Soluciones del 2026 AIME II

Desplázate hacia abajo para ver las soluciones preparadas profesionalmente de LIVE by Po-Shen Loh, imprime las soluciones en PDF, consulta la clave de respuestas, o haz el examen cronometrado completo.

Todos los problemas se usan con el permiso legal oficial de la Mathematical Association of America (MAA).

1.

Halla la suma de los 1010-ésimos términos de todas las sucesiones aritméticas de enteros cuyo primer término es igual a 44 y que incluyen tanto a 2424 como a 3434 entre sus términos.

Find the sum of the 1010th terms of all arithmetic sequences of integers that have first term equal to 44 and include both 2424 and 3434 as terms.

Conceptos:sucesión aritméticadivisibilidadmáximo común divisor
Nivel de dificultad: 1840
Pista pequeña:

Si la diferencia común es d,d, entonces que 2424 y 3434 sean ambos términos obliga a que dd divida tanto a 2020 como a 30.30.

If the common difference is d,d, then 2424 and 3434 both being terms forces dd to divide both 2020 and 30.30.

Pista grande:

La diferencia común debe ser un divisor positivo de 10;10; suma los 1010-ésimos términos 4+9d4 + 9d sobre todos esos d.d.

The common difference must be a positive divisor of 10;10; add up the 1010th terms 4+9d4 + 9d over all such d.d.

Solución:

Sea la diferencia común d.d. Como el primer término es 44 y aparecen tanto 2424 como 3434, dd divide a 244=2024 - 4 = 20 y a 344=30,34 - 4 = 30, así que dd divide a gcd(20,30)=10.\gcd(20, 30) = 10. La diferencia debe ser positiva para alcanzar 2424 y 3434 desde 4,4, por lo que d{1,2,5,10}d \in \{1, 2, 5, 10\} (y cada uno de ellos funciona, ya que 2020 y 3030 son ambos divisibles por d,d, lo que coloca ambos objetivos en la sucesión).

El 1010-ésimo término es 4+9d,4 + 9d, por lo que la suma pedida es d{1,2,5,10}(4+9d)=44+9(1+2+5+10)=16+162=178. \begin{aligned} &\sum_{d \in \{1,2,5,10\}} (4 + 9d) \\ &= 4 \cdot 4 + 9(1 + 2 + 5 + 10) \\ &= 16 + 162 = 178. \end{aligned}

Let the common difference be d.d. Since the first term is 44 and both 2424 and 3434 appear, dd divides 244=2024 - 4 = 20 and 344=30,34 - 4 = 30, so dd divides gcd(20,30)=10.\gcd(20, 30) = 10. The difference must be positive to reach 2424 and 3434 from 4,4, so d{1,2,5,10}d \in \{1, 2, 5, 10\} (and each of these works, since 2020 and 3030 are both divisible by d,d, which puts both targets in the sequence).

The 1010th term is 4+9d,4 + 9d, so the requested sum is d{1,2,5,10}(4+9d)=44+9(1+2+5+10)=16+162=178. \begin{aligned} &\sum_{d \in \{1,2,5,10\}} (4 + 9d) \\ &= 4 \cdot 4 + 9(1 + 2 + 5 + 10) \\ &= 16 + 162 = 178. \end{aligned}

2.

La figura de abajo muestra una cuadrícula de 1010 cuadrados en una fila. Cada cuadrado tiene una diagonal que conecta su vértice inferior izquierdo con su vértice superior derecho. Un insecto se mueve a lo largo de los segmentos de vértice a vértice, sin recorrer nunca el mismo segmento dos veces y sin moverse nunca de derecha a izquierda por un segmento horizontal o diagonal. Sea NN el número de caminos que el insecto puede tomar desde la esquina inferior izquierda (A)(A) hasta la esquina superior derecha (B).(B). Uno de esos caminos de AA a BB se muestra con los segmentos gruesos en la figura. Halla N.\sqrt{N}.

The figure below shows a grid of 1010 squares in a row. Each square has a diagonal connecting its lower left vertex to its upper right vertex. A bug moves along the line segments from vertex to vertex, never traversing the same segment twice and never moving from right to left along a horizontal or diagonal segment. Let NN be the number of paths the bug can take from the lower left corner (A)(A) to the upper right corner (B).(B). One such path from AA to BB is shown by the thick line segments in the figure. Find N.\sqrt{N}.

Nivel de dificultad: 2230
Pista pequeña:

Como el insecto nunca se mueve hacia la izquierda, cruza la franja entre x=ix = i y x=i+1x = i + 1 exactamente una vez, por el borde inferior, el borde superior o la diagonal.

Since the bug never moves left, it crosses the strip between x=ix = i and x=i+1x = i + 1 exactly once, along the bottom edge, the top edge, or the diagonal.

Pista grande:

Los segmentos verticales quedan completamente determinados por las diez elecciones de cruce, así que cada cuadrado contribuye independientemente con 33 opciones.

The vertical segments are completely determined by the ten crossing choices, so each square independently contributes 33 options.

Solución:

Pongamos A=(0,0)A = (0, 0) y B=(10,1).B = (10, 1). Todo movimiento horizontal y diagonal va hacia la derecha, así que la coordenada xx del insecto nunca disminuye, y cruza cada una de las 1010 franjas verticales exactamente una vez, usando exactamente uno de los tres segmentos hacia la derecha de ese cuadrado: el borde inferior, el borde superior, o la diagonal.

Estas diez elecciones determinan todo el camino. Cada cruce llega a una altura determinada (borde inferior: baja; borde superior o diagonal: alta) y sale a una altura determinada (borde inferior o diagonal: baja; borde superior: alta), así que en cada línea vertical el insecto recorre el segmento vertical exactamente cuando las alturas de llegada y de salida difieren, y cada segmento vertical se necesita a lo sumo una vez, por lo que ningún segmento se repite. Lo mismo vale en los extremos: el insecto empieza abajo en AA y termina arriba en B,B, usando los verticales de los extremos si es necesario. Recíprocamente, toda sucesión de elecciones produce un camino válido.

Por lo tanto N=310,N = 3^{10}, y N=35=243.\sqrt{N} = 3^5 = 243.

Put A=(0,0)A = (0, 0) and B=(10,1).B = (10, 1). Every horizontal and diagonal move goes rightward, so the bug’s xx-coordinate never decreases, and it crosses each of the 1010 vertical strips exactly once, using exactly one of that square’s three rightward segments: the bottom edge, the top edge, or the diagonal.

These ten choices determine the whole path. Each crossing arrives at a definite height (bottom edge: low; top edge or diagonal: high) and departs at a definite height (bottom edge or diagonal: low; top edge: high), so at each vertical line the bug traverses the vertical segment exactly when the arrival and departure heights differ — and each vertical segment is needed at most once, so no segment repeats. The same applies at the ends: the bug starts low at AA and finishes high at B,B, using the end verticals if necessary. Conversely, every sequence of choices yields a valid path.

Therefore N=310,N = 3^{10}, and N=35=243.\sqrt{N} = 3^5 = 243.

3.

Sea ABCDEABCDE un pentágono no convexo con ángulos internos A=E=90\angle A = \angle E = 90^\circ y B=D=45.\angle B = \angle D = 45^\circ. Supongamos que DE<AB,DE \lt AB, AE=20,AE = 20, BC=142,BC = 14\sqrt{2}, y que los puntos B,B, C,C, y DD están del mismo lado de la recta AE.AE. Supongamos además que ABAB es un entero con AB<2026AB \lt 2026 y que el área del pentágono ABCDEABCDE es un múltiplo entero de 16.16. Halla el número de valores posibles de AB.AB.

