Hmm so I guess I haven't posted here in a while
by djmathman, Jan 19, 2014, 4:50 AM
So... how's January been going for me? Well, interestingly, to say the least. yes that's an adverb, it's grammatically correct... darn you SAT Writing section
I guess this is a bit late, but I placed in Science Olympiad regionals! At the NJIT region, I received a 3rd place medal for Circuit Lab (that was my only event, as Astronomy conflicted) and our team placed sixth overall. The medal came as a complete shock (hehe, get it, electricity, hehe), since my teammate and I totally thought we failed that test. But I guess everyone else failed too, as it was probably a medium State level test, so it's ok. The sixth place finish means that it's our school's first year of qualifying for the State competition without the only-one-team-per-school rule! We're probably going to get killed in States, since powerhouses like West-Windsor Plainsboro, Montgomery, and Princeton weren't there, but hey, there's nothing like getting false hope beforehand, right?
School's been pretty hard so far. All my science courses, such as Chemistry and Physics, have been really easy so far, but all my humanities courses, such as Spanish, AP Lang, and APUSH II, have been incredibly difficult. I just got pummeled by an AP Lang Quarter-Exam as well D:. (Think of it as a college midterm of sorts; that's how hard it can get.)
As for math... idk. I've been continuing to work from 107 Geometry Problems and scattering through the fora/Contest pages for interesting problems. I'm still not at the level of USAMO yet; hopefully I will be when the time comes, of course assuming I qualify which is a nontrivial assumption.
So what specific math problems have I completed during these twenty days of not posting any math on my blog?
Solution
Solution I wrote during the contest (that can probably be modified to account for directed angles)![[asy]
import olympiad;
size(300);
defaultpen(linewidth(0.8));
pair A=(3,7),B=origin,C=(10,0),Hb=foot(B,A,C),Hc=foot(C,A,B),H=orthocenter(A,B,C),M=C/2;
path wb=circumcircle(H,B,M),wc=circumcircle(H,C,M);
pair P=intersectionpoint(wb,B--A),Q=intersectionpoint(wc,C--A);
pair Rx=10*H-9*P,Rs=2*H-P,R=intersectionpoint(wc,Rs--Rx);
pair Sx=10*H-9*Q,Ss=2*H-Q,S=intersectionpoint(wb,Ss--Sx);
draw(A--B--C--cycle^^C--Hc^^B--Hb^^wb^^wc^^P--R--C^^Q--S--B^^M--H);
draw(S--R,linetype("4 4"));
draw(rightanglemark(B,foot(B,A,C),C)^^rightanglemark(C,foot(C,A,B),B));
pair G=centroid(A,B,C);
label("$A$",A,dir(G--A));
label("$B$",B,dir(G--B));
label("$C$",C,dir(G--C));
label("$M$",M,2*dir(285));
label("$H$",H,2.5*dir(125));
label("$P$",P,dir(170));
label("$Q$",Q,dir(90));
label("$R$",R,dir(H--R));
label("$S$",S,dir(H--S));
[/asy]](//latex.artofproblemsolving.com/0/8/2/082b60f18a58805f78a5b4aff29bfbf91ea3339c.png)
Notice that
and
are both cyclic pentagons, whence
Next, remark that
implies that
, so
(here
is the projection of
onto
). Similarly,
, so
. Now consider the spiral similarity
sending
to
. This sends
to
and
to
, so
, which implies that
. But we know by cyclicity that
, so therefore
and
are collinear.
Finally, since
is the midpoint of
, we know that
and
. Adding both of these equations together gives
, as desired. 
Comment
I've also decided to go back and look at the British Olympiad Round 1 from 2008 that I printed out a while ago. I was looking through some stuff when I found it, and I wanted to see how much better I've gotten at math since I first found it... maybe two years ago?
Couldn't solve this one before so yay
Wait why is this higher than the previous one
is a point such that
is a cyclic quadrilateral, and since
, this quadrilateral must be an isosceles trapezoid.
If: We have
and
, so
, implying that
.
Only If: Since
is an isosceles trapezoid, we have
, so
. Now letting
, it can be seen that
, as desired.
We have proven the statement in both directions, so we are done.
Bashy but interesting
I guess this is a bit late, but I placed in Science Olympiad regionals! At the NJIT region, I received a 3rd place medal for Circuit Lab (that was my only event, as Astronomy conflicted) and our team placed sixth overall. The medal came as a complete shock (hehe, get it, electricity, hehe), since my teammate and I totally thought we failed that test. But I guess everyone else failed too, as it was probably a medium State level test, so it's ok. The sixth place finish means that it's our school's first year of qualifying for the State competition without the only-one-team-per-school rule! We're probably going to get killed in States, since powerhouses like West-Windsor Plainsboro, Montgomery, and Princeton weren't there, but hey, there's nothing like getting false hope beforehand, right?

