Power of A Point: Albert Zhu

Power of a Point
Albert Zhu
April 1st, 2021
This article is about the various uses of Power of a Point within olympiad
geometry. Sections 1 through 4 should be fairly accessible to beginners
who have had some practice with angle chasing, although Section 5 requires
some extra background on barycentric coordinates. If you don’t know what
something means, consult the first two Appendices or message me on AoPS;
my username is Kagebaka.
Acknowledgements
Special thanks to Evan Chen for providing the style file for this article, as well as AoPS
user amar 04 for giving many problem suggestions and proofreading. Much of the material
of this handout was inspired by Chapter 2 of Evan Chen’s book Euclidean Geometry in
Mathematical Olympiads, as well as Kenneth Peng’s linearity handout.
Contents
1 Introduction 3
2 Radical Axis 7
3 Coaxial Circles 12
4 Point Circles 17
5 Black Magic 19
5.1 Barycentric Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.2 Linearity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6 Final Remarks 26
6.1 Heuristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
6.2 Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
7 Problems 27
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
7.2 Radical Axis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
7.3 Coaxial Circles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7.4 Point Circles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7.5 Black Magic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
7.6 Challenge Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1
Albert Zhu (April 1st, 2021) Power of a Point
8 Hints 32
9 Appendices 34
9.1 Appendix A: Notations and conventions . . . . . . . . . . . . . . . . . . . . 34
9.2 Appendix B: Referenced theorems without proof . . . . . . . . . . . . . . . 34
9.3 Appendix C: Trigonometric bash for TSTST 2016/6 . . . . . . . . . . . . . 35
2
§1 Introduction
First, we begin with the basic theorem, which allows us to derive cyclic quadrilaterals
based on lengths:
Theorem 1.1 (Power of a Point)

Let A, B, C, D be points in the plane with P = AB ∩ CD such that P either lies
on both segments AB, CD or neither. Then A, B, C, D are concyclic if and only if
P A · P B = P C · P D.
A A′
C C′
P
D B′
B D′
′
P
Proof. Observe that
PA · PB = PC · PD
−
⇐⇒ 4P AC ∼ 4P DB
⇐⇒ ∠P AC = ∠P DB
⇐⇒ A, B, C, D concyclic,
so we’re done.
Exercise 1.2. Prove the tangent case; i.e. if P, A, B are on a line in that order, then if
C is one of the points on a circle passing through A, B such that P C is tangent to that
circle, we must have P A · P B = P C 2 .
One observation we can make about this configuration is that the quantity P A · P B
did not depend on the line AB, and just the position of point P with respect to the circle.
This motivates the following definition:
Definition 1.3. The power of point P with respect to a circle ω with center O and
radius R is
Pow(P, ω) = OP 2 − R2 .
Note that this quantity actually has a sign (positive if P is outside ω, 0 if P is on ω,

negative if P is inside ω), which is an important fact that you must remember!
Exercise 1.4. Check that the absolute value of the power of a point P returns the same
product of lengths from the first theorem and the tangent case.
Actually, before we develop any further theory, we can already now solve some problems
with just the basic PoP theorem:
3
Example 1.5 (Euler’s Theorem)

Let 4ABC be a triangle. If O and I are its circumcenter and incenter, respectively,
and R and r are its circumradius and inradius, respectively, then prove that OI 2 =
R(R − 2r), and in particular R ≥ 2r.
M2
A
F O
I
B C
M1
Proof. It suffices to show that R2 − OI 2 = 2Rr, which is clearly the power of point I with
respect to the circumcircle of 4ABC. In particular, if we let AI meet the circumcircle
again at M1 , the midpoint of the arc AB not containing A, then we just need to show
that AI · IM1 = 2Rr. This looks sort of like similar triangles (remember this idea: similar
triangles often come in handy when using PoP), so we are motivated to introduce the
F, the point where the incircle touches AB, and the midpoint M2 of arc BAC. Since
∠BM2 M1 = ∠BAM1 and ∠M2 BM1 = 90◦ = ∠AF I, we must have 4AF I ∼ 4M2 BM1 ,
which gives us 2R AI
= BMr , but we know that BM1 = M1 I by the Incenter-Excenter
1
Lemma, so we get the desired result upon cross multiplying. Of course, R ≥ 2r just
follows from the fact that OI 2 and R are both positive, so R − 2r must be as well unless
AB = BC = CA =⇒ O = I, which gives the equality case.
Exercise 1.6. Let H be the orthocenter of 4ABC. Prove that

Pow(H, (ABC)) = 8R2 cos A cos B cos C.
Example 1.7 (Nine-Point Circle)

In a triangle 4ABC, let
• H be the orthocenter,
• HA , HB , HC be the feet of the altitudes from A, B, C respectively,
• MA , MB , MC be the midpoints of BC, CA, AC, respectively, and
• XA , XB , XC be the midpoints of AH, BH, CH, respectively.
Then HA , HB , HC , MA , MB , MC , XA , XB , XC lie on a circle whose center is the

midpoint of H and the circumcenter O (and in particular lies on the Euler Line).
4
XA HB
MC MB
O
N
HC H
XB XC
B C
HA MA
Proof. The canonical proof of this theorem is with homothety, but we can still use Power
of a Point in a very nice way to deal with most of the points. WLOG let 4ABC be
acute with ∠B > ∠C, and let HA 0 , H 0 , H 0 be the reflections of H over BC, CA, AB,
B C
respectively, which are well-known to lie on (ABC). Then we have
0 0
−Pow(H, (ABC)) = AH · HHA = BH · HHB = CH · HHC0
1 1 0 1 1 0 1 1
=⇒ AH · HHA = BH · HHB = CH · HHC0
2 2 2 2 2 2
=⇒ XA H · HHA = XB H · HHB = XC H · HHC ,
so the converse of PoP implies that HA , HB , HC , XA , XB , XC all lie on a circle ω.

Now, to prove that the midpoints also lie on ω, we observe that since XA is the center
of (AHB HHC ), it lies on the perpendicular bisector of HB HC and we have ∠HC XA HB =
2∠A. On the other hand, we also know that MA is the center of (BHC HB C), so MA must
also lie on the perpendicular bisector of HB HC and we have ∠HC MA HA = 2∠HC CHB =
180◦ − 2∠A. Thus, MA lies on ω, and furthermore XA MA is a diameter.
Finally, we want to show that the center N of ω is the midpoint of OH, which we can
prove with some angle chasing. Observe that
∠HA XA MA = 90◦ − ∠XA MA HA

= 90◦ − ∠XA HB HA
= 90◦ − ∠XA HB HC − ∠HC HB HA
1
= 90◦ − (180◦ − 2∠A) − 2(90◦ − ∠B)
2
= ∠A − ∠BAHA − ∠OAC
= ∠HAO,
so AO k XA MA and in fact AXA MA O is a parallelogram as we also have AXA ⊥

BC ⊥ OMA . Since the opposite sides of a parallelogram are equal, we then must have
HXA = AXA = OMA , so 4HXA N ∼ = 4OMA N by SAS congruence, and N is the
midpoint of OH as desired.
5
Example 1.8 (IMO Shortlist 2013/G4)

Let ABC be a triangle with ∠B > ∠C. Let P and Q be two different points on
line AC such that ∠P BA = ∠QBA = ∠ACB and A is located between P and C.
Suppose that there exists an interior point D of segment BQ for which P D = P B.
Let the ray AD intersect the circle ABC at R 6= A. Prove that QB = QR.
A D R
B
C
Proof. The first thing we notice is that the angle conditions produce a number of
tangencies: specifically, we have P B tangent to (ABC) and AB tangent to both (BQC)
and (BDR). Now we would like to translate these tangencies into usable information, and
PoP is the key: the former tangency implies that P A · P C = P B 2 = P D2 and the latter
pair of tangencies implies that AD · AR = AB 2 = AQ · AC, so in fact P D is tangent to
(ADC) and additionally DQCR is cyclic, both by the converse of PoP.
Now the rest is just length chasing. By the Angle Bisector Theorem on 4P BQ,
·AQ
QB = BPAP , but PoP on A with respect to (DQCR) and P with respect to (ABC)
2
respectively tell us that AQ = AD·ARAC and AP = BPCP , so substituting gives us QB =
AR CP QR AR
AC · AD · BP . However, we also know that 4ADC ∼ 4AQR =⇒ CD = AC and
AD BP
4ADP ∼ 4CDP =⇒ CD = CP , so simplifying gives us
QR CP
QB = · AD · = QR
CD BP
as desired.
6
§2 Radical Axis
Now we look at what happens when the power of a point is equal with respect to two
circles.
Definition 2.1. If ω1 and ω2 are two circles, then the set of points P such that
Pow(P, ω1 ) = Pow(P, ω2 ) is the radical axis of the two circles.
The reason why this locus is special is because of the following characterization:
Theorem 2.2 (Radical Axis)

The radical axis of two circles ω1 and ω2 with centers O1 and O2 , respectively, is a
line perpendicular to O1 O2 . In particular, the radical axis is the common chord if
the two circles intersect and the common tangent if the two circles are tangent.
Proof. We use Cartesian coordinates. Let r1 and r2 be the radii of ω1 and ω2 , respectively,
and suppose that O1 = (a, 0), O2 = (b, 0). Then for any point P = (x, y), we have
Pow(P, ω1 ) = (x − a)2 + y 2 − r12

