It's been a while since I've posted math on here
by djmathman, Jul 19, 2024, 1:21 AM
so have a random collection of problems I've solved over the past
months, where
.
CMIMC 2024 Algebra 9 (Robert Trosten). Let
be the non-negative rational numbers,
such that
,
for
, and
Define a sequence
of non-negative integers recursively via
for every
. Find 
Solution
CMIMC 2024 Algebra 10 (Connor Gordon). There exists a unique pair of polynomials
such that
Compute
.
Solution
Romania District Olympiad 2024 Grade 12 Problem 4. Let
be a differentiable function, with a continous derivative. Given that
and
for every
prove that
for any positive integer
and real number 
Solution
OTIS Mock AIME Problem 15 (Wilbert Chu). A parabola in the Cartesian plane is tangent to the
-axis at
and to the
-axis at
. Find the sum of the coordinates of the vertex of the parabola.
Solution
OMMC PoTM March 2022 (Evan Chang). Define acute triangle
with circumcircle
Let
be the midpoint of minor arc
in
and let
be the reflection of
over
If the circle with diameter
is tangent to the external angle bisector of
at
show 
Solution
BAMO 2010 Problem 4. Acute triangle
has
. Point
lies in the interior of triangle
so that
and
. Point
is the reflection of
across line
, and point
is the reflection of
across line
. Prove that lines
and
are perpendicular.
Solution
Naoki Sato. In triangle
, sides
and
are extended to
and
, respectively, so that
is parallel to
and tangent to the
-excircle. Construct the circle that passes through
and
, and is tangent to the
-excircle. Prove that this circle is also tangent to the incircle of triangle
.
![[asy]
real area(pair A, pair B, pair C) {
return(abs((xpart(A)*ypart(B) + xpart(B)*ypart(C) + xpart(C)*ypart(A) - ypart(A)*xpart(B) - ypart(B)*xpart(C) - ypart(C)*xpart(A)))/2);
};
pair excentre(pair A, pair B, pair C) {
return((-abs(B - C)*A + abs(C - A)*B + abs(A - B)*C)/(-abs(B - C) + abs(C - A) + abs(A - B)));
};
real exradius(pair A, pair B, pair C) {
return(2*area(A,B,C)/(-abs(B - C) + abs(C - A) + abs(A - B)));
};
path excircle(pair A, pair B, pair C) {
return(Circle(excentre(A,B,C),exradius(A,B,C)));
};
unitsize (0.28 cm);
pair A, B, C, D, E, P, T, X;
A = (0,0);
B = (-1,-4);
C = (5,-4);
D = interp(A,B,exradius(A,B,C)/inradius(A,B,C));
E = interp(A,C,exradius(A,B,C)/inradius(A,B,C));
X = (-9.19135,-14.6049);
T = (incenter(A,D,E) + reflect(D,E)*(incenter(A,D,E)))/2;
P = reflect(X,incenter(A,D,E))*(T);
draw(incircle(A,B,C),red);
draw(excircle(A,B,C),red);
draw(circumcircle(D,E,P),blue);
draw(A--D--E--cycle);
draw(B--C);
label("$A$", A, N);
label("$B$", B, NW);
label("$C$", C, NE);
label("$D$", D, SW);
label("$E$", E, SE);
[/asy]](//latex.artofproblemsolving.com/7/3/b/73be5d3f064c85d2bfe424bec10c98f69e6f6da9.png)
Solution![[asy]
defaultpen(fontsize(10pt));
real area(pair A, pair B, pair C) {
return(abs((xpart(A)*ypart(B) + xpart(B)*ypart(C) + xpart(C)*ypart(A) - ypart(A)*xpart(B) - ypart(B)*xpart(C) - ypart(C)*xpart(A)))/2);
};
pair excentre(pair A, pair B, pair C) {
return((-abs(B - C)*A + abs(C - A)*B + abs(A - B)*C)/(-abs(B - C) + abs(C - A) + abs(A - B)));
};
real exradius(pair A, pair B, pair C) {
return(2*area(A,B,C)/(-abs(B - C) + abs(C - A) + abs(A - B)));
};
path excircle(pair A, pair B, pair C) {
return(Circle(excentre(A,B,C),exradius(A,B,C)));
};
unitsize (0.3 cm);
pair A, B, C, D, E, I, Ia, U, V, P, Q, R, S, T;
A = (0,0);
B = (-1,-4);
C = (5,-4);
D = interp(A,B,exradius(A,B,C)/inradius(A,B,C));
E = interp(A,C,exradius(A,B,C)/inradius(A,B,C));
I = incenter(A,B,C); Ia = excentre(A,B,C);
T = foot(I, B,C); U = foot(I,A,B); V = foot(I,A,C); S = foot(Ia, B, C); P = foot(Ia,A,D); Q = foot(Ia,A,E); R = foot(Ia,D,E);
draw(incircle(A,B,C),red);
draw(excircle(A,B,C),red);
draw(A--D--E--cycle);
draw(B--C);
label("$A$", A, N);
label("$B$", B, NW);
label("$C$", C, NE);
label("$D$", D, SW);
label("$E$", E, SE);
label("$U$",U,NW);
label("$V$",V,NE);
label("$S$",S,dir(270));
label("$T$",T, N);
label("$P$",P,NW);
label("$Q$",Q,NE);
label("$R$",R,dir(270));
[/asy]](//latex.artofproblemsolving.com/0/d/c/0dcf24d5798ec2381580050753191675c75b8e61.png)
Assume without loss (as in the diagram above) that
. Let
be the incircle of
and
its
-excircle. Let
and
be the tangency points of
and
, respectively, with
, and let
,
, and
be the tangency points of
with
,
, and
, also respectively. Using standard triangle notation, write
,
, and
. Therefore
Scaling up by a factor of
yields the equivalent equality
It follows by the converse to Casey's Theorem that there exists a circle passing through the point circles
and
which is internally tangent to
and externally tangent to
. This is precisely the blue circle in the problem statement. 