Let ABCDEABCDE be a nonconvex pentagon with internal angles A=E=90\angle A = \angle E = 90^\circ and B=D=45.\angle B = \angle D = 45^\circ. Suppose that DE<AB,DE \lt AB, AE=20,AE = 20, BC=142,BC = 14\sqrt{2}, and points B,B, C,C, and DD lie on the same side of line AE.AE. Suppose further that ABAB is an integer with AB<2026AB \lt 2026 and the area of pentagon ABCDEABCDE is an integer multiple of 16.16. Find the number of possible values of AB.AB.

Nivel de dificultad: 2510
Pista pequeña:

Pon AA en el origen y E=(20,0);E = (20, 0); los ángulos rectos en AA y EE hacen que ABAB y EDED sean verticales.

Put AA at the origin and E=(20,0);E = (20, 0); the right angles at AA and EE make ABAB and EDED vertical.

Pista grande:

Los ángulos de 4545^\circ en BB y DD obligan a que C=(14,AB14)C = (14, AB - 14) y DE=AB8;DE = AB - 8; la fórmula del zapatero convierte la condición de área en una congruencia para AB.AB.

The 4545^\circ angles at BB and DD force C=(14,AB14)C = (14, AB - 14) and DE=AB8;DE = AB - 8; the shoelace formula turns the area condition into a congruence for AB.AB.

Solución:

Coloquemos A=(0,0)A = (0, 0) y E=(20,0)E = (20, 0) con el pentágono por encima de la recta AE,AE, y escribamos h=AB.h = AB. Los ángulos rectos en AA y EE hacen que ABAB y EDED sean verticales: B=(0,h)B = (0, h) y D=(20,k)D = (20, k) con k=DE.k = DE. En BB el lado BC=142BC = 14\sqrt{2} forma un ángulo de 4545^\circ con el rayo descendente BA,BA, dirigiéndose hacia el interior del pentágono, así que C=(14,h14).C = (14, h - 14). De manera similar en D,D, el lado DCDC forma un ángulo de 4545^\circ con el rayo descendente DE,DE, así que C=(20s,ks)C = (20 - s, k - s) donde s=DC2.s = \frac{DC}{\sqrt{2}}. Igualar coordenadas da s=6s = 6 y k=h8.k = h - 8. El ángulo interior en CC es entonces el ángulo reflejo 270270^\circ (suma de ángulos 90+45+270+45+90=54090 + 45 + 270 + 45 + 90 = 540), y DE=h8<ABDE = h - 8 \lt AB automáticamente.

La fórmula del zapatero aplicada a A(0,0),A(0,0), B(0,h),B(0,h), C(14,h14),C(14, h-14), D(20,h8),D(20, h-8), E(20,0)E(20, 0) da el área [ABCDE]=1214h+(6h+168)+(20h+160)=20h164. \begin{aligned} [ABCDE] &= \frac{1}{2}\big|{-14h} + (-6h + 168) \\ &\qquad {}+ (-20h + 160)\big| \\ &= 20h - 164. \end{aligned} La condición de que 20h16420h - 164 sea divisible por 1616 se reduce a 4h4(mod16),4h \equiv 4 \pmod{16}, es decir, h1(mod4).h \equiv 1 \pmod 4. Para que CC esté estrictamente del mismo lado de la recta AEAE que BB y D,D, necesitamos h>14.h \gt 14.

Así que hh recorre 17,21,25,,2025,17, 21, 25, \ldots, 2025, lo que da 2025174+1=503\frac{2025 - 17}{4} + 1 = 503 valores.

Place A=(0,0)A = (0, 0) and E=(20,0)E = (20, 0) with the pentagon above line AE,AE, and write h=AB.h = AB. The right angles at AA and EE make ABAB and EDED vertical: B=(0,h)B = (0, h) and D=(20,k)D = (20, k) with k=DE.k = DE. At BB the side BC=142BC = 14\sqrt{2} makes a 4545^\circ angle with the downward ray BA,BA, heading into the pentagon, so C=(14,h14).C = (14, h - 14). Similarly at D,D, the side DCDC makes a 4545^\circ angle with the downward ray DE,DE, so C=(20s,ks)C = (20 - s, k - s) where s=DC2.s = \frac{DC}{\sqrt{2}}. Matching coordinates gives s=6s = 6 and k=h8.k = h - 8. The interior angle at CC is then the reflex angle 270270^\circ (angle sum 90+45+270+45+90=54090 + 45 + 270 + 45 + 90 = 540), and DE=h8<ABDE = h - 8 \lt AB automatically.

The shoelace formula on A(0,0),A(0,0), B(0,h),B(0,h), C(14,h14),C(14, h-14), D(20,h8),D(20, h-8), E(20,0)E(20, 0) gives area [ABCDE]=1214h+(6h+168)+(20h+160)=20h164. \begin{aligned} [ABCDE] &= \frac{1}{2}\big|{-14h} + (-6h + 168) \\ &\qquad {}+ (-20h + 160)\big| \\ &= 20h - 164. \end{aligned} The condition that 20h16420h - 164 be divisible by 1616 reduces to 4h4(mod16),4h \equiv 4 \pmod{16}, that is, h1(mod4).h \equiv 1 \pmod 4. For CC to lie strictly on the same side of line AEAE as BB and D,D, we need h>14.h \gt 14.

So hh runs over 17,21,25,,2025,17, 21, 25, \ldots, 2025, which is 2025174+1=503\frac{2025 - 17}{4} + 1 = 503 values.

4.

Para cada entero positivo nn sea f(n)f(n) el valor del numeral en base diez nn visto en base b,b, donde bb es el menor entero mayor que el mayor dígito de n.n. Por ejemplo, si n=72,n = 72, entonces b=8,b = 8, y 7272 como numeral en base 88 es igual a 78+2=58;7 \cdot 8 + 2 = 58; por lo tanto f(72)=58.f(72) = 58. Halla el número de enteros positivos nn menores que 10001000 tales que f(n)=n.f(n) = n.

For each positive integer nn let f(n)f(n) be the value of the base-ten numeral nn viewed in base b,b, where bb is the least integer greater than the greatest digit in n.n. For example, if n=72,n = 72, then b=8,b = 8, and 7272 as a numeral in base 88 equals 78+2=58;7 \cdot 8 + 2 = 58; therefore f(72)=58.f(72) = 58. Find the number of positive integers nn less than 10001000 such that f(n)=n.f(n) = n.

Nivel de dificultad: 2300
Pista pequeña:

Los números de un dígito se leen igual en toda base. Para numerales más largos compara los valores posicionales: si b<10,b \lt 10, cada dígito vale menos que en base diez.

One-digit numbers read the same in every base. For longer numerals compare place values: if b<10,b \lt 10, each digit is worth less than in base ten.

Pista grande:

Así que las soluciones de varios dígitos son exactamente los números que contienen un dígito 9;9; cuéntalos con conteo complementario.

So the multi-digit solutions are exactly the numbers containing a digit 9;9; count those with complementary counting.

Solución:

Si nn tiene un solo dígito d,d, entonces el numeral dd tiene valor dd en toda base, así que f(n)=n:f(n) = n: los 99 números de un dígito funcionan. Si nn tiene dígitos dk1d1d0d_{k-1} \ldots d_1 d_0 con k2,k \ge 2, entonces siempre b10b \le 10, y si b<10b \lt 10 entonces f(n)=dibi<di10i=nf(n) = \sum d_i b^i \lt \sum d_i 10^i = n porque el dígito principal cumple dk1bk1<dk110k1.d_{k-1} b^{k-1} \lt d_{k-1} 10^{k-1}. Así que un nn de varios dígitos cumple f(n)=nf(n) = n exactamente cuando b=10,b = 10, es decir, cuando algún dígito de nn es igual a 9.9.