Pow(P, ω2 ) = (x − b)2 + y 2 − r22 ,
so
0 = Pow(P, ω1 ) − Pow(P, ω2 )
= ((x − a)2 + y 2 − r12 ) − ((x − b)2 + y 2 − r22 )
= (−2a + 2b)x + (a2 − b2 + r22 − r12 ),
which is the equation of a line perpendicular to the x-axis as desired. Also, if we

let A, B are the intersections of ω1 and ω2 , then Pow(A, ω1 ) = 0 = Pow(B, ω2 ) and
Pow(A, ω1 ) = 0 = Pow(B, ω2 ), so AB must be the radical axis, and if ω1 and ω2 are
tangent, then clearly their common tangent is perpendicular to O1 O2 and their touch
point has an equal power of 0 to both circles.
7
Exercise 2.3. Prove this theorem synthetically with the lemma that AB is perpendicular
to CD if and only if AC 2 + BD2 = AD2 + BC 2 .
Remark 2.4. Observe that if we set r1 or r2 to 0, none of the algebra breaks. We will
discuss the use of radical axes with degenerate circles in the fourth section.
Remark 2.5. Furthermore, if we substitute the 0 with y, then Pow(P, ω1 ) − Pow(P, ω2 ) is

still a linear function in P. We will revisit this idea in the last section.
A natural extension of the radical axis is the radical center, which is the scenario
where the power of a point is equal with respect to three circles rather than two. As we
will see, this allows us to combine collinearity/concurrency with concyclicity.
Theorem 2.6 (Radical Center)

Let points A, B lie on ω1 and points C, D lie on ω2 . Then AB ∩ CD = P lies on
the radical axis of ω1 and ω2 if and only if ABCD is a cyclic quadrilateral whose
circumcenter does not lie on the line connecting the centers of ω1 and ω2 .
C A
P
D
Proof. For the “if” direction, observe that AB and CD are the radical axes of (ABCD), ω1
and (ABCD), ω2 , respectively. Thus,
Pow(P, ω1 ) = Pow(P, (ABCD)) = Pow(P, ω2 ),
which means that P must lie on the radical axis of ω1 and ω2 as desired.
For the “only if” direction, we know that
±P A · P B = Pow(P, ω1 ) = Pow(P, ω2 ) = ±P C · P D.
Since P A · P B is negative if and only if P is between A and B and positive if and

only if P is outside of segment AB, the position of P allows the converse of PoP to be
satisfied.
8
Remark 2.7. For the case where ABCD’s circumcenter does lie on the line connecting the
centers of ω1 and ω2 , the radical axes are either all the same line or pairwise parallel (as they
are all perpendicular to the line connecting all three centers), so they intersect at infinity.
This is an important potential configuration issue that should always be remembered when
using the radical center!
Now let’s do some examples.
Example 2.8
Let two circles ω1 and ω2 meet at A, B, and let C, D be points on ω1 , ω2 , respectively,
such that CD is a common tangent. Then if M is the midpoint of CD, we have
Pow(M, ω1 ) = CM 2 = DM 2 = Pow(M, ω2 ) = CM · DM = Pow(M, (ACD)),
so why isn’t M the radical center of ω1 , ω2 , and (ACD)?
Proof. The problem is that there are sign issues. The powers of M with respect to ω1
and ω2 are positive (note that M still lies on their radical axis), but the power of M
with respect to (ACD) is negative.
Remark 2.9. If two circles are non-intersecting, the midpoints of the internal tangents also
lie on the radical axis. This is an important property of the radical axis, so don’t forget it!
Example 2.10 (USAJMO 2012/1)

Given a triangle ABC, let P and Q be points on segments AB and AC, respectively,
such that AP = AQ. Let S and R be distinct points on segment BC such that S
lies between B and R, ∠BP S = ∠P RS, and ∠CQR = ∠QSR. Prove that P , Q, R
and S are concyclic
B C
S R
Proof. First, we observe that AB and AC are tangent to (P SR) and (QSR), respectively,
and we want to show that these two circles must actually be the same. Well, if (P SR) 6=
(QSR), then they must have radical axis SR. Thus, if we were to show that a point
not lying on SR must be on the radical axis of the two circles, we would be done by
contradiction, and in fact A does the trick:
Pow(A, (P SR)) = AP 2 = AQ2 = Pow(A, (QSR)),
so P, Q, S, R are concyclic as desired.
9
Exercise 2.11. Try this problem with phantom point instead of radical axis.
Example 2.12 (USAMO 2009/1)

Given circles ω1 and ω2 intersecting at points X and Y , let `1 be a line through the
center of ω1 intersecting ω2 at points P and Q and let `2 be a line through the center
of ω2 intersecting ω1 at points R and S. Prove that if P, Q, R and S lie on a circle
then the center of this circle lies on line XY .
S O3
X
Q
P R
O1 O2
Proof. Let O1 , O2 , O3 be the centers of ω1 , ω2 , (P RQS), respectively. Drawing the dia-

gram, we can see that by radical center, P Q and RS meet on XY, so we would like to
somehow tie this information to the other condition that the radical axes pass through
the centers of ω1 and ω2 . Well, we know that O1 Q ⊥ O2 O3 and O2 S ⊥ O1 O3 , so in fact
this implies that the radical center of the three circles is actually the orthocenter of
4O1 O2 O3 . We also know that O2 O3 ⊥ XY, so O3 , X, Y must be collinear as the three
points form the altitude from O3 to O1 O2 . However, we must remember to check whether
or not O1 , O2 , O3 can be collinear! In fact, as seen in the diagram below, this is actually
a possible configuration:
P
X
R
O O1 O2
S
Y
Q
10
Thankfully, this special case can just be addressed with a fairly simple phantom point
and Pythagorean Theorem argument: if we let O be the midpoint of XY, then
OP 2 = OO12 + O1 P 2
= OO12 + (O2 P 2 − O1 O22 )
= OO12 + O2 X 2 − O1 O22
= OO12 + (OO22 + OX 2 ) − O1 O22
2
2 2 1
= OO1 + OO2 + XY − O1 O22 ,
2
which is symmetrically equal for OQ2 , OR2 , OS 2 as well, so O = O3 .
Exercise 2.13. Prove that O3 lies on the radical axis of ω1 , ω2 with length chasing
rather than radical center. Note that this solution does not require the all parallel case
to be addressed separately!
11
§3 Coaxial Circles
Definition 3.1. If 3 or more circles share pairwise radical axes, then we say that they
are a coaxial pencil or just coaxial.
Note that three circles that share a point are coaxial if and only if their centers are
collinear; thus, coaxiality can come in handy if we wish to prove that three circles’ centers
are collinear. However, the main result of this section is the following relatively obscure
lemma (though it has recently become more popular), which is a very powerful tool for
proving concyclicity as it allows us to essentially ignore two of the points on the circle.
Lemma 3.2 (Forgotten Coaxiality Lemma)

Let two circles ω1 and ω2 intersect at points A and B. Then the locus of points P
such that Pow(P,ω1)
Pow(P,ω2 ) = k for some constant k is a circle that forms a coaxial pencil
with ω1 and ω2 .
D2
D
A
D1
C1
B C C2
Proof. WLOG let k be negative (i.e. the circle is contained within the union of the
areas of ω1 and ω2 ). Let C and D be points such that k = Pow(C,ω 1) Pow(D,ω1 )
Pow(C,ω2 ) = Pow(D,ω2 ) . If AD
meets ω1 , ω2 at D1 , D2 , respectively, and BC meets ω1 , ω2 at C1 , C2 , respectively, then
it follows that
CC1 · CB DD1 · DA CC1 DD1
k=− =− =⇒ = ;
CC2 · CB DD2 · DA CC2 DD2
along with C1 D1 k C2 D2 from Reim’s Theorem on ABC1 D1 , this gives us CD k C1 D1 k
C2 D2 , which means that ABCD is cyclic by the converse of Reim. These steps can
be reversed to prove the converse; assuming ABCD is cyclic, Reim’s implies that
CD k C1 D1 k C2 D2 , which means that
CC1 DD1 CC1 · CB DD1 · DA
= =⇒ =
CC2 DD2 CC2 · CB DD2 · DA
as desired.
First, we present two classical results.
Example 3.3 (Newton-Gauss Line)

In a convex quadrilateral ABCD, let P = AB ∩ CD and Q = AD ∩ BC. Then the
midpoints of P Q, AC, and BD are collinear.
12
D
H1
A
H4
H2
Q C
B
H3
Proof. We will show that the circles (P Q), (AC), and (BD) are coaxial, which suffices.
Let H1 , H2 , H3 , and H4 be the orthocenters of 4P AD, 4P BC, 4QAB, and 4QCD,
respectively. Then since H1 is the radical center of the triples {(P Q), (BD), (P D)},
{(P Q), (AC), (AP )}, and {(BD), (AC), (AD)}, it lies on the pairwise radical axes of
(P Q), (AC), and (BD). Similarly, H2 , H3 , and H4 also lie on the pairwise radical axes,
which means that the three circles must actually be coaxial as they would otherwise have
multiple radical centers, an impossibility.
Example 3.4 (Existence of Isodynamic Points)