CMIMC 2024 Algebra 9 (Robert Trosten). Let










Solution
This is a problem about continued fractions in disguise. Let
for any real numbers
.
We first make two claims about the behavior of
.
Lemma 1 (Behavior on simple continued fractions). If
is the simple continued fraction representation of a rational number
, then
Proof. Induction. The base case holds by inspection, and for the inductive step, remark that
![\begin{align*}
f([a_0;a_1,a_2,\ldots, a_n]) &= a_0 + f([0;a_1,a_2,\ldots, a_n]) \\
&= a_0 + f([a_1;a_2,\ldots, a_n]) \\
&\stackrel{\textrm{IH}}=a_0 + a_1 + \cdots + a_n. \quad\square
\end{align*}](//latex.artofproblemsolving.com/e/d/5/ed52335bf9cb0f091367ec622f71da2aac02221f.png)
Lemma 2 (Behavior on negative inputs). If
, where each
, then
![\[
f(q) = a_0 + a_1 + \cdots + a_n - n.
\]](//latex.artofproblemsolving.com/b/2/1/b214a5f947e5a15ff66c5c5a47d87cdadcb169d9.png)
Proof. Remark that
so
. That is, the minus sign serves to subtract
from the overall sum when
is applied. Another round of induction finishes. 
We now proceed with the original problem. Define
for each positive integer
; compute
and
. The characteristic polynomial of the recurrence
is
, which has roots
and
satisfying
and
. Thus
has characteristic polynomial with roots
and
, which can be computed by any method of choice to be
.
This means
or
It follows from induction and
that
so
.
![\[
[x_0;x_1,x_2,\ldots, x_n] := x_0 + \cfrac{1}{x_1 + \cfrac{1}{x_2 + \cfrac{1}{\cdots + \cfrac{1}{x_n}}}}
\]](http://latex.artofproblemsolving.com/0/7/b/07b589ccfcbfca216daa9fe0288a4270ac9d07f3.png)