School's been pretty hard so far. All my science courses, such as Chemistry and Physics, have been really easy so far, but all my humanities courses, such as Spanish, AP Lang, and APUSH II, have been incredibly difficult. I just got pummeled by an AP Lang Quarter-Exam as well D:. (Think of it as a college midterm of sorts; that's how hard it can get.)
As for math... idk. I've been continuing to work from 107 Geometry Problems and scattering through the fora/Contest pages for interesting problems. I'm still not at the level of USAMO yet; hopefully I will be when the time comes, of course assuming I qualify which is a nontrivial assumption.
So what specific math problems have I completed during these twenty days of not posting any math on my blog?
All-Russian MO 2001 wrote:
Let the circle
be internally tangent to another circle
at
.Take a point
on
and draw a tangent
which intersects
at
and
. Let
be the midpoint of the arc
which is on the opposite side of
. Prove that the circumradius of the
doesnt depend on the choice of
.














Solution
Note that by Archimedes' Lemma (or a simple homothety argument),
bisects
. (This follows from the fact that the homothety sending
to
sends
to the tangent at
, which implies that
is the midpoint of arc
.) Therefore
, and as such
. Let
be the ratio of similitude of the two triangles. We wish to show that
is constant, as since the circumradius of
is constant the circumradius of
will follow to be constant.
Note that
. Hence
Thus, we have
However, note that the homothety centered at
sending
to
sends
to
, and as such
. Solving gives
, which is fixed, as desired. 














Note that











Winter NIMO 2014 wrote:
Let
be an acute triangle with orthocenter
and let
be the midpoint of
. (The orthocenter is the point at the intersection of the three altitudes.) Denote by
the circle passing through
,
, and
, and denote by
the circle passing through
,
, and
. Lines
and
meet
and
again at
and
, respectively. Rays
and
meet
and
again at
and
, respectively. Show that
and
have the same area.


























Solution I wrote during the contest (that can probably be modified to account for directed angles)
![[asy]
import olympiad;
size(300);
defaultpen(linewidth(0.8));
pair A=(3,7),B=origin,C=(10,0),Hb=foot(B,A,C),Hc=foot(C,A,B),H=orthocenter(A,B,C),M=C/2;
path wb=circumcircle(H,B,M),wc=circumcircle(H,C,M);
pair P=intersectionpoint(wb,B--A),Q=intersectionpoint(wc,C--A);
pair Rx=10*H-9*P,Rs=2*H-P,R=intersectionpoint(wc,Rs--Rx);
pair Sx=10*H-9*Q,Ss=2*H-Q,S=intersectionpoint(wb,Ss--Sx);
draw(A--B--C--cycle^^C--Hc^^B--Hb^^wb^^wc^^P--R--C^^Q--S--B^^M--H);
draw(S--R,linetype("4 4"));
draw(rightanglemark(B,foot(B,A,C),C)^^rightanglemark(C,foot(C,A,B),B));
pair G=centroid(A,B,C);
label("$A$",A,dir(G--A));
label("$B$",B,dir(G--B));
label("$C$",C,dir(G--C));
label("$M$",M,2*dir(285));
label("$H$",H,2.5*dir(125));
label("$P$",P,dir(170));
label("$Q$",Q,dir(90));
label("$R$",R,dir(H--R));
label("$S$",S,dir(H--S));
[/asy]](http://latex.artofproblemsolving.com/0/8/2/082b60f18a58805f78a5b4aff29bfbf91ea3339c.png)
Notice that


![\[\angle APH=\angle BSH=\angle BMH=\angle HQC=\angle HRC.\]](http://latex.artofproblemsolving.com/8/1/7/817d673631382cef8753ee7290eecad188fc25b2.png)




















Finally, since


![$[BSM]=[MSC]$](http://latex.artofproblemsolving.com/f/c/a/fca3f61ab6468bfd6562d1caec6d1d2899d9492c.png)
![$[BRM]=[MRC]$](http://latex.artofproblemsolving.com/8/7/6/8768ad740dc23cd630672c116d9c23dc98e49e39.png)
![$[BRS]=[CRS]$](http://latex.artofproblemsolving.com/d/9/b/d9b3bfa24e20aa4e37b82e4d7c64ffdfea7d9ecd.png)