In a triangle 4ABC, the locus of points P such that PP B AB
C = AC is known as the
A-Apollonian Circle. Prove that the three Apollonian Circles of a non-equilateral
triangle are coaxial.
M
A
E L B D C
Proof. First we show that the locus is actually a circle. To do so, let D and E be the
intersections of the A-internal and A-external bisectors of 4ABC with BC, respectively;
observe that D and E must lie on the A-Apollonian Circle because of the Angle Bisector
Theorem on 4ABC, and we also must have ∠DAE = 90◦ . This suggests that the locus
is the circle with diameter DE; indeed, for any other point P satisfying PP B AB
C = AC , D
and E are also the intersections of the P -internal and P -external bisectors of 4P BC
with BC, so again we must have ∠DP E = 90◦ as desired.
13
Next we show that the three Apollonian Circles are coaxial. Let J be an intersection
of the A- and B-Apollonian Circles of 4ABC. Then we must have
JB AB JA BA JB AC
= , = =⇒ = ,
JC AC JC BC JC BC
so J lies on the C-Apollonian Circle as well. Let L, M, N be the centers of the A-, B-,
C-Apollonian Circles, respectively; it now suffices to show that L, M, N are collinear.
Since L is the midpoint of DE, we have
1 1
∠LAB + ∠A = ∠ADL = ∠A + ∠C =⇒ ∠LAB = ∠C,
2 2
so LA is tangent to (ABC), and by symmetry M B and N C are as well, which means
that
BC 2 LA2 LB · LC LB
4LAB ∼ 4LCA =⇒ = = = .
CA2 LC 2 LC 2 LC
Thus,
LB M C N A BC 2 CA2 AB 2
· · = · · = 1,
LC M A N B CA2 AB 2 BC 2
so we’re done by the converse of Menelaus’s Theorem on 4ABC with traversal LM N .
Now we obliterate some problems with the Forgotten Coaxiality Lemma.
Example 3.5 (CGMO 2017/7)

Let the ABCD be a cyclic quadrilateral with circumcircle ω1 . Lines AC and BD
intersect at point E, and lines AD, BC intersect at point F . Circle ω2 is tangent to
segments EB, EC at points M, N respectively, and intersects with circle ω1 at points
Q, R. Lines BC, AD intersect line M N at S, T respectively. Show that Q, R, S, T
are concyclic.
A
D
E F
M N T
C S
B R
Proof. We’re motivated to use the Forgotten Coaxiality Lemma because there are lots
of similar triangles, so we know a lot of lengths, and there’s not really any obvious
way to characterize Q and R. Through some easy angle chasing, we can observe that
4SCN ∼ 4T DM =⇒ TSN SC SM SB
M = T D and 4SBM ∼ 4T AN =⇒ T N = T A , so it follows
that
SC · SB TD · TA Pow(S, ω1 ) Pow(T, ω1 )
= ⇐⇒ = ,
SN · SM TN · TM Pow(S, ω2 ) Pow(T, ω2 )
and we’re done.
14

Let ABC be a triangle with circumcircle ω and ` a line without common points
with ω. Denote by P the foot of the perpendicular from the center of ω to `. The
side-lines BC, CA, AB intersect ` at the points X, Y, Z different from P . Prove
that the circumcircles of the triangles AXP , BY P and CZP have a common point
different from P or are mutually tangent at P .
Y D
Z
E P
A
O C
Proof. WLOG suppose the diagram is as shown, and let D = AB ∩ (BY P ) 6= B and
E = AC ∩ (CZP ) 6= C. By the Forgotten Coaxiality Lemma, we just need to show that
Pow(A, (BY P )) Pow(X, (BY P )) AD · AB XP · XY
= ⇐⇒ = .
Pow(A, (CZP )) Pow(X, (CZP )) AC · AE XP · XZ
By Menelaus’s Theorem on 4AY Z and traversal BCX, we find that the RHS is equal
to AB·CY AD CY
BZ·AC , so it suffices to prove that AE = BZ . Now this is just length chasing: we have
AD · BZ − AE · CY + ZP · ZY − Y P · ZY
= AD · BZ − AE · CY + DZ · BZ − EY · CY
= AZ · BZ − AY · CY
= Pow(Z, (ABC)) − Pow(Y, (ABC))
= OZ 2 − OY 2
= (OP 2 + ZP 2 ) − (OP 2 + Y P 2 )
= ZY (ZP − Y P ),
so we’re done upon rearranging.
For our final example, we will use the Forgotten Coaxiality Lemma on a circle with
radius 0(!) to prove that two circles are tangent.
Example 3.7 (IMO 2015/3)

Let ABC be an acute triangle with AB > AC. Let Γ be its circumcircle, H its
orthocenter, and F the foot of the altitude from A. Let M be the midpoint of BC.
Let Q be the point on Γ such that ∠HQA = 90◦ and let K be the point on Γ such
that ∠HKQ = 90◦ . Assume that the points A, B, C, K and Q are all different and
lie on Γ in this order. Prove that the circumcircles of triangles KQH and F KM are
tangent to each other.
15
A
Q
K
L
B M F C
Q′
A′ D
Proof. Let L = Q∞BC ∩AH. The problem is equivalent to showing that (KQH), (F KM ),
and the circle centered at K with radius 0 are coaxial, so by the Forgotten Coaxiality
Lemma we need to prove that
Pow(M, K) Pow(F, K) M K2 MH · MQ
= ⇐⇒ 2
= .
Pow(M, (KQH)) Pow(F, (KQH)) FK FH · FL
+ QH MQ M H+QH
However, 4HM F ∼ 4HQL implies that M H MH
F H = LH , so F L = F H+LH = F H , which
means that we just want to show that M K MH
F K = F H ; this is equivalent to showing that
K lies on ω, the H-Apollonian Circle of 4HM F. If P is the center of ω and D is the
reflection of H over F, then it’s clear that D must lie on the ω as DF = HF and
DM = HM, and we also know that P H is tangent to (HM F ), so HM is tangent to ω.
Thus, it suffices to show that HM is tangent to (DHK).
Let A0 be the reflection of H over M, `0 be the tangent to (DKH) at H, and KH ∩ Γ 6=
K = Q0 ; clearly Q, H, M, A0 are collinear because ∠AQH = 90◦ = ∠AQA0 , and it’s
well-known that A0 is the A-antipode in Γ. Additionally, Q0 must be the antipode of Q in
(ABC) because ∠QKQ0 = ∠QKH = 90◦ , implying that AQA0 Q0 is a rectangle, which
combined with `0 k AQ0 from Reim’s Theorem on degenerate quadrilateral HHKD gives
HM = `0 as desired.
16
§4 Point Circles
We now take a brief moment to examine the uses of the radical axis when we consider
points to be circles with radius 0. Although this technique is fairly niche and extremely
difficult to spot in the wild, it often leads to very elegant solutions when used effectively.
Example 4.1 (Circumcenter)

In 4ABC, prove that the perpendicular bisectors of its sides meet at a single point.
Proof. If we consider A and B to be circles of radius 0, then it follows that their radical
axis is the perpendicular bisector of AB. Symmetrically, it follows that the radical center
of A, B, C is the concurrent point of the three perpendicular bisectors.
Example 4.2 (Polish MO 2018/5)

An acute triangle ABC in which AB < AC is given. Points E and F are feet of
its heights from B and C, respectively. The line tangent in point A to the circle
escribed on ABC crosses BC at P . The line parallel to BC that goes through point
A crosses EF at Q. Prove P Q is perpendicular to the median from A of triangle
ABC.
A Q
F
C
P B M
Proof. We observe that if we were to show that P Q was the radical axis of (BC) and
some circle centered at A, we would be done, which motivates our next step: since EF is
the radical axis of (BC) and (AEF ), it follows that Q is the radical center of the point
circle at A, so Q lies on the radical axis of A and (BC), and obviously we have
Pow(P, A) = P A2 = Pow(P, (ABC)) = P B · P C = Pow(P, (BC))
so P lies on the radical axis of A and (BC) as well.
Example 4.3 (CGMO 2015/6)

Let Γ1 and Γ2 be two non-overlapping circles. A, C are on Γ1 and B, D are on
Γ2 such that AB is an external common tangent to the two circles, and CD is an
internal common tangent to the two circles. AC and BD meet at E. F is a point
on Γ1 , the tangent line to Γ1 at F meets the perpendicular bisector of EF at M .
M G is a line tangent to Γ2 at G. Prove that M F = M G.
17
B
X
A
O1 G E O2
D
F
M
Proof. We observe that there are a lot of equal lengths in this problem, so this motivates
us to try to use radical axes somehow. Since we know that the midpoints of AB and CD
form the radical axis of Γ1 and Γ2 (see Example 2.8 if you’re not sure why), we let these
points be X and Y, respectively. We also know that AC ⊥ BD through some simple
angle chasing, so we now introduce the key idea of using the point circle E :
Pow(X, E) = XE 2 = XA2 = Pow(X, A) = XB 2 = Pow(X, B)
Pow(Y, E) = Y E 2 = Y A2 = Pow(Y, A) = Y B 2 = Pow(Y, B),
implying that Γ1 , Γ2 , and E are coaxial. Thus, Pow(M, Γ1 ) = M F 2 = M E 2 = Pow(M, E)
implies that M ∈ XY as well, which means that M F 2 = M G2 =⇒ M F = M G.
This next solution is in my opinion less intuitive than the previous ones, but it’s too
good not to share.
Example 4.4 (Balkan MO 2015/2)

Let 4ABC be a scalene triangle with incentre I and circumcircle ω. Lines AI, BI, CI
intersect ω for the second time at points D, E, F , respectively. The parallel lines from
I to the sides BC, AC, AB intersect EF, DF, DE at points K, L, M , respectively.
Prove that the points K, L, M are collinear.
A
E
C
K B
D
M L
−
Proof. Since ∠F IK = ∠F CB = ∠KEI, we have 4KF I ∼ 4KIE. This means that
KF · KE = KI 2 , so K lies on the radical axis of I and ω, and by symmetry M and L
do too, which means that K, M, L are collinear as desired.
18
§5 Black Magic
The main idea of this section is to combine Power of a Point with more analytical
methods, which tends to be more versatile than just using PoP by itself albeit at the
cost of generally being more computationally intensive.
§5.1 Barycentric Coordinates

The use of barycentric coordinates concurrently with Power of a Point is actually fairly
natural due to the following fact:
Theorem 5.1 (Barycentric Power of a Point)