We first make two claims about the behavior of

Lemma 1 (Behavior on simple continued fractions). If
![$q\sim [a_0;a_1,a_2,\ldots, a_n]$](http://latex.artofproblemsolving.com/9/d/a/9dad8b74b7b2e832dc4267954207326c07981fff.png)

![\[
f(q) = a_0 + a_1 + a_2 + \cdots + a_n.
\]](http://latex.artofproblemsolving.com/a/d/1/ad1d3439077e871af7f05df0dde30352fce0440d.png)
![\begin{align*}
f([a_0;a_1,a_2,\ldots, a_n]) &= a_0 + f([0;a_1,a_2,\ldots, a_n]) \\
&= a_0 + f([a_1;a_2,\ldots, a_n]) \\
&\stackrel{\textrm{IH}}=a_0 + a_1 + \cdots + a_n. \quad\square
\end{align*}](http://latex.artofproblemsolving.com/e/d/5/ed52335bf9cb0f091367ec622f71da2aac02221f.png)
Lemma 2 (Behavior on negative inputs). If
![$q = [a_0;-a_1,-a_2,\ldots, -a_n]$](http://latex.artofproblemsolving.com/d/b/1/db1aeefe3c3e5348abaee63575c92d3e2adfb40c.png)

![\[
f(q) = a_0 + a_1 + \cdots + a_n - n.
\]](http://latex.artofproblemsolving.com/b/2/1/b214a5f947e5a15ff66c5c5a47d87cdadcb169d9.png)
Proof. Remark that
![\[
a - \frac 1q = a - 1 + \frac{q-1}q = a - 1 + \cfrac{1}{\frac{q}{q-1}} = a - 1 + \cfrac{1}{1 + \cfrac{1}{q-1}},
\]](http://latex.artofproblemsolving.com/a/f/7/af75e83c81abb9ea00551a8dc1c0681de485c4e1.png)




We now proceed with the original problem. Define














This means
![\[
Q_n = 34 Q_{n-1} - Q_{n-2}\text{ for every }n\geq 2,
\]](http://latex.artofproblemsolving.com/f/4/c/f4cd879461ea3c10aac7be5b061c5528dfcdb1ea.png)
![\[
\frac{Q_{n}}{Q_{n-1}} = 34 - \frac{Q_{n-2}}{Q_{n-1}}.
\]](http://latex.artofproblemsolving.com/5/f/0/5f0e4fd6ca5f27f004feb90362108439181c9771.png)

![\[
\frac{Q_6}{Q_5} = [34;-34,-34,-34,-34],
\]](http://latex.artofproblemsolving.com/5/c/4/5c48d00b817766ca497d36c6b7ff1a95524a044a.png)

CMIMC 2024 Algebra 10 (Connor Gordon). There exists a unique pair of polynomials



Solution
By comparing degrees, deduce that
and
are both quadratics. Furthermore, the leading coefficients of
and
are either
or
.
Let
and
be the roots of
, and define
and
analogously. Observe that
is the product of two factors, each of whose sum of roots is
. Comparing the sum of the roots of both sides in the first equation yields the relation
. Analogous reasoning with the second equation yields the relation
. Solving yields
and
. That is,
for some independent choices of signs.
To compute
and
, observe that
, so
. It follows that
Regardless of the leading coefficient of
, this equation simplifies to
, or
. Analogously,
, so
and
Regardless of the leading coefficient of
, this equation simplifies to
, or
.
The above analysis implies
and
. To determine the exact signs, note
, so
, which can only occur if
. Analogously,
, so
, which can only occur if
.
Finally, compute
![\[
P(10) + Q(-10) = (10^2 - 4\cdot 10 + 2) - [(-10)^2 - 5\cdot(-10) + 2] = \boxed{-90}.
\]](//latex.artofproblemsolving.com/1/a/1/1a1dc6dd3e0275664b5c859e27c698d22a5ef159.png)