Números de dos dígitos que contienen un 9:9: los números 9090 hasta 9999 más 19,29,,89,19, 29, \ldots, 89, para un total de 10+8=18.10 + 8 = 18. Números de tres dígitos que contienen un 9:9: 900899=252,900 - 8 \cdot 9 \cdot 9 = 252, restando los números sin ningún 99 (dígito principal 118,8, los demás 0088).

El total es 9+18+252=279.9 + 18 + 252 = 279.

If nn has a single digit d,d, then the numeral dd has value dd in every base, so f(n)=n:f(n) = n: all 99 one-digit numbers work. If nn has digits dk1d1d0d_{k-1} \ldots d_1 d_0 with k2,k \ge 2, then b10b \le 10 always, and if b<10b \lt 10 then f(n)=dibi<di10i=nf(n) = \sum d_i b^i \lt \sum d_i 10^i = n because the leading digit satisfies dk1bk1<dk110k1.d_{k-1} b^{k-1} \lt d_{k-1} 10^{k-1}. So a multi-digit nn satisfies f(n)=nf(n) = n exactly when b=10,b = 10, that is, when some digit of nn equals 9.9.

Two-digit numbers containing a 9:9: the numbers 9090 through 9999 plus 19,29,,89,19, 29, \ldots, 89, for 10+8=18.10 + 8 = 18. Three-digit numbers containing a 9:9: 900899=252,900 - 8 \cdot 9 \cdot 9 = 252, subtracting the numbers with no 99 (leading digit 118,8, others 0088).

The total is 9+18+252=279.9 + 18 + 252 = 279.

5.

Una urna contiene nn canicas. Cada canica es roja o azul, y hay al menos 77 canicas de cada color. Cuando se extraen al azar 77 canicas de la urna sin reemplazo, la probabilidad de que exactamente 44 de ellas sean rojas es igual a la probabilidad de que exactamente 55 de ellas sean rojas. Halla la suma de los cinco menores valores de nn para los cuales esto es posible.

An urn contains nn marbles. Each marble is either red or blue, and there are at least 77 marbles of each color. When 77 marbles are drawn randomly from the urn without replacement, the probability that exactly 44 of them are red equals the probability that exactly 55 of them are red. Find the sum of the five least values of nn for which this is possible.

Nivel de dificultad: 2390
Pista pequeña:

Escribe ambas probabilidades con coeficientes binomiales, (r4)(b3)\binom{r}{4}\binom{b}{3} y (r5)(b2),\binom{r}{5}\binom{b}{2}, y cancela los factores comunes.

Write both probabilities with binomial coefficients, (r4)(b3)\binom{r}{4}\binom{b}{3} and (r5)(b2),\binom{r}{5}\binom{b}{2}, and cancel the common factors.

Pista grande:

La ecuación se reduce a 5(b2)=3(r4),5(b - 2) = 3(r - 4), así que r4(mod5);r \equiv 4 \pmod 5; recuerda que ambos colores necesitan al menos 77 canicas.

The equation reduces to 5(b2)=3(r4),5(b - 2) = 3(r - 4), so r4(mod5);r \equiv 4 \pmod 5; remember both colors need at least 77 marbles.

Solución:

Digamos que hay rr canicas rojas y bb azules, r,b7.r, b \ge 7. La condición es (r4)(b3)=(r5)(b2).\binom{r}{4}\binom{b}{3} = \binom{r}{5}\binom{b}{2}. Como (r5)=(r4)r45\binom{r}{5} = \binom{r}{4}\frac{r-4}{5} y (b3)=(b2)b23,\binom{b}{3} = \binom{b}{2}\frac{b-2}{3}, al cancelar se obtiene b23=r45,\frac{b - 2}{3} = \frac{r - 4}{5}, es decir, b=3r25.b = \frac{3r - 2}{5}.

Así que r4(mod5),r \equiv 4 \pmod 5, y b7b \ge 7 requiere 3r235,3r - 2 \ge 35, por lo que r14.r \ge 14. Las cinco elecciones más pequeñas son r=14,19,24,29,34r = 14, 19, 24, 29, 34 con b=8,11,14,17,20,b = 8, 11, 14, 17, 20, que dan n=22,30,38,46,54.n = 22, 30, 38, 46, 54.

La suma es 22+30+38+46+54=190.22 + 30 + 38 + 46 + 54 = 190.

Say there are rr red and bb blue marbles, r,b7.r, b \ge 7. The condition is (r4)(b3)=(r5)(b2).\binom{r}{4}\binom{b}{3} = \binom{r}{5}\binom{b}{2}. Since (r5)=(r4)r45\binom{r}{5} = \binom{r}{4}\frac{r-4}{5} and (b3)=(b2)b23,\binom{b}{3} = \binom{b}{2}\frac{b-2}{3}, cancelling gives b23=r45,\frac{b - 2}{3} = \frac{r - 4}{5}, that is, b=3r25.b = \frac{3r - 2}{5}.

So r4(mod5),r \equiv 4 \pmod 5, and b7b \ge 7 requires 3r235,3r - 2 \ge 35, so r14.r \ge 14. The five smallest choices are r=14,19,24,29,34r = 14, 19, 24, 29, 34 with b=8,11,14,17,20,b = 8, 11, 14, 17, 20, giving n=22,30,38,46,54.n = 22, 30, 38, 46, 54.

The sum is 22+30+38+46+54=190.22 + 30 + 38 + 46 + 54 = 190.

6.

Halla la suma de todos los números reales rr tales que haya al menos un punto donde la circunferencia de radio rr centrada en (4,39)(4, 39) sea tangente a la parábola de ecuación 2y=x28x+12.2y = x^2 - 8x + 12.

Find the sum of all real numbers rr such that there is at least one point where the circle with radius rr centered at (4,39)(4, 39) is tangent to the parabola with equation 2y=x28x+12.2y = x^2 - 8x + 12.

Nivel de dificultad: 2650
Pista pequeña:

El centro (4,39)(4, 39) está sobre el eje de la parábola. Escribe los puntos de la parábola como (4+u, u222)(4 + u,\ \frac{u^2}{2} - 2) y estudia la distancia al cuadrado al centro como función de u.u.

The center (4,39)(4, 39) lies on the parabola’s axis. Write points of the parabola as (4+u, u222)(4 + u,\ \frac{u^2}{2} - 2) and study the squared distance to the center as a function of u.u.

Pista grande:

La tangencia ocurre exactamente donde el radio es normal a la parábola, es decir, en los puntos críticos de la distancia; no olvides el vértice, donde ambas curvas tienen tangentes horizontales.

Tangency happens exactly where the radius is normal to the parabola, i.e. at critical points of the distance; don’t forget the vertex, where both curves have horizontal tangents.