Let some circle ω have equation
−a2 yz − b2 zx − c2 xy + (x + y + z)(ux + vy + wz) = 0.
Then if P = (x, y, z) is any point in the plane,
Pow(P, ω) = −a2 yz − b2 zx − c2 xy + (x + y + z)(ux + vy + wz).
Proof. Since the formula of a circle is derived from

(distance from (x, y, z) to the center)2 − radius2 = 0,
plugging in an arbitrary point yields the power of that point. Note that (x, y, z) cannot
be homogenized, because otherwise the distance formula breaks!
As a demonstration of how useful this formula can be even by itself, we present an

extremely clean solution to a problem from the AoPS Quarantine Geometry Olympiad.
Example 5.2 (AQGO 2020/2)

Let 4ABC be a triangle with D as a random point on BC such that BD < BM
where M is the midpoint of BC and let E, F be on sides AB, AC such that EF k BC.
Let D0 be the reflection of D across the midpoint of BC and define X, Y as the
intersections of (BD0 E) and (CDF ). Then prove that (AXY ) passes through
a fixed point as D varies other than A.
X
E F
B D Y D′ C
19
Proof. Upon inspection, the fixed point appears to be the intersection of the A-symmedian
with (ABC); call this point G. Since D is a variable point, it’s hard to get any information
about it, so we are motivated to use the Forgotten Coaxiality Lemma in order to avoid
having to deal with it. We use barycentric coordinates with reference triangle 4ABC.
Let E = (e, 1 − e, 0) and F = (e, 0, 1 − e) (this comes from [EBC] [F BC]
[ABC] = [ABC] ) be fixed
points, and let D = (0, d, 1 − d) and D0 = (0, 1 − d, d) be variable points. Plugging in
B, E, D0 and C, F, D implies that (BED0 ) and (CF D) are parametrized by
a2 yz + b2 zx + c2 xy = (x + y + z)(c2 (1 − e)x + a2 (1 − d)z)
and
a2 yz + b2 zx + c2 xy = (x + y + z)(b2 (1 − e)x + a2 (1 − d)y),
respectively. Since G = (−a2 : 2b2 : 2c2 ), it follows that
Pow(BED0 ) (G)
Pow(CF D) (G)
−4a2 b2 c2 + 2a2 b2 c2 + 2a2 b2 c2 + (2b2 + 2c2 − a2 )(−a2 c2 (1 − e) + 2a2 c2 (1 − d))
=
−4a2 b2 c2 + 2a2 b2 c2 + 2a2 b2 c2 + (2b2 + 2c2 − a2 )(−a2 b2 (1 − e) + 2a2 b2 (1 − d))
c2
= 2
b
AE · AB
=
AF · AC
Pow(BED0 ) (A)
= ,
Pow(CF D) (A)
so A, G, X, Y are concyclic, which is what we wanted to show. Note that we were able to
use homogenized coordinates here because the common denominator of unhomogenized
coordinates would cancel out in the fraction. This will not always be the case!
Our next example is an occasionally useful result that allows us to compute arbitrary
powers of points with respect to a circle without having to involve its center.
Example 5.3
Let ABC be a triangle and P be an arbitary point. Let the parallel line to BC
through P meet AB, AC at A1 , A2 . Similarly define B1 ∈ BC, B2 ∈ BA and
C1 ∈ CA, C2 ∈ CB. Then using directed lengths, i.e. P A1 · P A2 is given a positive
sign if P~A1 is in the same direction of P~A2 and negative otherwise, we have
Pow(P, (ABC)) = P A1 · P A2 + P B1 · P B2 + P C1 · P C2 .
Proof. We use barycentric coordinates with reference triangle 4ABC. Suppose that
P = (x, y, z). Observing that [P BC] = [A1 BC] = [A2 BC] since the three triangles share
the same base and height, we must have A1 = (x, 1 − x, 0), A2 = (x, 0, 1 − x), so the
barycentric distance formula implies that
p
P A1 = −a2 (x + y − 1)z = az, P A2 = ay,
and similar computations yield
P B1 = bx, P B2 = bz, P C1 = cy, P C2 = cx.
20
Note that if the directed length product P A1 · P A2 is positive, then P is outside of

4ABC and one of y, z must be negative, and if P A1 · P A2 is negative, then P is inside
4ABC and both y and z must be positive. Thus, we must have P A1 · P A2 = −a2 yz in
order for both sides to have the same sign, so we get
P A1 · P A2 + P B1 · P B2 + P C1 · P C2
= −a2 yz + −b2 zx + −c2 xy
= Pow(P, (ABC))
as desired.
§5.2 Linearity
We begin with Remark 2.5 from the second section, but formalized:
Theorem 5.4 (Linearity of PoP)

Let ω1 and ω2 be two circles and P be the set of points in the plane. Then the
function f : P → R defined by f (•) = Pow(•, ω1 ) − Pow(•, ω2 ) is linear; in particular,
if P~ = tA
~ + (1 − t)B,
~ then f (P ) = tf (A) + (1 − t)f (B).
Proof. We already proved this earlier, but we now present an alternate, more synthetic
~ = kA
proof. Let A, B, C be points such that C ~ + (1 − k)B.
~ If O1 and O2 are the centers
of ω1 and ω2 , respectively, then Stewart’s Theorem on 4O1 AB, 4O2 AB respectively
imply that
k(1 − k) + CO12 = kAO12 + (1 − k)BO12 ,
k(1 − k) + CO22 = kAO22 + (1 − k)BO22 .
Subtracting the second equality from the first, we have
CO12 − CO22 = k(AO12 − AO22 ) + (1 − k)(BO12 − BO22 )

CO12 − r12 − CO22 + r22 = k(AO12 − r12 − AO22 + r22 ) + (1 − k)(BO12 − r12 − BO22 + r22 )
f (C) = kf (A) + (1 − k)f (B)
as desired.
The way that this theorem is mostly used in practice is the following corollary:
Corollary 5.5
Let ω1 , ω2 be two circles and let P be a point on line BC. Then if f (•) = f (•, ω1 ) −
f (•, ω2 ), we have
PC PB
f (P ) = f (B) + f (C).
BC BC
More generally, if P = (x, y, z) in barycentric coordinates with reference triangle
4ABC, we have
f (P ) = xf (A) + yf (B) + zf (C).
Now let’s do some examples, first with the line form of Corollary 5.5, and then with
the general form.
21
Example 5.6
In 4ABC, let M be the midpoint of BC, D be the projection of A onto BC, and O
be the circumcenter. OD meets AM at P . Prove that P lies on the radical axis of
(BOC) and the nine-point circle of triangle ABC.
H
P
B C
D M
Proof. WLOG let 4ABC be acute. Let ω be the nine-point circle, and define f (•) =
Pow(•, (BOC)) − Pow(•, ω). Here, we actually have two choices for f (P ), since
OP ~ DP ~ PM ~ AP ~
P~ = D+ O= A+ M.
OD OH AM AM
We will present the former solution path and leave the latter as an exercise. Let H be the
orthocenter of 4ABC, and let AD meet ω again at E (well-known to be the midpoint of
AH). Observe that it suffices to show that
OP DP OP f (O)
0 = f (P ) = f (D) + f (O) ⇐⇒ =− .
OD OH DP f (D)
It’s well-known that O and H are reflections over the center of ω, so we actually have
f (O) = −Pow(O, ω) = −Pow(H, ω) = EH · HD = (R cos ∠A)(2R cos ∠B cos ∠C),

f (D) = Pow(D, (BOC)) = −BD · DC = −(AB cos ∠B)(AC cos ∠C),
f (O) 2R2 cos ∠A
=⇒ − = .
f (D) AB · AC
We also know that OM k AD, so
OP OM R cos ∠A 2R2 cos ∠A

4OP M ∼ 4DP A =⇒ = = = ,
DP DA AB sin ∠B AB · AC
where the last equality comes from the Law of Sines on 4ABC, as desired.
AP f (A)
Exercise 5.7. Prove that MP = f (M ) .
22
Example 5.8 (USAMO 2015/2)

Quadrilateral AP BQ is inscribed in circle ω with ∠P = ∠Q = 90◦ and AP = AQ <
BP . Let X be a variable point on segment P Q. Line AX meets ω again at S (other
than A). Point T lies on arc AQB of ω such that XT is perpendicular to AX. Let
M denote the midpoint of chord ST . As X varies on segment P Q, show that M
moves along a circle.
U
X
A O B
M
T
Q
Proof. Let O be the center of ω. In locus problems, it’s always a good idea to check the
extreme cases, so setting X = P, X = P Q ∩ AB, and X = Q, we can see that M lies on
the circle passing through the midpoint of P B, the midpoint of QB, and Q. Thus, we
predict that the center of the locus circle of M is the midpoint of AO. This motivates us
to construct the circle with diameter AO as well as the midpoints U and V of AS and
AT, respectively, which clearly lie on (AO). Now let f (•) = Pow(•, (AO)) − Pow(•, ω);
then we have
1 1
Pow(M, (AO)) + M S · M T = f (M ) = (f (S) + f (T )) = (SU · SA + V T · AT ),
2 2
1 2 2 2
which implies that Pow(M, (AO)) = 4 (SA + AT − T S ). By Law of Cosines on 4AST,
this quantity equals
1 1
(SA · AT cos ∠SAT ) = (SA · AX),
2 2
but we know that 4AP X ∼ 4ASP =⇒ SA · AX = AP 2 , so Pow(M, (AO)) = 12 AP 2 .
In particular, this quantity does not depend on X, so if O1 is the midpoint of AO, it
follows that M O12 is fixed from the definition of power of a point, meaning that M lies
on a fixed circle with center O1 as desired.
Like the Balkan problem in the previous section, this next solution is definitely far less
natural than the previous ones, but it’s extremely slick.