Let





![\[
P(Q(x)) = \pm(Q(x) - r_P)(Q(x) - r_Q)
\]](http://latex.artofproblemsolving.com/c/a/b/cab19121e6f91dca559585ac4d6ff0d8716eaeb4.png)





![\[
P(x) = \pm(x^2 - 4x + b)\quad\text{and}\quad Q(x) = \pm(x^2 - 5x + d)
\]](http://latex.artofproblemsolving.com/8/a/8/8a8fa51cad7e1da4240c6b081cf72d45847347c8.png)
To compute




![\[
P(0)(0^2 - 6\cdot 0 + 7) = P(5)(5^2 - 6\cdot 5 + 7).
\]](http://latex.artofproblemsolving.com/e/0/1/e014ecb505e45384d0d7eb2d0f28d156d6897464.png)





![\[
Q(0)(0^2 - 3\cdot 0 - 2) = Q(4)(4^2 - 3\cdot 4 - 2).
\]](http://latex.artofproblemsolving.com/9/e/b/9eb73713432fb448236f791d9fd2e0d130cf467a.png)



The above analysis implies








Finally, compute
![\[
P(10) + Q(-10) = (10^2 - 4\cdot 10 + 2) - [(-10)^2 - 5\cdot(-10) + 2] = \boxed{-90}.
\]](http://latex.artofproblemsolving.com/1/a/1/1a1dc6dd3e0275664b5c859e27c698d22a5ef159.png)
Romania District Olympiad 2024 Grade 12 Problem 4. Let




![\[\frac{1}{n+1}\int_0^af(t)^{2n+1}\mathrm{d}t\leqslant\left(\int_0^af(t)^n\mathrm{d}t\right)^2,\]](http://latex.artofproblemsolving.com/5/f/1/5f19be19c27318882cecb0e6e013cf3e2678f0a9.png)


Solution
We'll improve the bound by a factor of
.
First remark that
because
for all
.
Let
be arbitrary. Starting with the bound
, integrating both sides from
to
yields
Multiplying both sides by
gives
But, by FTC Part 1, the left hand side is the derivative of
evaluated at
, while the right hand side is the derivative of
evaluated at
. Integrating both sides and evaluating at
gives the desired inequality improved by a factor of
.
Equality holds when
, which works when plugged in:
![\[ \frac{1}{n+1}\int_0^a t^{2n+1}\,dt = \frac12 a^{2(n+1)} = \frac12\left(a^{n+1}\right)^2 = \frac12\left(\int_0^a t^n\,dt\right)^2. \]](//latex.artofproblemsolving.com/e/8/6/e866ec7cda141a8642327b410e286868417581ec.png)

First remark that



Let




![\[ \frac{f(u)^{n+1}}{n+1} \leq \int_0^u f(t)^n\,dt. \]](http://latex.artofproblemsolving.com/c/0/f/c0fc7b01fc565b60f65e589469f820644a2e498b.png)

![\[ \frac{f(u)^{2n+1}}{n+1} \leq f(u)^n\int_0^u f(t)^n\,dt. \]](http://latex.artofproblemsolving.com/0/6/7/06734e97a43a4f80b46bf69bd916b6b781bec90f.png)






Equality holds when

![\[ \frac{1}{n+1}\int_0^a t^{2n+1}\,dt = \frac12 a^{2(n+1)} = \frac12\left(a^{n+1}\right)^2 = \frac12\left(\int_0^a t^n\,dt\right)^2. \]](http://latex.artofproblemsolving.com/e/8/6/e866ec7cda141a8642327b410e286868417581ec.png)
OTIS Mock AIME Problem 15 (Wilbert Chu). A parabola in the Cartesian plane is tangent to the