Solución:

Completando el cuadrado, 2y=(x4)24,2y = (x - 4)^2 - 4, así que con u=x4u = x - 4 la parábola es el conjunto de puntos (4+u, u222)\left(4 + u,\ \frac{u^2}{2} - 2\right) y el centro (4,39)(4, 39) está sobre su eje. La circunferencia es tangente a la parábola en un punto exactamente cuando las dos curvas comparten allí una recta tangente, es decir, cuando el radio hasta ese punto es normal a la parábola, lo que ocurre exactamente en los puntos críticos de la distancia al cuadrado D(u)=u2+(u2241)2,D(u)=2u+u(u282)=u(u280). \begin{aligned} &D(u) = u^2 + \left(\frac{u^2}{2} - 41\right)^2, \\ &D'(u) = 2u + u\left(u^2 - 82\right) \\ &\quad {}= u\left(u^2 - 80\right). \end{aligned}

En u=±80:u = \pm\sqrt{80}: D=80+(4041)2=81,D = 80 + (40 - 41)^2 = 81, así que r=9r = 9 (la circunferencia toca la parábola en dos puntos simétricos). En u=0,u = 0, el punto es el vértice (4,2)(4, -2) a distancia 41,41, donde la parábola y la circunferencia de radio 4141 tienen ambas rectas tangentes horizontales, así que r=41r = 41 también funciona.

La suma es 9+41=50.9 + 41 = 50.

Completing the square, 2y=(x4)24,2y = (x - 4)^2 - 4, so with u=x4u = x - 4 the parabola is the set of points (4+u, u222)\left(4 + u,\ \frac{u^2}{2} - 2\right) and the center (4,39)(4, 39) lies on its axis. The circle is tangent to the parabola at a point exactly when the two curves share a tangent line there, i.e. when the radius to that point is normal to the parabola — which happens exactly at critical points of the squared distance D(u)=u2+(u2241)2,D(u)=2u+u(u282)=u(u280). \begin{aligned} &D(u) = u^2 + \left(\frac{u^2}{2} - 41\right)^2, \\ &D'(u) = 2u + u\left(u^2 - 82\right) \\ &\quad {}= u\left(u^2 - 80\right). \end{aligned}

At u=±80:u = \pm\sqrt{80}: D=80+(4041)2=81,D = 80 + (40 - 41)^2 = 81, so r=9r = 9 (the circle touches the parabola at two symmetric points). At u=0,u = 0, the point is the vertex (4,2)(4, -2) at distance 41,41, where the parabola and the circle of radius 4141 both have horizontal tangent lines, so r=41r = 41 also works.

The sum is 9+41=50.9 + 41 = 50.

7.

Se lanza repetidamente un dado estándar y justo de seis caras. Cada vez que el dado muestra 11 o 2,2, Alice recibe una moneda; cada vez que muestra 33 o 4,4, Bob recibe una moneda; y cada vez que muestra 55 o 6,6, Carol recibe una moneda. La probabilidad de que Alice y Bob reciban cada uno al menos dos monedas antes de que Carol reciba alguna moneda se puede escribir como mn,\frac{m}{n}, donde mm y nn son enteros positivos primos entre sí. Halla 100m+n.100m + n.

A standard fair six-sided die is rolled repeatedly. Each time the die reads 11 or 2,2, Alice gets a coin; each time it reads 33 or 4,4, Bob gets a coin; and each time it reads 55 or 6,6, Carol gets a coin. The probability that Alice and Bob each receive at least two coins before Carol receives any coins can be written as mn,\frac{m}{n}, where mm and nn are relatively prime positive integers. Find 100m+n.100m + n.

Nivel de dificultad: 2840
Pista pequeña:

Condiciona sobre que el primer lanzamiento de Carol sea el lanzamiento k+1:k + 1: los kk lanzamientos anteriores son lanzamientos de Alice/Bob que deben incluir al menos dos de cada uno.

Condition on the first Carol roll being roll k+1:k + 1: the kk rolls before it are Alice/Bob rolls that must include at least two of each.

Pista grande:

Dados kk lanzamientos de Alice/Bob, los 2k2^k patrones son todos igualmente probables, y 2k+22k + 2 de ellos tienen a lo sumo un Alice o a lo sumo un Bob. Suma la serie resultante.

Given kk Alice/Bob rolls, all 2k2^k patterns are equally likely, and 2k+22k + 2 of them have at most one Alice or at most one Bob. Sum the resulting series.

Solución:

Cada lanzamiento es un lanzamiento de Alice, de Bob, o de Carol, cada uno con probabilidad 13.\frac{1}{3}. El evento tiene éxito exactamente cuando los lanzamientos anteriores al primer lanzamiento de Carol incluyen al menos dos lanzamientos de Alice y al menos dos de Bob. El primer lanzamiento de Carol es el lanzamiento k+1k + 1 con probabilidad (23)k13,\left(\frac{2}{3}\right)^k \frac{1}{3}, y dado esto, los primeros kk lanzamientos forman una de 2k2^k cadenas de Alice/Bob igualmente probables. Para k3k \ge 3 las cadenas malas, a lo sumo un Alice o a lo sumo un Bob, suman (k+1)+(k+1)=2k+2,(k + 1) + (k + 1) = 2k + 2, y ninguna cadena es mala de ambas formas. Por lo tanto P=k413(23)k(12k+22k). \begin{aligned} P &= \sum_{k \ge 4} \frac{1}{3}\left(\frac{2}{3}\right)^k \\ &\quad {}\cdot \left(1 - \frac{2k + 2}{2^k}\right). \end{aligned}

La primera parte es k4(23)k=1627.\sum_{k \ge 4} \left(\frac{2}{3}\right)^k = \frac{16}{27}. Para la segunda, k0k+13k=1(113)2=94,\sum_{k \ge 0} \frac{k + 1}{3^k} = \frac{1}{(1 - \frac{1}{3})^2} = \frac{9}{4}, así que k42k+23k=2(9412313427)=211108=1154. \begin{aligned} &\sum_{k \ge 4} \frac{2k + 2}{3^k} \\ &= 2\left(\frac{9}{4} - 1 - \frac{2}{3} - \frac{1}{3} - \frac{4}{27}\right) \\ &= 2 \cdot \frac{11}{108} = \frac{11}{54}. \end{aligned}

Por lo tanto P=13(16271154)P = \frac{1}{3}\left(\frac{16}{27} - \frac{11}{54}\right) =132154=754,= \frac{1}{3} \cdot \frac{21}{54} = \frac{7}{54}, y 100m+n=700+54=754.100m + n = 700 + 54 = 754.

Each roll is an Alice roll, a Bob roll, or a Carol roll, each with probability 13.\frac{1}{3}. The event succeeds exactly when the rolls before the first Carol roll include at least two Alice rolls and at least two Bob rolls. The first Carol roll is roll k+1k + 1 with probability (23)k13,\left(\frac{2}{3}\right)^k \frac{1}{3}, and given this, the first kk rolls form one of 2k2^k equally likely Alice/Bob strings. For k3k \ge 3 the bad strings — at most one Alice, or at most one Bob — number (k+1)+(k+1)=2k+2,(k + 1) + (k + 1) = 2k + 2, and no string is bad in both ways. Hence P=k413(23)k(12k+22k). \begin{aligned} P &= \sum_{k \ge 4} \frac{1}{3}\left(\frac{2}{3}\right)^k \\ &\quad {}\cdot \left(1 - \frac{2k + 2}{2^k}\right). \end{aligned}

The first piece is k4(23)k=1627.\sum_{k \ge 4} \left(\frac{2}{3}\right)^k = \frac{16}{27}. For the second, k0k+13k=1(113)2=94,\sum_{k \ge 0} \frac{k + 1}{3^k} = \frac{1}{(1 - \frac{1}{3})^2} = \frac{9}{4}, so k42k+23k=2(9412313427)=211108=1154. \begin{aligned} &\sum_{k \ge 4} \frac{2k + 2}{3^k} \\ &= 2\left(\frac{9}{4} - 1 - \frac{2}{3} - \frac{1}{3} - \frac{4}{27}\right) \\ &= 2 \cdot \frac{11}{108} = \frac{11}{54}. \end{aligned}

Therefore P=13(16271154)P = \frac{1}{3}\left(\frac{16}{27} - \frac{11}{54}\right) =132154=754,= \frac{1}{3} \cdot \frac{21}{54} = \frac{7}{54}, and 100m+n=700+54=754.100m + n = 700 + 54 = 754.