Let ABC be a triangle with circumcenter O and incenter I. The points D, E and
F on the sides BC, CA and AB respectively are such that BD + BF = CA and
CD + CE = AB. The circumcircles of the triangles BF D and CDE intersect at
P 6= D. Prove that OP = OI.
23
F
I E
O
P
B C
D
Proof. First, it’s clear that AE + AF = BC, and we also observe that P lies on (AEF )
by Miquel’s Pivot Theorem. It suffices to show that Pow(P, (ABC)) = Pow(I, (ABC)),
which happens if and only if Pow(P, (ABC)) = −2Rr by Euler’s Theorem. Let P have
barycentric coordinates (x, y, z) with reference triangle 4ABC. Then if we define
fA (•) = Pow(•, (AEF )) − Pow(•, (ABC)),

fB (•) = Pow(•, (BF D)) − Pow(•, (ABC)),
fC (•) = Pow(•, (CDE)) − Pow(•, (ABC)),
we have
−Pow(P, (ABC)) = xfA (A) + yfA (B) + zfA (C) = y(c · BF ) + z(b · CE)
= xfB (A) + yfB (B) + zfB (C) = z(a · CD) + x(c · AF )
= xfC (A) + yfC (B) + zfC (C) = x(b · AE) + y(a · BD),
so
X X X X
− Pow(P, (ABC)) a= yca(BF ) + zba(CE) = xbc(AE + AF ) = abc x = abc.
cyc cyc cyc cyc
abc
Thus, Pow(P, (ABC)) = − a+b+c = − 4R[ABC]
2[ABC] = −2Rr as desired.
r
Now to finish, we will brutally murder a classic problem to showcase how powerful
linearity can be.
Example 5.10 (TSTST 2016/6)

Let ABC be a triangle with incenter I, and whose incircle is tangent to BC, CA,
AB at D, E, F , respectively. Let K be the foot of the altitude from D to EF .
Suppose that the circumcircle of 4AIB meets the incircle at two distinct points C1
and C2 , while the circumcircle of 4AIC meets the incircle at two distinct points B1
and B2 . Prove that the radical axis of the circumcircles of 4BB1 B2 and 4CC1 C2
passes through the midpoint M of DK.
24
C1
B2
E
D0
K
F
G I
F0
E0 M
B1
C2
B D C
Proof. Call the incircle ω. After some messing around, we let D0 , E0 , F0 be the mid-
points of EF, F D, DE, respectively: we can observe that D0 is the radical center of
(AIC), ω, (AI) and F0 is the radical center of (AIC), ω, (CI), so B1 B2 is the E-midline
in 4DEF (of course, C1 C2 is the F -midline as well by symmetry). As the Gergonne
Point G of 4ABC is the Symmedian Point of 4D0 E0 F0 , it’s well-known that D0 , M, G
are collinear (the so-called D-Schwatt Line of 4DEF ), so now we just need to show that
G lies on the radical axis of (BB1 BC ) and (CC1 C2 ).
To do so, we will use linearity: define fB (•) = Pow(•, ω) − Pow(•, (BB1 B2 )) and fC
similarly. Using 4BF0 D0 as a reference triangle, we have
[GF0 D0 ] [GBF0 ] [GD0 B] dist(G, F0 D0 )
fB (G) = fB (B) + fB (F0 ) + fB (E0 ) = BD2 .
[BF0 D0 ] [BF0 D0 ] [BF0 D0 ] dist(B, F0 D0 )
Thus, by symmetry it suffices to show that
fB (G) CD2 dist(G, D0 E0 ) dist(B, F0 D0 )
= 1 ⇐⇒ · · = 1.
fC (G) BD2 dist(G, F0 D0 ) dist(C, D0 E0 )
Noting that there are many right and equal angles from the tangents in the configuration,
we are motivated to use trigonometry; we complete the rest of this proof in Appendix
9.3, as it is not very enlightening and fairly long.
Exercise 5.11. For an alternate approach, let the line through G parallel to F D meet
AB, BC at Q0 , S 0 . The crux of the solution is showing that (BQ0 S 0 ), (DEF ), (AIC) are
coaxial, which can be done in many ways (see here, here, or here).
25
§6 Final Remarks
§6.1 Heuristics
Of course, knowing a theorem’s statement doesn’t mean anything unless you also know
when and how to apply it. Therefore, this section is dedicated to listing some situations
where Power of a Point is likely going to be useful.
• Power of a Point is first and foremost a length chasing tool. If a problem involves
a circle and asks for some condition that can be encoded into length ratios, start
computing powers of points!
• The orthocenter plays very nicely with Power of a Point. Thus, even when there
might not be any obvious circles in a problem, if there is a prominent orthocenter,
don’t count Power of a Point out of the equation just yet.
• It’s unlikely that the correct points to use PoP on will be immediately obvious, so
don’t be afraid to introduce lots of extra points until you find the right ones.
• If a problem involves proving that four points are concyclic but there’s very little
information about one of them, don’t start blindly flailing about with the basic
PoP theorem; the Forgotten Coaxiality Lemma is probably going to be more useful.
• Power of a Point is extremely powerful (no pun intended) for problems involving
tangencies if the examples haven’t made that obvious already. The first thing you
should try should just be the basic PoP theorem, but don’t be afraid to get clever
like with Example 3.7 or Example 4.3.
• Any time a concurrency and collinearity problem has one or more important circles,
you should think of radical axes/center. If the problem statement asks you to show
that three lines are “all parallel or concurrent,” that’s also a big red flag.
• Point circles may occasionally be useful for problems involving lots of equal lengths,
especially if there are perpendicular bisectors.
• Power of a Point (and especially linearity) can be helpful for problems involving
fixed points or circles as we saw in Example 5.8; showing that the power of some
point is fixed is a pretty underrated tool in my opinion.
§6.2 Further Reading

For more reading on length chasing, see AoPS user mira74’s Ratio Lemma handout or
SinaQane’s length bashing formula sheet. The topic of geometric inversion is also closely
related to Power of a Point, for which Gunmay Handa’s inversion handout (download
link!) and Chapter 8 of Evan Chen’s textbook Euclidean Geometry in Mathematical
Olympiads are useful references. Another useful article on an underrated geometric
transformation is Radek Olšák’s polar duality handout.
26
§7 Problems
Problems in each section are roughly sorted by difficulty. Please only refer to the hints if
you are well and truly stuck, as many are giveaways! I’ve also linked AoPS threads for
most of the problems, and once you solve or simply cannot make any more progress on
a problem, feel free to check out other people’s solutions; I personally have also posted
solutions on many of the threads.
§7.1 Introduction
Problem 7.1. Let AD be the A-altitude in triangle 4ABC. If H is a point on AD,
prove that H is the orthocenter of 4ABC if and only if BD · CD = AD · AH.
Problem 7.2 (AoPS). Let BT be the altitude and H be the intersection point of
the altitudes of triangle ABC. Point N is symmetric to H with respect to BC. The
circumcircle of triangle AT N intersects BC at points F and K. Prove that F B = BK.
Problem 7.3 (Sharygin CR 2021/16). Let circles Ω and ω touch internally at point A.
A chord BC of Ω touches ω at point K. Let O be the center of ω. Prove that the circle
BOC bisects segment AK.
Problem 7.4 (USAMO 1998/2). Let C1 and C2 be concentric circles, with C2 in the
interior of C1 . From a point A on C1 one draws the tangent AB to C2 (B ∈ C2 ). Let C
be the second point of intersection of ray AB and C1 , and let D be the midpoint of AB.
A line passing through A intersects C2 at E and F in such a way that the perpendicular
AM
bisectors of DE and CF intersect at a point M on AB. Find, with proof, the ratio M C.
Problem 7.5 (Canadian MO 2013/3). Let G be the centroid of a right-angled triangle

ABC with ∠BCA = 90◦ . Let P be the point on ray AG such that ∠CP A = ∠CAB, and
let Q be the point on ray BG such that ∠CQB = ∠ABC. Prove that the circumcircles
of triangles AQG and BP G meet at a point on side AB.
Problem 7.6 (AIME I 2020/15). Let ABC be an acute triangle with circumcircle ω
and orthocenter H. Suppose the tangent to the circumcircle of 4HBC at H intersects ω
at points X and Y with HA = 3, HX = 2, HY = 6. The area of 4ABC can be written
√
as m n, where m and n are positive integers, and n is not divisible by the square of any
prime. Find m + n.
Problem 7.7 (IMO 2019/2). In triangle ABC, point A1 lies on side BC and point B1
lies on side AC. Let P and Q be points on segments AA1 and BB1 , respectively, such
that P Q is parallel to AB. Let P1 be a point on line P B1 , such that B1 lies strictly
between P and P1 , and ∠P P1 C = ∠BAC. Similarly, let Q1 be the point on line QA1 ,
such that A1 lies strictly between Q and Q1 , and ∠CQ1 Q = ∠CBA. Prove that points
P, Q, P1 , and Q1 are concyclic.
Problem 7.8 (IMO Shortlist 2008/G4). In an acute triangle ABC segments BE and
CF are altitudes. Two circles passing through the point A and F and tangent to the
line BC at the points P and Q so that B lies between C and Q. Prove that lines P E
and QF intersect on the circumcircle of triangle AEF .
Problem 7.9 (IMO 2009/2). Let ABC be a triangle with circumcentre O. The points
P and Q are interior points of the sides CA and AB respectively. Let K, L and M be
the midpoints of the segments BP, CQ, and P Q. respectively, and let Γ be the circle
passing through K, L and M . Suppose that the line P Q is tangent to the circle Γ. Prove
that OP = OQ.
27
§7.2 Radical Axis

