An Amazing Geometry Problem
by djmathman, Sep 8, 2020, 7:47 PM
This problem statement is so cool 
KöMaL A.774 (March 2020), Lajos Fonyó. Let
be the circumcenter of triangle
, and
be an arbitrary point on the circumcircle of
. Let points
,
and
be the orthogonal projections of point
onto lines
,
and
, respectively. Prove that the incenter of triangle
is on the Simson-Wallace line of triangle
corresponding to point
.
Solution
I know I say this all the time, but I'm very happy to have solved this
.

KöMaL A.774 (March 2020), Lajos Fonyó. Let














Solution
Hopefully my use of directed angles is correct (though tbh it probably isn't).
Without loss of generality assume that
lies on the arc
not containing
. Denote by
the orthocenter of
; let
,
, and
be the feet of the altitudes from
,
, and
, respectively; and let
,
, and
be the reflections of
about
,
, and
, also respectively. Furthermore, let
be the Simson line of
with respect to
, and let
,
, and
be the reflections of
about
,
, and
, also respectively. Finally, denote by
the incenter of
. (Points
,
, and
are omitted from the diagram below for clarity reasons.)
![[asy]
size(300);
defaultpen(linewidth(0.6)+fontsize(10));
pair A = (0,6), B = (-2,0), C = (4,0), H = orthocenter(A,B,C), O = circumcenter(A,B,C), D = rotate(114,O)*C;
pair Ap = foot(D,B,C), Bp = foot(D,A,C), Cp = foot(D,A,B), X = foot(D,O,A), Y = foot(D,O,B), Z = foot(D,O,C);
pair P = foot(A,B,C), Q = foot(B,A,C), R = foot(C,A,B);
pair Pp = 2* P - H, Qp = 2*Q-H, Rp = 2*R-H, Xp = 2*X-D, Yp = 2*Y-D, Zp = 2*Z-D;
pair I = incenter(Xp,Yp,Zp);
draw(A--B--C--A,lightblue);
draw(circumcircle(A,B,C)^^O--A^^O--B^^O--C,heavygreen);
draw(Xp--Yp--Zp--cycle,red);
draw(Ap--Cp,orange);
draw(H--I,orange+linetype("3 3"));
draw(A--Pp^^B--Qp^^C--Rp,purple);
draw(A--Cp--D^^Bp--D^^Ap--D,magenta+linetype("3 3"));
draw(Pp--Qp--Rp--cycle,red+linetype("3 3"));
dot("$A$",A,NW,linewidth(3.3));
dot("$B$",B,dir(O--B),linewidth(3.3));
dot("$C$",C,dir(O--C),linewidth(3.3));
dot("$D$",D,dir(O--D),linewidth(3.3));
dot("$H$",H,2.3*dir(65),linewidth(3.3));
dot("$I$",I,dir(H--I),linewidth(3.3));
dot("$O$",O,S,linewidth(3.3));
dot("$X'$",Xp,dir(O--Xp),linewidth(3.3));
dot("$Y'$",Yp,dir(O--Yp),linewidth(3.3));
dot("$Z'$",Zp,dir(O--Zp),linewidth(3.3));
dot("$P'$",Pp,dir(O--Pp),linewidth(3.3));
dot("$Q'$",Qp,dir(O--Qp),linewidth(3.3));
dot("$R'$",Rp,dir(O--Rp),linewidth(3.3));
dot("$A'$",Ap,dir(D--Ap),linewidth(3.3));
dot("$B'$",Bp,dir(D--Bp),linewidth(3.3));
dot("$C'$",Cp,dir(D--Cp),linewidth(3.3));
[/asy]](//latex.artofproblemsolving.com/0/9/1/091686c2bbfa34b3a267336e0eda65a84efedeaf.png)
We make the following crucial claim.
Claim. Triangles
and
are congruent.
Proof. Observe that points
,
, and
all lie on the circle with diameter
. Some angle chasing reveals that
Analogous (though not entirely congruent) angle chases hold for the other two pairs of angles. This means that
; combined with the fact that the circumcircles coincide, we deduce the desired claim. 
Since the two triangles are inversely congruent, there exists a reflection
mapping
to
. Furthermore,
is the incenter of
and is hence also the incenter of
; thus,
also maps
to
, meaning
, for instance.
Now another angle chase reveals
That is, the angle line
makes with the line
is equal to the angle it makes with
. Combining both previous claims, therefore, yields that
.
Finally, observe that the homothety
sends
to
. Further recall that the Simson line of
bisects
, so
also sends
to a point on
. This means that
also sends
to a point on
. But
also sends
to the incenter of
, and so this incenter must lie on
, which is what we were after. 
Without loss of generality assume that

































![[asy]
size(300);
defaultpen(linewidth(0.6)+fontsize(10));
pair A = (0,6), B = (-2,0), C = (4,0), H = orthocenter(A,B,C), O = circumcenter(A,B,C), D = rotate(114,O)*C;
pair Ap = foot(D,B,C), Bp = foot(D,A,C), Cp = foot(D,A,B), X = foot(D,O,A), Y = foot(D,O,B), Z = foot(D,O,C);
pair P = foot(A,B,C), Q = foot(B,A,C), R = foot(C,A,B);
pair Pp = 2* P - H, Qp = 2*Q-H, Rp = 2*R-H, Xp = 2*X-D, Yp = 2*Y-D, Zp = 2*Z-D;
pair I = incenter(Xp,Yp,Zp);
draw(A--B--C--A,lightblue);
draw(circumcircle(A,B,C)^^O--A^^O--B^^O--C,heavygreen);
draw(Xp--Yp--Zp--cycle,red);
draw(Ap--Cp,orange);
draw(H--I,orange+linetype("3 3"));
draw(A--Pp^^B--Qp^^C--Rp,purple);
draw(A--Cp--D^^Bp--D^^Ap--D,magenta+linetype("3 3"));
draw(Pp--Qp--Rp--cycle,red+linetype("3 3"));
dot("$A$",A,NW,linewidth(3.3));
dot("$B$",B,dir(O--B),linewidth(3.3));
dot("$C$",C,dir(O--C),linewidth(3.3));
dot("$D$",D,dir(O--D),linewidth(3.3));
dot("$H$",H,2.3*dir(65),linewidth(3.3));
dot("$I$",I,dir(H--I),linewidth(3.3));
dot("$O$",O,S,linewidth(3.3));
dot("$X'$",Xp,dir(O--Xp),linewidth(3.3));
dot("$Y'$",Yp,dir(O--Yp),linewidth(3.3));
dot("$Z'$",Zp,dir(O--Zp),linewidth(3.3));
dot("$P'$",Pp,dir(O--Pp),linewidth(3.3));
dot("$Q'$",Qp,dir(O--Qp),linewidth(3.3));
dot("$R'$",Rp,dir(O--Rp),linewidth(3.3));
dot("$A'$",Ap,dir(D--Ap),linewidth(3.3));
dot("$B'$",Bp,dir(D--Bp),linewidth(3.3));
dot("$C'$",Cp,dir(D--Cp),linewidth(3.3));
[/asy]](http://latex.artofproblemsolving.com/0/9/1/091686c2bbfa34b3a267336e0eda65a84efedeaf.png)
We make the following crucial claim.
Claim. Triangles


Proof. Observe that points







Since the two triangles are inversely congruent, there exists a reflection










Now another angle chase reveals





Finally, observe that the homothety
















I know I say this all the time, but I'm very happy to have solved this

This post has been edited 5 times. Last edited by djmathman, Sep 9, 2020, 4:38 PM