8.

El triángulo isósceles ABC\triangle ABC tiene AB=BC.AB = BC. Sea II el incentro de ABC.\triangle ABC. Los perímetros de ABC\triangle ABC y AIC\triangle AIC están en la razón 125:6,125 : 6, y todos los lados de ambos triángulos tienen longitudes enteras. Halla el valor mínimo posible de AB.AB.

Isosceles triangle ABC\triangle ABC has AB=BC.AB = BC. Let II be the incenter of ABC.\triangle ABC. The perimeters of ABC\triangle ABC and AIC\triangle AIC are in the ratio 125:6,125 : 6, and all the sides of both triangles have integer lengths. Find the minimum possible value of AB.AB.

Nivel de dificultad: 2990
Pista pequeña:

Con AB=BC=aAB = BC = a y AC=b,AC = b, la circunferencia inscrita toca a ACAC en su punto medio, así que AI2=r2+b24;AI^2 = r^2 + \frac{b^2}{4}; la fórmula de Herón simplifica esto a AI2=ab22a+b.AI^2 = \frac{ab^2}{2a + b}.

With AB=BC=aAB = BC = a and AC=b,AC = b, the incircle touches ACAC at its midpoint, so AI2=r2+b24;AI^2 = r^2 + \frac{b^2}{4}; Heron’s formula simplifies this to AI2=ab22a+b.AI^2 = \frac{ab^2}{2a + b}.

Pista grande:

Pon a2a+b=pq\sqrt{\frac{a}{2a + b}} = \frac{p}{q} en forma irreducible; la razón de perímetros se convierte en 125(q22p2)(q+2p)=6q3,125(q^2 - 2p^2)(q + 2p) = 6q^3, y la divisibilidad fija q.q.

Set a2a+b=pq\sqrt{\frac{a}{2a + b}} = \frac{p}{q} in lowest terms; the perimeter ratio becomes 125(q22p2)(q+2p)=6q3,125(q^2 - 2p^2)(q + 2p) = 6q^3, and divisibility pins down q.q.

Solución:

Sea a=AB=BCa = AB = BC y b=AC,b = AC, de modo que s=a+b2.s = a + \frac{b}{2}. La circunferencia inscrita toca a ACAC en su punto medio (la longitud de la tangente desde AA es sa=b2s - a = \frac{b}{2}), así que AI2=CI2=r2+b24.AI^2 = CI^2 = r^2 + \frac{b^2}{4}. Por la fórmula de Herón, r2=(sa)2(sb)s=b242ab2a+b,r^2 = \frac{(s-a)^2(s-b)}{s} = \frac{b^2}{4} \cdot \frac{2a - b}{2a + b}, y por lo tanto AI2=b24(2ab2a+b+1)=ab22a+b. \begin{aligned} AI^2 &= \frac{b^2}{4}\left(\frac{2a - b}{2a + b} + 1\right) \\ &= \frac{ab^2}{2a + b}. \end{aligned} La condición de perímetros es 125(2AI+b)=6(2a+b).125\,(2\,AI + b) = 6\,(2a + b).

Como AIAI es racional, escribamos a2a+b=pq\sqrt{\frac{a}{2a + b}} = \frac{p}{q} en forma irreducible. Entonces aq2=p2(2a+b)aq^2 = p^2(2a + b) obliga a que aa sea divisible por p2;p^2; escribiendo a=mp2a = mp^2 se obtiene b=m(q22p2),b = m(q^2 - 2p^2), 2a+b=mq2,2a + b = mq^2, y AI=mp(q22p2)q.AI = \frac{mp(q^2 - 2p^2)}{q}. La condición de perímetros pierde entonces por completo a mm: 125(q22p2)(2p+q)=6q3.125\,(q^2 - 2p^2)(2p + q) = 6q^3. Como gcd(125,6)=1\gcd(125, 6) = 1, obtenemos que qq es divisible por 5;5; y qq debe ser par, ya que para qq impar ambos factores de la izquierda son impares mientras que el lado derecho es par. Escribiendo q=10wq = 10w y simplificando, (50w2p2)(p+5w)=12w3.(50w^2 - p^2)(p + 5w) = 12w^3. Ambos factores de la izquierda son coprimos con ww (pues gcd(p,q)=1\gcd(p, q) = 1). Si un primo dividiera a w,w, el lado derecho obligaría a que también dividiera su producto, una contradicción; por tanto w=1.w = 1. Entonces (50p2)(p+5)=12(50 - p^2)(p + 5) = 12 tiene la única solución positiva p=7.p = 7.

Así que a=49m,a = 49m, b=2m,b = 2m, y AI=7m5,AI = \frac{7m}{5}, que es un entero exactamente cuando mm es divisible por 5.5. Tomando m=5m = 5 se obtiene ABC\triangle ABC con lados 245,245,10245, 245, 10 y AIC\triangle AIC con lados 7,7,10,7, 7, 10, cuyos perímetros 500500 y 2424 están efectivamente en la razón 125:6.125 : 6. El mínimo posible de ABAB es 245.245.

Let a=AB=BCa = AB = BC and b=AC,b = AC, so s=a+b2.s = a + \frac{b}{2}. The incircle touches ACAC at its midpoint (tangent length from AA is sa=b2s - a = \frac{b}{2}), so AI2=CI2=r2+b24.AI^2 = CI^2 = r^2 + \frac{b^2}{4}. By Heron’s formula, r2=(sa)2(sb)s=b242ab2a+b,r^2 = \frac{(s-a)^2(s-b)}{s} = \frac{b^2}{4} \cdot \frac{2a - b}{2a + b}, and therefore AI2=b24(2ab2a+b+1)=ab22a+b. \begin{aligned} AI^2 &= \frac{b^2}{4}\left(\frac{2a - b}{2a + b} + 1\right) \\ &= \frac{ab^2}{2a + b}. \end{aligned} The perimeter condition is 125(2AI+b)=6(2a+b).125\,(2\,AI + b) = 6\,(2a + b).

Since AIAI is rational, write a2a+b=pq\sqrt{\frac{a}{2a + b}} = \frac{p}{q} in lowest terms. Then aq2=p2(2a+b)aq^2 = p^2(2a + b) forces aa to be divisible by p2;p^2; writing a=mp2a = mp^2 gives b=m(q22p2),b = m(q^2 - 2p^2), 2a+b=mq2,2a + b = mq^2, and AI=mp(q22p2)q.AI = \frac{mp(q^2 - 2p^2)}{q}. The perimeter condition then loses mm entirely: 125(q22p2)(2p+q)=6q3.125\,(q^2 - 2p^2)(2p + q) = 6q^3. Since gcd(125,6)=1\gcd(125, 6) = 1 we get that qq is divisible by 5;5; and qq must be even, since for odd qq both factors on the left are odd while the right side is even. Writing q=10wq = 10w and simplifying, (50w2p2)(p+5w)=12w3.(50w^2 - p^2)(p + 5w) = 12w^3. Both factors on the left are coprime to ww (as gcd(p,q)=1\gcd(p, q) = 1). If a prime divided w,w, the right side would make it divide their product, a contradiction; hence w=1.w = 1. Now (50p2)(p+5)=12(50 - p^2)(p + 5) = 12 has the unique positive solution p=7.p = 7.