Problem 7.10 (Orthocenter). Prove that the three altitudes of a triangle are concurrent.
Problem 7.11 (APMO 2020/1). Let Γ be the circumcircle of 4ABC. Let D be a point
on the side BC. The tangent to Γ at A intersects the parallel line to BA through D at
point E. The segment CE intersects Γ again at F . Suppose B, D, F , E are concyclic.
Prove that AC, BF , DE are concurrent.
Problem 7.12 (AoPS). Let A, B, C, D be points in the plane such that CD is a line
perpendicular to AB and C, D 6∈ AB. If E and F are the orthocenters of 4ABC and
4ABD, prove that the radical axis of the circles with diameters CE and DF is AB.
Problem 7.13 (AoPS). In 4ABC, let I be the incenter and D, E, F be the touch points
of the incircle with BC, CA, AB, respectively. If AI meets (ABC) again at M and K, Q
are the points where the tangent to (ABC) at M meets AC, AB, respectively, show that
D lies on the radical axis of (EIK) and (F IQ).
Problem 7.14 (Orthic Axis). Let AD, BE, CF be altitudes in 4ABC, and let A1 =
BC ∩ EF, B1 = CA ∩ F D, and C1 = AB ∩ DE. Prove that A1 , B1 , C1 lie on a line
perpendicular to the Euler Line of 4ABC.
Problem 7.15 (IMO 2008/1). Let H be the orthocenter of an acute-angled triangle
ABC. The circle ΓA centered at the midpoint of BC and passing through H intersects
the sideline BC at points A1 and A2 . Similarly, define the points B1 , B2 , C1 , and C2 .
Prove that six points A1 , A2 , B1 , B2 , C1 , and C2 are concyclic.
Problem 7.16 (PUMaC Finals 2017/A3). Triangle ABC has incenter I. The line
through I perpendicular to AI meets the circumcircle of ABC at points P and Q, where
P and B are on the same side of AI. Let X be the point such that P X||CI and QX||BI.
Show that P B, QC, and IX intersect at a common point.
Problem 7.17 (Sharygin CR 2018/20, GOTEEM 2020/1). Let the incircle of a non-
isosceles triangle ABC touch AB, AC and BC at points D, E and F respectively. The
corresponding excircle touches the side BC at point N . Let T be the common point of
AN and the incircle, closest to N , and K be the common point of DE and F T . Prove
that AK||BC.
Problem 7.18 (TSTST 2017/1). Let ABC be a triangle with circumcircle Γ, circum-
center O, and orthocenter H. Assume that AB = 6 AC and that ∠A 6= 90◦ . Let M and
N be the midpoints of sides AB and AC, respectively, and let E and F be the feet of
the altitudes from B and C in 4ABC, respectively. Let P be the intersection of line
M N with the tangent line to Γ at A. Let Q be the intersection point, other than A, of Γ
with the circumcircle of 4AEF . Let R be the intersection of lines AQ and EF . Prove
that P R ⊥ OH.
Problem 7.19 (EGMO 2014/2). Let D and E be points in the interiors of sides AB
and AC, respectively, of a triangle ABC, such that DB = BC = CE. Let the lines CD
and BE meet at F . Prove that the incentre I of triangle ABC, the orthocentre H of
triangle DEF and the midpoint M of the arc BAC of the circumcircle of triangle ABC
are collinear.
Problem 7.20 (IMO 2012/5). Let ABC be a triangle with ∠BCA = 90◦ , and let D be
the foot of the altitude from C. Let X be a point in the interior of the segment CD. Let
K be the point on the segment AX such that BK = BC. Similarly, let L be the point
on the segment BX such that AL = AC. Let M be the point of intersection of AL and
BK. Show that M K = M L.
28
§7.3 Coaxial Circles

Problem 7.21. Let AD, BE, CF be the altitudes of a scalene triangle with circumcenter
O. Prove that (AOD), (BOE), and (COF ) are coaxial and furthermore that their
common radical axis is the Euler Line.
Problem 7.22 (AoPS). In scalene 4ABC with circumcircle Γ, let D be the midpoint
of BC. Line AD meets Γ again at A1 . The tangent to Γ at A1 intersects BC at A2 .
Denote by ωA the circumcircle of 4DA1 A2 , and define ωB , ωC similarly. Prove that ωA ,
ωB , and ωC are coaxial.
Problem 7.23 (AoPS). In 4ABC, let D, E, and F be the touch points of the incircle
with BC, CA, and AB, respectively. If P is the foot from D to EF and X = AB ∩ CP
and Y = AC ∩ BP, then prove that (AXY ), (AEF ), and (ABC) meet at a point other
than A.
Problem 7.24 (IMO Shortlist 2005/G5). Let 4ABC be an acute-angled triangle with
AB = 6 AC. Let H be the orthocenter of triangle ABC, and let M be the midpoint of
the side BC. Let D be a point on the side AB and E a point on the side AC such that
AE = AD and the points D, H, E are on the same line. Prove that the line HM is
perpendicular to the common chord of the circumscribed circles of triangle 4ABC and
triangle 4ADE.
Problem 7.25 (SMO 2020/5). In triangle 4ABC, let E and F be points on sides AC
and AB, respectively, such that BF EC is cyclic. Let lines BE and CF intersect at point
P , and M and N be the midpoints of BF and CE, respectively. If U is the foot of the
perpendicular from P to BC, and the circumcircles of triangles 4BM U and 4CN U
intersect at second point V different from U , prove that A, P, and V are collinear.
Problem 7.26 (European Mathematical Cup 2016/4). Let C1 , C2 be circles intersecting
in X, Y . Let A, D be points on C1 and B, C on C2 such that A, X, C are collinear and
D, X, B are collinear. The tangent to circle C1 at D intersects BC and the tangent to
C2 at B in P , R respectively. The tangent to C2 at C intersects AD and tangent to C1
at A, in Q, S respectively. Let W be the intersection of AD with the tangent to C2 at B
and Z the intersection of BC with the tangent to C1 at A. Prove that the circumcircles
of triangles Y W Z, RSY and P QY have two points in common, or are tangent in the
same point.
Problem 7.27 (IMO 1979/3). Two circles in a plane intersect. A is one of the points
of intersection. Starting simultaneously from A two points move with constant speed,
each travelling along its own circle in the same sense. The two points return to A
simultaneously after one revolution. Prove that there is a fixed point P in the plane such
that the two points are always equidistant from P.
Problem 7.28 (AoPS, taken from here). Let ABC be a triangle and let L, M, N be
points on sides BC, CA, AB respectively. Let Q be the Miquel point of L, M, N with
respect to 4ABC. Show that if AL, BM, CN concur, then (AQL), (BQM ), (CQN ) are
coaxial.
§7.4 Point Circles

Problem 7.29 (Iran TST 2011/1). In acute triangle ABC, ∠B is greater than ∠C. Let
M be the midpoint of BC and let E and F be the feet of the altitudes from B and C,
respectively. Let K and L be the midpoints of M E and M F , respectively, and let T be
on line KL such that T A k BC. Prove that T A = T M .
29
Problem 7.30 (Iran MO 2017 Round 3/G3). Let ABC be an acute-angle triangle.
Suppose that M be the midpoint of BC and H be the orthocenter of ABC. Let
F ≡ BH ∩ AC and E ≡ CH ∩ AB. Suppose that X be a point on EF such that
∠XM H = ∠HAM and A, X are in the distinct side of M H. Prove that AH bisects
M X.
Problem 7.31 (2015 Sharygin CR/19). Let L and K be the feet of the internal and
the external bisector of angle A of a triangle ABC. Let P be the common point of the
tangents to the circumcircle of the triangle at B and C. The perpendicular from L to
BC meets AP at point Q. Prove that Q lies on the medial line of triangle LKP .
Problem 7.32 (IMO Shortlist 2009/G3). Let ABC be a triangle. The incircle of
ABC touches the sides AB and AC at the points Z and Y , respectively. Let G be the
point where the lines BY and CZ meet, and let R and S be points such that the two
quadrilaterals BCY R and BCSZ are parallelogram. Prove that GR = GS.
Problem 7.33 (ARMO 2011 Grade 10/4). The perimeter of triangle ABC is 4. Point
X is marked on ray AB and point Y is marked on ray AC such that AX = AY = 1. If
BC intersects XY at point M , prove that perimeter of one of triangles ABM or ACM
is 2.
Problem 7.34 (NICE MO 2021/2). Let O be the circumcenter of triangle ABC. Suppose
the perpendicular bisectors of OB and OC intersect lines AB and AC at D 6= A and
E=6 A, respectively. Determine the maximum possible number of distinct intersection
points between line BC and the circumcircle of 4ADE.
§7.5 Black Magic

Problem 7.35 (IMO Shortlist 2011/G2). Let A1 A2 A3 A4 be a non-cyclic quadrilateral.
Let O1 and r1 be the circumcentre and the circumradius of the triangle A2 A3 A4 . Define
O2 , O3 , O4 and r2 , r3 , r4 in a similar way. Prove that
1 1 1 1
+ + + = 0.
O1 A21 − r12 O2 A22 − r22 O3 A23 − r32 O4 A24 − r42
Problem 7.36. In triangle 4ABC, let O, G, and K be the circumcenter, centroid, and
Symmedian Point, respectively. Prove that OK ≥ OG.
Problem 7.37 (Taiwan TST 2016 Round 2/1). Let O be the circumcenter of triangle
ABC, and ω be the circumcircle of triangle BOC. Line AO intersects with circle ω again
at the point G. Let M be the midpoint of side BC, and the perpendicular bisector of BC
meets circle ω at the points O and N . Prove that the midpoint of the segment AN lies
on the radical axis of the circumcircle of triangle OM G, and the circle whose diameter is
AO.
Problem 7.38 (ELMO Shortlist 2013/G3). In 4ABC, a point D lies on line BC. The
circumcircle of ABD meets AC at F (other than A), and the circumcircle of ADC meets
AB at E (other than A). Prove that as D varies, the circumcircle of AEF always passes
through a fixed point other than A, and that this point lies on the median from A to BC.
Problem 7.39 (USA December TST 2012/1). In acute triangle ABC, ∠A < ∠B and
∠A < ∠C. Let P be a variable point on side BC. Points D and E lie on sides AB and
AC, respectively, such that BP = P D and CP = P E. Prove that as P moves along side
BC, the circumcircle of triangle ADE passes through a fixed point other than A.
30
§7.6 Challenge Problems