Solution
Let
and
be the tangency points of the parabola with the
and
axes, respectively, let
be the origin, and let
and
be the focus and vertex, respectively, of the parabola. Furthermore, let
and
be the projections of
and
, respectively, onto the directrix of the parabola. We make the following series of claims about the configuration.
Claim. The following facts are true.
![[asy]
size(180);
defaultpen(linewidth(0.7));
pair O = origin, X = (2,0), Y = (-2,0), F = 2 * dir(58), V =(F.x,F.y/2), T = (F.x,0);
pair AA = rotate(90,F) * O, P = extension(F,AA,X,(2,1)), Q = extension(F,AA,Y,(-2,1));
real f (real x)
{
real c = (P.y-V.y)/(P.x-V.x)^2;
return c*(x - V.x)^2 + V.y;
}
draw(graph(f, -2.3,2.3),gray(0.6),Arrows(size=7));
draw(P--Q--O--cycle^^rightanglemark(P,O,Q,5),rgb(0.9,0.1,0.1));
draw(O--F^^rightanglemark(O,F,P,5),rgb(0.95,0.35,0.25));
draw(X--Y, rgb(0.1,0.1,0.7));
draw(Q--Y^^F--T^^P--X^^rightanglemark(Q,Y,O,5)^^rightanglemark(F,T,X,5)^^rightanglemark(P,X,O,5), rgb(0.6,0.1,0.6));
dot("$F$",F,NE,linewidth(3.3));
dot("$P$",P,SE,linewidth(3.3));
dot("$Q$",Q,NE,linewidth(3.3));
dot("$X$",X,S,linewidth(3.3));
dot("$Y$",Y,S,linewidth(3.3));
dot("$T$",T,S,linewidth(3.3));
dot("$V$",V,SW,linewidth(3.3));
dot("$O$",O,S,linewidth(3.3));
[/asy]](//latex.artofproblemsolving.com/0/1/3/0137bd9deb49c89e0a049f1e4bb9f0d30cacd4c0.png)
Proof. This is a combination of several known facts, but in the interest of completeness we prove this only assuming the reflection property for parabolas.
Observe that
is the perpendicular bisector of
, and analogously
is the perpendicular bisector of
. It follows that
is the circumcenter of
. Because
, it follows that
too, and that
lies on
. This proves part 1.
For part 2, remark that
is the reflection of
about line
, so
. Analogously,
. This proves part 2.
Finally, for part 3, let
and
. Denote by
the foot of the perpendicular from
to
; remark that
is the midpoint of
. Remark that the length
is a weighted average of the lengths
and
; more specifically,
Thus
. It follows by similarity ratios that
and
, proving all parts of part 3. 
From here, we may finish. Because
has coordinates
and
has coordinates
, the point
has coordinates
. It follows that
has coordinates
, so
Hence the sum of the coordinates is
.











Claim. The following facts are true.
- The directrix
of the parabola passes through
.
- Point
lies on
, and
is the foot of the altitude from
to
.
- Points
,
, and
are collinear, as are
,
, and
. In particular,
![[asy]
size(180);
defaultpen(linewidth(0.7));
pair O = origin, X = (2,0), Y = (-2,0), F = 2 * dir(58), V =(F.x,F.y/2), T = (F.x,0);
pair AA = rotate(90,F) * O, P = extension(F,AA,X,(2,1)), Q = extension(F,AA,Y,(-2,1));
real f (real x)
{
real c = (P.y-V.y)/(P.x-V.x)^2;
return c*(x - V.x)^2 + V.y;
}
draw(graph(f, -2.3,2.3),gray(0.6),Arrows(size=7));
draw(P--Q--O--cycle^^rightanglemark(P,O,Q,5),rgb(0.9,0.1,0.1));
draw(O--F^^rightanglemark(O,F,P,5),rgb(0.95,0.35,0.25));
draw(X--Y, rgb(0.1,0.1,0.7));
draw(Q--Y^^F--T^^P--X^^rightanglemark(Q,Y,O,5)^^rightanglemark(F,T,X,5)^^rightanglemark(P,X,O,5), rgb(0.6,0.1,0.6));
dot("$F$",F,NE,linewidth(3.3));
dot("$P$",P,SE,linewidth(3.3));
dot("$Q$",Q,NE,linewidth(3.3));
dot("$X$",X,S,linewidth(3.3));
dot("$Y$",Y,S,linewidth(3.3));
dot("$T$",T,S,linewidth(3.3));
dot("$V$",V,SW,linewidth(3.3));
dot("$O$",O,S,linewidth(3.3));
[/asy]](http://latex.artofproblemsolving.com/0/1/3/0137bd9deb49c89e0a049f1e4bb9f0d30cacd4c0.png)
Proof. This is a combination of several known facts, but in the interest of completeness we prove this only assuming the reflection property for parabolas.
Observe that