So a=49m,a = 49m, b=2m,b = 2m, and AI=7m5,AI = \frac{7m}{5}, which is an integer exactly when mm is divisible by 5.5. Taking m=5m = 5 gives ABC\triangle ABC with sides 245,245,10245, 245, 10 and AIC\triangle AIC with sides 7,7,10,7, 7, 10, whose perimeters 500500 and 2424 are indeed in ratio 125:6.125 : 6. The minimum possible ABAB is 245.245.

9.

Sea SS el valor de la suma infinita 19+199+1999+19999+\frac{1}{9} + \frac{1}{99} + \frac{1}{999} + \frac{1}{9999} + \cdots Halla el resto cuando el mayor entero menor o igual que 10100S10^{100} S se divide entre 1000.1000.

Let SS denote the value of the infinite sum 19+199+1999+19999+\frac{1}{9} + \frac{1}{99} + \frac{1}{999} + \frac{1}{9999} + \cdots Find the remainder when the greatest integer less than or equal to 10100S10^{100} S is divided by 1000.1000.

Nivel de dificultad: 2920
Pista pequeña:

Desarrolla cada 110k1\frac{1}{10^k - 1} como una serie geométrica en 10k10^{-k} y agrupa las potencias iguales de 10:10: el coeficiente de 10n10^{-n} cuenta los divisores de n.n.

Expand each 110k1\frac{1}{10^k - 1} as a geometric series in 10k10^{-k} and collect equal powers of 10:10: the coefficient of 10n10^{-n} counts the divisors of n.n.

Pista grande:

Después de multiplicar por 10100,10^{100}, solo n=98,99,100n = 98, 99, 100 afectan a las últimas tres cifras, y la cola sobrante está entre 00 y 1.1.

After multiplying by 10100,10^{100}, only n=98,99,100n = 98, 99, 100 affect the last three digits, and the leftover tail is between 00 and 1.1.

Solución:

Cada término es 110k1=j110kj,\frac{1}{10^k - 1} = \sum_{j \ge 1} 10^{-kj}, así que sumando sobre kk y agrupando por el exponente n=kjn = kj, obtenemos S=n1d(n)10n,S = \sum_{n \ge 1} \frac{d(n)}{10^n}, donde d(n)d(n) es el número de divisores de n.n. Por lo tanto 10100S=n=1100d(n)10100n10^{100} S = \sum_{n = 1}^{100} d(n)\,10^{100 - n} +T+ T con T=m1d(100+m)10m.T = \sum_{m \ge 1} d(100 + m)\,10^{-m}.

A partir de d(101)=2,d(101) = 2, d(102)=8,d(102) = 8, d(103)=2,d(103) = 2, d(104)=8,d(104) = 8, la cola empieza 0.20.2 +0.08+ 0.08 +0.002+ 0.002 +0.0008=0.2828,+ 0.0008 = 0.2828, y como d(N)<2N,d(N) \lt 2\sqrt{N}, los términos restantes aportan menos de m52100+m10m<0.001.\sum_{m \ge 5} \frac{2\sqrt{100 + m}}{10^m} \lt 0.001. Así que 0<T<10 \lt T \lt 1 y 10100S=n=1100d(n)10100n.\left\lfloor 10^{100} S \right\rfloor = \sum_{n = 1}^{100} d(n)\,10^{100 - n}.

Módulo 1000,1000, todo término con n97n \le 97 es un múltiplo de 1000,1000, dejando d(98)100+d(99)10d(98) \cdot 100 + d(99) \cdot 10 +d(100).+ d(100). Como d(98)=6,d(98) = 6, d(99)=6,d(99) = 6, y d(100)=9,d(100) = 9, el resto es 600+60+9=669.600 + 60 + 9 = 669.

Each term is 110k1=j110kj,\frac{1}{10^k - 1} = \sum_{j \ge 1} 10^{-kj}, so summing over kk and collecting the exponent n=kj,n = kj, S=n1d(n)10n,S = \sum_{n \ge 1} \frac{d(n)}{10^n}, where d(n)d(n) is the number of divisors of n.n. Hence 10100S=n=1100d(n)10100n10^{100} S = \sum_{n = 1}^{100} d(n)\,10^{100 - n} +T+ T with T=m1d(100+m)10m.T = \sum_{m \ge 1} d(100 + m)\,10^{-m}.

From d(101)=2,d(101) = 2, d(102)=8,d(102) = 8, d(103)=2,d(103) = 2, d(104)=8,d(104) = 8, the tail starts 0.20.2 +0.08+ 0.08 +0.002+ 0.002 +0.0008=0.2828,+ 0.0008 = 0.2828, and since d(N)<2N,d(N) \lt 2\sqrt{N}, the remaining terms contribute less than m52100+m10m<0.001.\sum_{m \ge 5} \frac{2\sqrt{100 + m}}{10^m} \lt 0.001. So 0<T<10 \lt T \lt 1 and 10100S=n=1100d(n)10100n.\left\lfloor 10^{100} S \right\rfloor = \sum_{n = 1}^{100} d(n)\,10^{100 - n}.

Modulo 1000,1000, every term with n97n \le 97 is a multiple of 1000,1000, leaving d(98)100+d(99)10d(98) \cdot 100 + d(99) \cdot 10 +d(100).+ d(100). Since d(98)=6,d(98) = 6, d(99)=6,d(99) = 6, and d(100)=9,d(100) = 9, the remainder is 600+60+9=669.600 + 60 + 9 = 669.

10.

Sea ABC\triangle ABC un triángulo con DD en BC\overline{BC} tal que AD\overline{AD} biseca BAC.\angle BAC. Sea ω\omega la circunferencia que pasa por AA y es tangente al segmento BC\overline{BC} en D.D. Sean EAE \ne A y FAF \ne A las intersecciones de ω\omega con los segmentos AB\overline{AB} y AC,\overline{AC}, respectivamente. Supongamos que AB=200,AB = 200, AC=225,AC = 225, y que todos AE,AE, AF,AF, BD,BD, y CDCD son enteros positivos. Halla el mayor valor posible de BC.BC.

Let ABC\triangle ABC be a triangle with DD on BC\overline{BC} such that AD\overline{AD} bisects BAC.\angle BAC. Let ω\omega be the circle that passes through AA and is tangent to segment BC\overline{BC} at D.D. Let EAE \ne A and FAF \ne A be the intersections of ω\omega with segments AB\overline{AB} and AC,\overline{AC}, respectively. Suppose that AB=200,AB = 200, AC=225,AC = 225, and all of AE,AE, AF,AF, BD,BD, and CDCD are positive integers. Find the greatest possible value of BC.BC.

Nivel de dificultad: 2840
Pista pequeña:

La tangencia en DD da las potencias de BB y C:C: BD2=BEBABD^2 = BE \cdot BA y CD2=CFCA;CD^2 = CF \cdot CA; la bisectriz da BD:DC=8:9.BD : DC = 8 : 9.

Tangency at DD gives the powers of BB and C:C: BD2=BEBABD^2 = BE \cdot BA and CD2=CFCA;CD^2 = CF \cdot CA; the bisector gives BD:DC=8:9.BD : DC = 8 : 9.

Pista grande:

Escribe BD=8t,BD = 8t, CD=9t;CD = 9t; entonces t=CDBDt = CD - BD es un entero, y que AEAE y AFAF sean enteros obliga a que tt sea divisible por 5,5, mientras que AE>0AE \gt 0 acota t.t.