Problem 7.40 (Haruki’s Lemma). Let AB, CD be two fixed chords and P be a variable
point on a circle. If Q = P C ∩ AB and R = P D ∩ AB, prove that the quantity AQ·RBQR
does not depend on P. As a bonus, try using this lemma to prove the Butterfly Theorem!
Problem 7.41 (USA TST 2006/6). Let ABC be a triangle. Triangles P AB and QAC
are constructed outside of triangle ABC such that AP = AB and AQ = AC and
∠BAP = ∠CAQ. Segments BQ and CP meet at R. Let O be the circumcenter of
triangle BCR. Prove that AO ⊥ P Q.
Problem 7.42 (Iran MO 2003 Round 3/P6). Let the incircle of a triangle ABC touch
BC, AC, AB at A1 , B1 , C1 respectively. Let M and N be the midpoints of AB1 and AC1 ,
respectively, and M N meet A1 C1 at T . Draw two tangents T P and T Q through T to
incircle. P Q meets M N at L and B1 C1 meets P Q at K. If I is the center of the incircle,
prove that IK is parallel to AL.
Problem 7.43 (USA December TST 2020/2). Two circles Γ1 and Γ2 have common
external tangents `1 and `2 meeting at T . Suppose `1 touches Γ1 at A and `2 touches Γ2
at B. A circle Ω through A and B intersects Γ1 again at C and Γ2 again at D, such that
quadrilateral ABCD is convex. Suppose lines AC and BD meet at point X, while lines
AD and BC meet at point Y . Show that T , X, Y are collinear.
Problem 7.44 (Crucial lemma of Sharygin CR 2021/18). Let 4ABC be a triangle
with incenter I, let D be some point on BC, let K be the midpoint of ID, and P =
AI ∩ (ABC) 6= A. If ω1 is the circle tangent to segments AD, BD, and circle (ABC)
internally and ω2 is the circle tangent to segments AD, CD, and (ABC) internally, prove
that P K is the radical axis of ω1 and ω2 .
Problem 7.45 (USA February TST 2021/2). Points A, V1 , V2 , B, U2 , U1 lie fixed
on a circle Γ, in that order, and such that BU2 > AU1 > BV2 > AV1 . Let X be a
variable point on the arc V1 V2 of Γ not containing A or B. Line XA meets line U1 V1
at C, while line XB meets line U2 V2 at D. Let O and ρ denote the circumcenter and
circumradius of 4XCD, respectively. Prove there exists a fixed point K and a real
number c, independent of X, for which OK 2 − ρ2 = c always holds regardless of the
choice of X.
Problem 7.46 (IMO Shortlist 2017/G7). A convex quadrilateral ABCD has an inscribed
circle with center I. Let Ia , Ib , Ic and Id be the incenters of the triangles DAB, ABC, BCD
and CDA, respectively. Suppose that the common external tangents of the circles AIb Id
and CIb Id meet at X, and the common external tangents of the circles BIa Ic and DIa Ic
meet at Y . Prove that ∠XIY = 90◦ .
Problem 7.47 (USEMO 2019/6). Let ABC be an acute scalene triangle with circum-
center O and altitudes AD, BE, CF . Let X, Y , Z be the midpoints of AD, BE, CF .
Lines AD and Y Z intersect at P , lines BE and ZX intersect at Q, and lines CF and
XY intersect at R. Suppose that lines Y Z and BC intersect at A0 , and lines QR and
EF intersect at D0 . Prove that the perpendiculars from A, B, C, O, to the lines QR,
RP , P Q, A0 D0 , respectively, are concurrent.
Problem 7.48 (IMO 2019/6). Let I be the incentre of acute triangle ABC with
AB = 6 AC. The incircle ω of ABC is tangent to sides BC, CA, and AB at D, E, and F ,
respectively. The line through D perpendicular to EF meets ω at R. Line AR meets ω
again at P . The circumcircles of triangle P CE and P BF meet again at Q. Prove that
lines DI and P Q meet on the line through A perpendicular to AI.
31
§8 Hints
7.1. Reflect H over BC.
7.2. Reflect H over B.
7.3. Draw the intersection of BC and the common tangent at A.
7.4. Proving that DEF C is cyclic implies that M is the midpoint of CD.
7.5. Show that (CGP ) is tangent to AC.
7.6. Reflect H over BC and construct the tangent to ω at that point.
7.7. Extend AA1 and BB1 to hit the circumcircle again.
7.8. Prove that AEP Q is cyclic, then angle chase.

−
7.9. Show that 4ABQ ∼ 4M KL.
7.10. H is the radical center of three circles.
7.12. Let P = AB ∩ CD, and use PoP.
7.13. KI is the diameter of (EIK).
7.14. Show that A1 , B1 , C1 lie on the radical axis of the nine-point circle and (ABC).
7.15. Either show that if (A1 A2 C1 C2 ) 6= (B1 B2 C1 C2 ), then C would lie on their radical
axis, or show that (A1 A2 C1 C2 ), (B1 B2 C1 C2 ), (C1 C2 A1 A2 ) share a common center.
7.16. Draw D = P X ∩ CQ, E = QX ∩ BP. Then use radical center twice.
7.17. Radical center on (AEF ), (DEF ), (ADT ).
7.18. Use radical center twice to show that P and R lie on the radical axis of (ABC)
and the nine-point circle.
7.19. M, I, H form the radical axis of two circles.
7.20. Introduce E and F, the second intersections of BX with the circle centered at A
with radius AC and AX with the circle centered at B with radius BC, respectively.
7.21. Show that H lies on the pairwise radical axes.
7.22. Prove that the circumcenter and centroid lie on the radical axes.
7.23. Use the Forgotten Coaxiality Lemma on X, Y with respect to (AEF ), (ABC).
7.24. Use the Forgotten Coaxiality lemma on D, E with respect to (AH), (ABC).
7.25. Use the Forgotten Coaxiality Lemma on P, Q = EF ∩ BC with respect to

(BM U ), (CM U ).
7.26. Use the Forgotten Coaxiality Lemma on W, Z with respect to (P QY ) and (RSY ).
7.27. If the two points are X and Y, prove that the midpoint of XY lies on a fixed circle
with the Forgotten Coaxiality Lemma.
7.28. Use the Forgotten Coaxiality Lemma 5 times.
32
7.29. Consider the radical axis of M and (AEF ).
7.30. AH ∩ M X is the radical center of M and two other circles.
7.31. Q is the radical center of L and two other circles.
7.32. Introduce the A-excircle; there are lots of equal lengths that result.
7.33. Reflect X over B and Y over C to get the tangency points of the A-excircle.
7.34. Consider the radical axis of O and (ADE).
7.35. Pick any 3 of the Ai as your reference triangle, then just use the bary PoP formula.
7.36. It suffices to prove that Pow(O, (ABC)) ≥ Pow(G, (ABC)), which can be done
with either straight bary or Example 5.3.
7.37. Plug P into f (•) = Pow(•, (OM G)) − Pow(•, (AO)).
7.38. Show that Pow(M, (AEF )) is fixed.
7.39. Show that Pow(H, (ADE)) is fixed using H = (tan A : tan B : tan C).
7.40. Extend AB to meet (P QD) at E. Is there anything special about the length of
BE?
7.41. Introduce X = AP ∩ (ACQR), Y = AP ∩ (ABP R). Prove that P O2 − P A2 =

Pow(A, (P QX)).
7.42. L is the radical center of A, (AP E), and the incircle.
7.43. Consider the radical axis of (ADE) and (F CB).
7.44. For P, simple PoP manipulations do the trick. For K, let I1 , I2 be the centers of
ω1 , ω2 , and let V be the reflection of D over M. If f (•) = Pow(•, ω1 ) − Pow(•, ω2 ),
after some manipulations of the equation f (K) = 12 (f (I) + f (D)) = 0, it suffices to
show that IV ⊥ I1 I2 . Also, don’t forget the Sawayama-Thebault Theorem!
7.45. Let A0 = A∞U1 V1 ∩ Γ 6= A and B 0 = B∞U2 V2 ∩ Γ 6= B. Then prove that K =