Comment
I like the above solution because it involves less angle chasing than the others that I've seen.
I've also decided to go back and look at the British Olympiad Round 1 from 2008 that I printed out a while ago. I was looking through some stuff when I found it, and I wanted to see how much better I've gotten at math since I first found it... maybe two years ago?
Problem 2 wrote:
Find all real values of
,
, and
such that
![\[(x+1)yz=12,\,\,\,(y+1)zx=4,\,\,\,\text{and}\,\,\, (z+1)xy=4.\]](//latex.artofproblemsolving.com/1/0/e/10ee77616816a04e776dc3adbdfc13bf5c941b2c.png)



![\[(x+1)yz=12,\,\,\,(y+1)zx=4,\,\,\,\text{and}\,\,\, (z+1)xy=4.\]](http://latex.artofproblemsolving.com/1/0/e/10ee77616816a04e776dc3adbdfc13bf5c941b2c.png)
Couldn't solve this one before so yay
Multiplying the first equation by
, the second by
, and the third by
gives
Let all three expressions equal
. Then we can solve to get
which immediately implies that
. Now substituting this into our first two equations (the third will now be equivalent to the second) gives
and
. Dividing the first equation by the second gives
Now substituting this expression for
into the first equation and simplifying gives
Therefore
or
. Substituting each of these in gives
and
respectively, so we have the solutions
. Both check.



![\[\left.\begin{array}{cc} (x+1)xyz&=12x\\(y+1)xyz&=4y\\(z+1)xyz&=4z\end{array}\right\}\implies \dfrac{3x}{x+1}=\dfrac{y}{y+1}=\dfrac{z}{z+1}.\]](http://latex.artofproblemsolving.com/5/9/a/59ab9539a0b3f02d5387f92570c454387e4c4f0a.png)





![\[\dfrac{(x+1)y}{x(y+1)}=3\implies x=\dfrac y{2y+3}.\]](http://latex.artofproblemsolving.com/f/a/9/fa92efde888f82789c98e3f6a3a249676a103b0a.png)

![\[y^2\left(\dfrac y{2y+3}+1\right)=12\implies (y-3)(y+2)^2=0.\]](http://latex.artofproblemsolving.com/2/d/b/2dbe80576567b025948b1b596fb71e6778bccede.png)





Problem 3 wrote:
Let
be a parallelogram such that
is an acute-angled triangle. The circumcircle of triangle
meets the line
again at
. Prove that
if, and only if,
. The circumcircle of a triangle is the circle which passes through its vertices.







Wait why is this higher than the previous one



If: We have




Only If: Since





We have proven the statement in both directions, so we are done.

Problem 6 wrote:
The obtuse-angled triangle
has sides of length
and
opposite the angles
respectively. Prove that
![\[a^3\cos A+b^3\cos B+c^3\cos C<abc.\]](//latex.artofproblemsolving.com/f/e/e/feef5a9a5fbf120385956cfc6de7813ce174a4dd.png)




![\[a^3\cos A+b^3\cos B+c^3\cos C<abc.\]](http://latex.artofproblemsolving.com/f/e/e/feef5a9a5fbf120385956cfc6de7813ce174a4dd.png)
Bashy but interesting
In the following solution, all summations are considered to be symmetric.
By the Law of Cosines, we have
. We can write similar expressions for the other two cosines, so substituting these in gives
Let
. The inequality now reduces to
If we can prove this, then we are done.
Now we focus on the interesting obtuse condition. WLOG let
be the longest side. Then the condition translates to
. Rearranging gives
, and cubing gives
Next, I claim that
. Indeed, note that by moving everything to one side we get
Finally, adding
to the nasty inequality just above gives
Dividing both sides by three gives us the desired inequality, and we are done. 
By the Law of Cosines, we have

![\[\sum a^3\left(\dfrac{b^2+c^2-a^2}{2bc}\right)<abc\implies \sum a^4(b^2+c^2-a^2)<2a^2b^2c^2.\]](http://latex.artofproblemsolving.com/1/d/d/1dd59cec067f1e79179da16ed57ca25da2f2f032.png)

![\[\sum x^2(y+z-x)<2xyz\implies \sum x^2y-\sum x^3<2xyz.\]](http://latex.artofproblemsolving.com/5/4/d/54dbd34cce917d68aa51e148c03e27421bccdbfe.png)
Now we focus on the interesting obtuse condition. WLOG let





![\begin{align*}z^3-3z(x^2+y^2)+2x^3+2y^2&=z[z^2-3(x^2+y^2)]+2x^3+2y^2\\&>(x+y)[(x+y)^2-3(x^2+y^2)]+2x^3+2y^3\\&=(x+y)[-2x^2+2xy-2y^2]+2x^3+2y^3=0.\end{align*}](http://latex.artofproblemsolving.com/a/7/d/a7d83d03b9bec3e27b26022d03cd1f50cd91678d.png)



This post has been edited 4 times. Last edited by djmathman, Apr 29, 2016, 4:06 AM