Write BD=8t,BD = 8t, CD=9t;CD = 9t; then t=CDBDt = CD - BD is an integer, and integrality of AEAE and AFAF forces tt to be divisible by 5,5, while AE>0AE \gt 0 bounds t.t.

Solución:

Como ω\omega es tangente a BCBC en D,D, la potencia de BB da BD2=BEBABD^2 = BE \cdot BA y la potencia de CC da CD2=CFCA.CD^2 = CF \cdot CA. La bisectriz del ángulo da BDDC=ABAC=89,\frac{BD}{DC} = \frac{AB}{AC} = \frac{8}{9}, así que BD=8tBD = 8t y CD=9t,CD = 9t, donde t=CDBDt = CD - BD es un entero positivo. Entonces BE=64t2200=8t225,CF=81t2225=9t225, \begin{aligned} &BE = \frac{64t^2}{200} = \frac{8t^2}{25}, \\ &CF = \frac{81t^2}{225} = \frac{9t^2}{25}, \end{aligned} así que AE=2008t225AE = 200 - \frac{8t^2}{25} y AF=2259t225.AF = 225 - \frac{9t^2}{25}.

Para que AEAE y AFAF sean enteros necesitamos que t2t^2 sea divisible por 25,25, es decir, t=5s.t = 5s. Entonces AE=2008s2>0AE = 200 - 8s^2 \gt 0 obliga a s4,s \le 4, y BC=17t=85s.BC = 17t = 85s. Con s=4:s = 4: BC=340,BC = 340, con BD=160,BD = 160, CD=180,CD = 180, AE=72,AE = 72, AF=81AF = 81 todos enteros positivos, y los lados 200,225,340200, 225, 340 forman un triángulo válido puesto que 200+225>340.200 + 225 \gt 340.

El mayor valor posible de BCBC es 340.340.

Since ω\omega is tangent to BCBC at D,D, the power of BB gives BD2=BEBABD^2 = BE \cdot BA and the power of CC gives CD2=CFCA.CD^2 = CF \cdot CA. The angle bisector gives BDDC=ABAC=89,\frac{BD}{DC} = \frac{AB}{AC} = \frac{8}{9}, so BD=8tBD = 8t and CD=9t,CD = 9t, where t=CDBDt = CD - BD is a positive integer. Then BE=64t2200=8t225,CF=81t2225=9t225, \begin{aligned} &BE = \frac{64t^2}{200} = \frac{8t^2}{25}, \\ &CF = \frac{81t^2}{225} = \frac{9t^2}{25}, \end{aligned} so AE=2008t225AE = 200 - \frac{8t^2}{25} and AF=2259t225.AF = 225 - \frac{9t^2}{25}.

For AEAE and AFAF to be integers we need t2t^2 to be divisible by 25,25, that is, t=5s.t = 5s. Then AE=2008s2>0AE = 200 - 8s^2 \gt 0 forces s4,s \le 4, and BC=17t=85s.BC = 17t = 85s. At s=4:s = 4: BC=340,BC = 340, with BD=160,BD = 160, CD=180,CD = 180, AE=72,AE = 72, AF=81AF = 81 all positive integers, and the sides 200,225,340200, 225, 340 form a valid triangle since 200+225>340.200 + 225 \gt 340.

The greatest possible value of BCBC is 340.340.

11.

Halla el mayor entero nn tal que el polinomio cúbico x3n6x2+(n11)x400x^3 - \frac{n}{6}x^2 + (n - 11)x - 400 tenga raíces α2,\alpha^2, β2,\beta^2, y γ2,\gamma^2, donde α,\alpha, β,\beta, y γ\gamma son números complejos, y haya exactamente siete valores posibles distintos para α+β+γ.\alpha + \beta + \gamma.

Find the greatest integer nn such that the cubic polynomial x3n6x2+(n11)x400x^3 - \frac{n}{6}x^2 + (n - 11)x - 400 has roots α2,\alpha^2, β2,\beta^2, and γ2,\gamma^2, where α,\alpha, β,\beta, and γ\gamma are complex numbers, and there are exactly seven different possible values for α+β+γ.\alpha + \beta + \gamma.

Nivel de dificultad: 3060
Pista pequeña:

Fija raíces cuadradas s1,s2,s3s_1, s_2, s_3 de las tres raíces; las sumas posibles son ±s1±s2±s3,\pm s_1 \pm s_2 \pm s_3, ocho expresiones que se emparejan como ±v.\pm v.

Fix square roots s1,s2,s3s_1, s_2, s_3 of the three roots; the possible sums are ±s1±s2±s3,\pm s_1 \pm s_2 \pm s_3, eight expressions that pair off as ±v.\pm v.

Pista grande:

Aparecen exactamente siete valores cuando una combinación es 0;0; el producto de las cuatro expresiones s1±s2±s3s_1 \pm s_2 \pm s_3 es igual a 4e2e124e_2 - e_1^2 en las funciones simétricas de las raíces.

Exactly seven values occur when one combination is 0;0; the product of the four expressions s1±s2±s3s_1 \pm s_2 \pm s_3 equals 4e2e124e_2 - e_1^2 in the roots’ symmetric functions.

Solución:

Las raíces de la cúbica son α2,β2,γ2.\alpha^2, \beta^2, \gamma^2. Fijemos raíces cuadradas s1,s2,s3s_1, s_2, s_3; entonces α+β+γ\alpha + \beta + \gamma recorre las ocho expresiones ±s1±s2±s3,\pm s_1 \pm s_2 \pm s_3, que forman cuatro pares ±v.\pm v. En general, las ocho son distintas. El producto de las tres raíces cuadradas elegidas es no nulo porque el término constante de la cúbica es 400,-400, así que cada sis_i es no nulo. Una coincidencia v(ε)=v(ε)v(\varepsilon) = v(\varepsilon') entre elecciones no opuestas debe diferir en dos signos y obliga a si=±sjs_i = \pm s_j para algún ij,i \ne j, lo que reduce los ocho valores a lo sumo a seis. Por tanto, aparecen exactamente siete valores precisamente cuando alguna elección satisface ±s1±s2±s3=0\pm s_1 \pm s_2 \pm s_3 = 0 —su opuesta da entonces el mismo valor 00— y no hay más degeneraciones.

Esta condición es que se anule (s1+s2+s3)(s1+s2+s3)(s1s2+s3)(s1+s2s3)=2i<jrirjiri2=4e2e12, \begin{aligned} &(s_1 + s_2 + s_3)(-s_1 + s_2 + s_3) \\ &\quad {}\cdot (s_1 - s_2 + s_3)(s_1 + s_2 - s_3) \\ &= 2\sum_{i \lt j} r_i r_j - \sum_i r_i^2 \\ &= 4e_2 - e_1^2, \end{aligned} donde ri=si2r_i = s_i^2 son las raíces y e1,e2e_1, e_2 sus funciones simétricas elementales. Por las fórmulas de Vieta, e1=n6e_1 = \frac{n}{6} y e2=n11,e_2 = n - 11, así que n236=4(n11),\frac{n^2}{36} = 4(n - 11), es decir, n2144n+1584=0,n^2 - 144n + 1584 = 0, con raíces n=12n = 12 y n=132.n = 132.

Para n=132n = 132, la cúbica se factoriza como (x16)(x26x+25),(x - 16)(x^2 - 6x + 25), con raíces distintas 16,16, 3+4i,3 + 4i, y 34i.3 - 4i. Elijamos sus raíces cuadradas como 4,4, 2+i,2 + i, y 2i;2 - i; entonces 4(2+i)(2i)=0.4 - (2+i) - (2-i) = 0. Las ocho elecciones de signos dan 00 dos veces y los seis valores no nulos distintos ±8,\pm 8, ±(4+2i),\pm(4+2i), ±(42i),\pm(4-2i), así que aparecen exactamente siete sumas. El mayor entero de este tipo es 132.132.