For part 2, remark that





Finally, for part 3, let










![\[
FT = \frac{x}{x+y}\cdot QY + \frac{y}{x+y}\cdot PX = \frac{2xy}{x+y}.
\]](http://latex.artofproblemsolving.com/7/9/d/79d1665171b6c7d98a929090df64d0be62215a3f.png)




From here, we may finish. Because








![\[
V = \frac{9}{10}X + \frac{1}{10}Q = \left(\frac{81}{100},\frac{3}{100}\right).
\]](http://latex.artofproblemsolving.com/d/b/3/db3ada37e0910a1a2bea15862728c18393d2c572.png)

OMMC PoTM March 2022 (Evan Chang). Define acute triangle












Solution
Let
be the midpoint of
and
the intersection point of
and
. Further, let
be the circle with diameter
. We proceed with two claims.
Claim 1. Lines
and
are perpendicular.
Proof. Let
be the intersection point of
and
. From the perspective of
,
is the external angle bisector and
is the internal angle bisector, ergo
. But this means
lies on the polar of
with respect to
, ergo
. 
Claim 2. Segments
and
have the same length.
Proof. By Power of a Point,
Additionally, by Claim 1, triangles
and
are similar, so
. It follows that
. 
From here, we see that
, so
. Equivalently,
. 







Claim 1. Lines


Proof. Let












Claim 2. Segments


Proof. By Power of a Point,
![\[
TA\cdot TQ = TB\cdot TC = TP^2.
\]](http://latex.artofproblemsolving.com/0/8/5/085cadc1afe96cb4b9672feefdc39d6bc6999276.png)





From here, we see that




BAMO 2010 Problem 4. Acute triangle














Solution
Let
be the intersection point of
and
. Compute
from an angle chase, hence
is cyclic. Because
is the circumcenter of
, it follows that
is the circumcenter of
as well. Rephrasing the problem from the perpsective of
, we thus get the following equivalent statement.
Equivalent Problem. Let
be a triangle with circumcenter
and
-excenter
. Points
and
lie on rays
and
such that
. Prove that
.
There are several ways to prove this. One way to do so is to let
be the circumcircle of
and observe the computations
and
The claim follows by the Perpendicularity Lemma.
Another way is to observe that
, where
and
are the feet of the
- and
- internal angle bisectors of
, and then nuke it.