AB 0 ∩ AA0 is fixed with the function Pow(•, (XCD)) − Pow(•, Γ).
7.46. Use the Forgotten Coaxiality Lemma on X, I with respect to (AIb Ic ), (CIb Ic ), and
use XIIb Ic cyclic to show that XI bisects ∠Ib IIc ; symmetry finishes the problem
from there.
7.47. Prove that A0 D0 is the radical axis of (DEF ) and (XY Z).
7.48. Define f1 (•) = Pow(•, (BDIF )) − Pow(•, (CDIE)) and f2 (•) = Pow(•, (BP F )) −
Pow(•, (CP E)). The crucial step is proving that for any point T on AX, f1 (T ) =
f2 (T ).
33
§9 Appendices
§9.1 Appendix A: Notations and conventions
• ∩ denotes the intersection of two objects. For example, if a point P is defined as
the intersection of two lines `1 and `2 , it would be written as P = `1 ∩ `2 .
• ∈ denotes that a point is on a figure. For example, if a point P lies on circle ω, it
would be written as P ∈ ω.
• Parentheses denote the circumcircle of a figure or the diameter circle if we only list
two points. For example, the circumcircle of a triangle 4ABC is (ABC), and the
circle with diameter AB is (AB).
• The abbreviation PoP means “Power of a Point,” and WLOG means “Without
Loss Of Generality.” I have no idea why the O is inconsistently capitalized either.
• The circumradius of the reference triangle (usually 4ABC) is denoted by R, and
the inradius is denoted by r. Also, in 4ABC, a, b, and c denote the lengths of
sides BC, CA, and AB, respectively, and we use ∠A, ∠B, and ∠C as shorthand for
∠CAB, ∠ABC, and ∠BCA, respectively.
+ −
• The symbol ∼ means directly similar, and the symbol ∼ means inversely similar.
+
For example, if 4ABC ∼ 4DEF, it means that in terms of directed angles,
−
]ABC = ]DEF, while if 4ABC ∼ 4DEF, it means that ]ABC = ]F ED.
• ∞` denotes the point at infinity on a line `, i.e. the point that all lines parallel to `
pass through. In particular, P ∞` is the line through point P parallel to line `.
§9.2 Appendix B: Referenced theorems without proof
Theorem 9.1 (Angle Bisector Theorem)

In triangle 4ABC, let D be a point on segment BC. Then AB BD
AC = CD if and only if D
lies on the A-internal angle bisector. If D is outside of segment BC, then AB
AC = CD
BD
if and only if D lies on the A-external angle bisector.
Theorem 9.2 (Reim’s Theorem)

Let ABP Q be a cyclic quadrilateral. If C and D are two points on lines AQ and
AP, respectively, then ABCD is cyclic if and only if CD k P Q.
Theorem 9.3 (Menelaus’s Theorem)

In triangle 4ABC, let D, E, F be points on lines BC, CA, AB, respectively. Then
D, E, F are collinear if and only if
DB CE AF
· · = −1,
DC AE BF
where lengths are directed (i.e. DB ~ is in the same direction as DC

is positive if DB ~
DC
and negative otherwise).
34
Theorem 9.4 (Stewart’s Theorem)

In triangle 4ABC, let D be a point on BC. Then if BD = m and CD = n, we have
a(mn + d2 ) = b2 m + c2 n ⇐⇒ man + dad = bmb + cnc.
Theorem 9.5 (Law of Sines)

In triangle 4ABC, we have
a b c
= = = 2R.
sin ∠A sin ∠B sin ∠C
Theorem 9.6 (Law of Cosines)

In triangle 4ABC, we have
c2 = a2 + b2 − 2ab cos C.
Theorem 9.7 (Miquel Pivot Theorem)

In triangle 4ABC, let D, E, F be points on BC, CA, AB, respectively. Then circles
(AEF ), (BF D), (CDE) have one point in common, the Miquel Point of D, E, F
with respect to 4ABC.
Theorem 9.8 (Schwatt Line)

In triangle 4ABC, let XA , XB , XC be the midpoints of the A-, B-, C-altitudes,
respectively, and let MA , MB , MC be the midpoints of BC, CA, AB, respectively.
Then XA MA , XB MB , XC MC concur at the Symmedian Point of 4ABC.
§9.3 Appendix C: Trigonometric bash for TSTST 2016/6
Claim 9.9 — We wish to show that

CD2 dist(G, D0 E0 ) dist(B, F0 D0 )
· · = 1.
BD2 dist(G, F0 D0 ) dist(C, D0 E0 )
Proof. Abbreviate ∠D = ∠EDF, ∠E = ∠F ED, and ∠F = ∠DF E. For the first quotient,
we have
◦ − 1 ∠CBA) 2
!
1
CD2 2 DE cos(90 2 DE 2 cos2 ∠E
= 1 · 1 = · .
BD2 ◦
2 F D cos(90 − 2 ∠ACB)
F D2 cos2 ∠F
For the second quotient, it’s well-known that the Gergonne Point is the isotomic conju-
gate of the orthocenter in 4D0 E0 F0 , so the barycentric coordinates of G in 4D0 E0 F0
are (cot ∠D : cot ∠E : cot ∠F ), giving
1
dist(G, D0 E0 ) [GD0 E0 ] D0 E0 cot ∠F sin ∠E cos ∠F F D2
= · 1 = · = ·
dist(G, F0 D0 ) [GF0 D0 ] F0 D0
cot ∠E sin ∠F cos ∠E DE 2
35
with the second and third equalities coming from the Law of Sines on 4D0 E0 F0 and
4DEF, respectively.
For the third quotient, we have
1
dist(B, F0 D0 ) = BD sin ∠E + DE sin ∠D
2
DE · EF
= BD sin ∠E + sin ∠E
2F D
FD DE · EF
= sin ∠E +
2 cos ∠E 2F D
sin ∠E
F D2 + DE · EF · cos ∠E

=
2F D cos ∠E
DE 2 + EF 2 − F D2

sin ∠E 2
= F D + DE · EF ·
2F D cos ∠E 2DE · EF
sin ∠E
F D2 + DE 2 + EF 2 ,

=
4F D cos ∠E
where the second and fifth equality come from Law of Sines and Law of Cosines, respec-
tively, on 4DEF. Thus,
dist(B, F0 D0 ) cos ∠F F D sin ∠E cos ∠F

= · = ,
dist(C, D0 E0 ) cos ∠E DE sin ∠F cos ∠E
with the second equality coming from Law of Sines on 4DEF, so multiplying all three
quotients yields 1 as desired.
36

Power of A Point: Albert Zhu

Uploaded by

Copyright:

Available Formats

Power of A Point: Albert Zhu

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Power of A Point: Albert Zhu

Uploaded by

Copyright:

Available Formats

Power of a Point

Theorem 1.1 (Power of a Point)

Proof. Observe that

Note that this quantity actually has a sign (positive if P is outside ω, 0 if P is on ω,

Example 1.5 (Euler’s Theorem)

Exercise 1.6. Let H be the orthocenter of 4ABC. Prove that

Example 1.7 (Nine-Point Circle)

• HA , HB , HC be the feet of the altitudes from A, B, C respectively,

• MA , MB , MC be the midpoints of BC, CA, AC, respectively, and

• XA , XB , XC be the midpoints of AH, BH, CH, respectively.

Then HA , HB , HC , MA , MB , MC , XA , XB , XC lie on a circle whose center is the

so the converse of PoP implies that HA , HB , HC , XA , XB , XC all lie on a circle ω.

∠HA XA MA = 90◦ − ∠XA MA HA

so AO k XA MA and in fact AXA MA O is a parallelogram as we also have AXA ⊥

Example 1.8 (IMO Shortlist 2013/G4)

Theorem 2.2 (Radical Axis)

Pow(P, ω1 ) = (x − a)2 + y 2 − r12

which is the equation of a line perpendicular to the x-axis as desired. Also, if we

Remark 2.5. Furthermore, if we substitute the 0 with y, then Pow(P, ω1 ) − Pow(P, ω2 ) is

Theorem 2.6 (Radical Center)

Pow(P, ω1 ) = Pow(P, (ABCD)) = Pow(P, ω2 ),

Since P A · P B is negative if and only if P is between A and B and positive if and

Now let’s do some examples.

Pow(M, ω1 ) = CM 2 = DM 2 = Pow(M, ω2 ) = CM · DM = Pow(M, (ACD)),

so why isn’t M the radical center of ω1 , ω2 , and (ACD)?

Example 2.10 (USAJMO 2012/1)

Pow(A, (P SR)) = AP 2 = AQ2 = Pow(A, (QSR)),

so P, Q, S, R are concyclic as desired.

Example 2.12 (USAMO 2009/1)

Proof. Let O1 , O2 , O3 be the centers of ω1 , ω2 , (P RQS), respectively. Drawing the dia-

which is symmetrically equal for OQ2 , OR2 , OS 2 as well, so O = O3 .

Lemma 3.2 (Forgotten Coaxiality Lemma)

First, we present two classical results.

Example 3.3 (Newton-Gauss Line)

Example 3.4 (Existence of Isodynamic Points)

Now we obliterate some problems with the Forgotten Coaxiality Lemma.

Example 3.5 (CGMO 2017/7)

Example 3.6 (IMO Shortlist 2012/G8)

Example 3.7 (IMO 2015/3)

Example 4.1 (Circumcenter)

Example 4.2 (Polish MO 2018/5)

Example 4.3 (CGMO 2015/6)

Example 4.4 (Balkan MO 2015/2)

§5.1 Barycentric Coordinates

Theorem 5.1 (Barycentric Power of a Point)

−a2 yz − b2 zx − c2 xy + (x + y + z)(ux + vy + wz) = 0.

Then if P = (x, y, z) is any point in the plane,

Pow(P, ω) = −a2 yz − b2 zx − c2 xy + (x + y + z)(ux + vy + wz).

Proof. Since the formula of a circle is derived from

As a demonstration of how useful this formula can be even by itself, we present an

Example 5.2 (AQGO 2020/2)

a2 yz + b2 zx + c2 xy = (x + y + z)(c2 (1 − e)x + a2 (1 − d)z)

and similar computations yield

P B1 = bx, P B2 = bz, P C1 = cy, P C2 = cx.

Note that if the directed length product P A1 · P A2 is positive, then P is outside of

Theorem 5.4 (Linearity of PoP)

CO12 − CO22 = k(AO12 − AO22 ) + (1 − k)(BO12 − BO22 )

f (O) = −Pow(O, ω) = −Pow(H, ω) = EH · HD = (R cos ∠A)(2R cos ∠B cos ∠C),

We also know that OM k AD, so