example the 20-gon in the question doesn't seem to have such a cut.

Matt McNabb - 7 years, 9 months ago

Log in to reply

Agreed!

The diagrams that I worked with similar to Hariram's. Like you mentioned, the ability to "cut the polygon ... and has 2 more sides" has further conditions with regards to the gradients / slopes of the adjacent lines, and hence may not always continue.

I ended up having to construct a family of stars (similar to your comment "Example with N = 17 N=17 >"). I do not believe that an inductive approach will work (in the sense of using the same base star and cutting more parts of it).

Calvin Lin Staff - 7 years, 9 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...