Equivalent Problem. Let










There are several ways to prove this. One way to do so is to let




Another way is to observe that






Naoki Sato. In triangle












![[asy]
real area(pair A, pair B, pair C) {
return(abs((xpart(A)*ypart(B) + xpart(B)*ypart(C) + xpart(C)*ypart(A) - ypart(A)*xpart(B) - ypart(B)*xpart(C) - ypart(C)*xpart(A)))/2);
};
pair excentre(pair A, pair B, pair C) {
return((-abs(B - C)*A + abs(C - A)*B + abs(A - B)*C)/(-abs(B - C) + abs(C - A) + abs(A - B)));
};
real exradius(pair A, pair B, pair C) {
return(2*area(A,B,C)/(-abs(B - C) + abs(C - A) + abs(A - B)));
};
path excircle(pair A, pair B, pair C) {
return(Circle(excentre(A,B,C),exradius(A,B,C)));
};
unitsize (0.28 cm);
pair A, B, C, D, E, P, T, X;
A = (0,0);
B = (-1,-4);
C = (5,-4);
D = interp(A,B,exradius(A,B,C)/inradius(A,B,C));
E = interp(A,C,exradius(A,B,C)/inradius(A,B,C));
X = (-9.19135,-14.6049);
T = (incenter(A,D,E) + reflect(D,E)*(incenter(A,D,E)))/2;
P = reflect(X,incenter(A,D,E))*(T);
draw(incircle(A,B,C),red);
draw(excircle(A,B,C),red);
draw(circumcircle(D,E,P),blue);
draw(A--D--E--cycle);
draw(B--C);
label("$A$", A, N);
label("$B$", B, NW);
label("$C$", C, NE);
label("$D$", D, SW);
label("$E$", E, SE);
[/asy]](http://latex.artofproblemsolving.com/7/3/b/73be5d3f064c85d2bfe424bec10c98f69e6f6da9.png)
Solution
![[asy]
defaultpen(fontsize(10pt));
real area(pair A, pair B, pair C) {
return(abs((xpart(A)*ypart(B) + xpart(B)*ypart(C) + xpart(C)*ypart(A) - ypart(A)*xpart(B) - ypart(B)*xpart(C) - ypart(C)*xpart(A)))/2);
};
pair excentre(pair A, pair B, pair C) {
return((-abs(B - C)*A + abs(C - A)*B + abs(A - B)*C)/(-abs(B - C) + abs(C - A) + abs(A - B)));
};
real exradius(pair A, pair B, pair C) {
return(2*area(A,B,C)/(-abs(B - C) + abs(C - A) + abs(A - B)));
};
path excircle(pair A, pair B, pair C) {
return(Circle(excentre(A,B,C),exradius(A,B,C)));
};
unitsize (0.3 cm);
pair A, B, C, D, E, I, Ia, U, V, P, Q, R, S, T;
A = (0,0);
B = (-1,-4);
C = (5,-4);
D = interp(A,B,exradius(A,B,C)/inradius(A,B,C));
E = interp(A,C,exradius(A,B,C)/inradius(A,B,C));
I = incenter(A,B,C); Ia = excentre(A,B,C);
T = foot(I, B,C); U = foot(I,A,B); V = foot(I,A,C); S = foot(Ia, B, C); P = foot(Ia,A,D); Q = foot(Ia,A,E); R = foot(Ia,D,E);
draw(incircle(A,B,C),red);
draw(excircle(A,B,C),red);
draw(A--D--E--cycle);
draw(B--C);
label("$A$", A, N);
label("$B$", B, NW);
label("$C$", C, NE);
label("$D$", D, SW);
label("$E$", E, SE);
label("$U$",U,NW);
label("$V$",V,NE);
label("$S$",S,dir(270));
label("$T$",T, N);
label("$P$",P,NW);
label("$Q$",Q,NE);
label("$R$",R,dir(270));
[/asy]](http://latex.artofproblemsolving.com/0/d/c/0dcf24d5798ec2381580050753191675c75b8e61.png)
Assume without loss (as in the diagram above) that




















![\begin{align*}
BC\cdot TS + BT\cdot EV &= a(b-c) + (s-b)\left[a + \frac{(s-c)s}{s-a}\right]\\
&= a(b-c+s-b) + \frac{s(s-b)(s-c)}{s-a}\\
&= a(s-c) + (s-c)\cdot\frac{s(s-b)}{s-a} = CT\cdot DU.
\end{align*}](http://latex.artofproblemsolving.com/6/3/9/639556b1c44b845439b970df20e83432a1d76479.png)

![\[
DE\cdot TS + DR\cdot EV = ER\cdot DU.
\]](http://latex.artofproblemsolving.com/f/e/5/fe521d06e1828d35816824f80267aff0f212f309.png)





This post has been edited 2 times. Last edited by djmathman, Jul 19, 2024, 6:19 PM