The roots of the cubic are α2,β2,γ2.\alpha^2, \beta^2, \gamma^2. Fix square roots s1,s2,s3s_1, s_2, s_3 of them; then α+β+γ\alpha + \beta + \gamma ranges over the eight expressions ±s1±s2±s3,\pm s_1 \pm s_2 \pm s_3, which come in four pairs ±v.\pm v. Generically all eight are distinct. The product of the three chosen square roots is nonzero because the cubic’s constant term is 400,-400, so every sis_i is nonzero. A coincidence v(ε)=v(ε)v(\varepsilon) = v(\varepsilon') between choices that are not opposite must differ in two signs and forces si=±sjs_i = \pm s_j for some ij,i \ne j, which collapses the eight values to at most six. So exactly seven values occur precisely when one choice satisfies ±s1±s2±s3=0\pm s_1 \pm s_2 \pm s_3 = 0 — its opposite is then the same value 00 — and no further degeneracies occur.

That condition is the vanishing of (s1+s2+s3)(s1+s2+s3)(s1s2+s3)(s1+s2s3)=2i<jrirjiri2=4e2e12, \begin{aligned} &(s_1 + s_2 + s_3)(-s_1 + s_2 + s_3) \\ &\quad {}\cdot (s_1 - s_2 + s_3)(s_1 + s_2 - s_3) \\ &= 2\sum_{i \lt j} r_i r_j - \sum_i r_i^2 \\ &= 4e_2 - e_1^2, \end{aligned} where ri=si2r_i = s_i^2 are the roots and e1,e2e_1, e_2 their elementary symmetric functions. By Vieta’s formulas e1=n6e_1 = \frac{n}{6} and e2=n11,e_2 = n - 11, so n236=4(n11),\frac{n^2}{36} = 4(n - 11), i.e. n2144n+1584=0,n^2 - 144n + 1584 = 0, with roots n=12n = 12 and n=132.n = 132.

For n=132n = 132 the cubic factors as (x16)(x26x+25),(x - 16)(x^2 - 6x + 25), with distinct roots 16,16, 3+4i,3 + 4i, and 34i.3 - 4i. Choose their square roots as 4,4, 2+i,2 + i, and 2i;2 - i; then 4(2+i)(2i)=0.4 - (2+i) - (2-i) = 0. The eight sign choices give 00 twice and the six distinct nonzero values ±8,\pm 8, ±(4+2i),\pm(4+2i), ±(42i),\pm(4-2i), so exactly seven sums occur. The greatest such integer is 132.132.

12.

Considera un tetraedro con dos caras triangulares isósceles de lados 510,5\sqrt{10}, 510,5\sqrt{10}, y 1010 y dos caras triangulares isósceles de lados 510,5\sqrt{10}, 510,5\sqrt{10}, y 18.18. Los cuatro vértices del tetraedro están sobre una esfera de centro S,S, y las cuatro caras del tetraedro son tangentes a una esfera de centro R.R. La distancia RSRS se puede escribir como mn,\frac{m}{n}, donde mm y nn son enteros positivos primos entre sí. Halla m+n.m + n.

Consider a tetrahedron with two isosceles triangle faces with side lengths 510,5\sqrt{10}, 510,5\sqrt{10}, and 1010 and two isosceles triangle faces with side lengths 510,5\sqrt{10}, 510,5\sqrt{10}, and 18.18. The four vertices of the tetrahedron lie on a sphere with center S,S, and the four faces of the tetrahedron are tangent to a sphere with center R.R. The distance RSRS can be written as mn,\frac{m}{n}, where mm and nn are relatively prime positive integers. Find m+n.m + n.

Nivel de dificultad: 2990
Pista pequeña:

Las aristas son 1010 y 1818 opuestas entre sí más cuatro de 510.5\sqrt{10}. Coloca esas dos aristas sobre ejes horizontales perpendiculares; ambos centros están sobre el eje de simetría vertical.

The edges are 1010 and 1818 opposite each other plus four of 510.5\sqrt{10}. Place those two edges on perpendicular horizontal axes; both centers lie on the vertical symmetry axis.

Pista grande:

Las dos aristas están separadas 1212 verticalmente. Halla cada centro sobre el eje: distancias iguales a los vértices para S,S, distancias iguales a dos caras adyacentes para R.R.

The two edges sit 1212 apart vertically. Find each center on the axis: equal distances to vertices for S,S, equal distances to two adjacent faces for R.R.

Solución:

Las cuatro caras tienen el multiconjunto de lados {510×8, 10×2, 18×2},\{5\sqrt{10} \times 8,\ 10 \times 2,\ 18 \times 2\}, y cada arista está en dos caras, así que el tetraedro ABCDABCD tiene AB=10AB = 10 y CD=18CD = 18 como aristas opuestas y las otras cuatro aristas iguales a 510.5\sqrt{10}. Colocamos A=(5,0,12),B=(5,0,12),C=(0,9,0),D=(0,9,0), \begin{aligned} &A = (-5, 0, 12), \\ &B = (5, 0, 12), \\ &C = (0, -9, 0), \\ &D = (0, 9, 0), \end{aligned} lo cual es consistente puesto que AC2=25+81+144AC^2 = 25 + 81 + 144 =250= 250 =(510)2.= (5\sqrt{10})^2. La configuración es simétrica bajo xxx \to -x y bajo yy,y \to -y, así que ambos centros están sobre el eje zz.

Para S=(0,0,s),S = (0, 0, s), igualar las distancias a AA y a CC da 25+(12s)2=81+s2,25 + (12 - s)^2 = 81 + s^2, así que s=113.s = \frac{11}{3}. Para R=(0,0,t),R = (0, 0, t), la cara ABCABC tiene plano 4y3z+36=04y - 3z + 36 = 0 y la cara ACDACD tiene plano 12x+5z=0,12x + 5z = 0, así que las distancias iguales requieren 363t5=5t13    t=11716,\frac{36 - 3t}{5} = \frac{5t}{13} \implies t = \frac{117}{16}, y por las dos simetrías especulares este punto es equidistante (a distancia 4516\frac{45}{16}) de las cuatro caras.

Por lo tanto RS=11716113RS = \frac{117}{16} - \frac{11}{3} =35117648=17548,= \frac{351 - 176}{48} = \frac{175}{48}, que ya es irreducible, así que m+n=175+48=223.m + n = 175 + 48 = 223.

The four faces have side multiset {510×8, 10×2, 18×2},\{5\sqrt{10} \times 8,\ 10 \times 2,\ 18 \times 2\}, and each edge lies on two faces, so the tetrahedron ABCDABCD has AB=10AB = 10 and CD=18CD = 18 as opposite edges and the other four edges equal to 510.5\sqrt{10}. Place A=(5,0,12),B=(5,0,12),C=(0,9,0),D=(0,9,0), \begin{aligned} &A = (-5, 0, 12), \\ &B = (5, 0, 12), \\ &C = (0, -9, 0), \\ &D = (0, 9, 0), \end{aligned} which is consistent since AC2=25+81+144AC^2 = 25 + 81 + 144 =250= 250 =(510)2.= (5\sqrt{10})^2. The configuration is symmetric under xxx \to -x and under yy,y \to -y, so both centers lie on the zz-axis.

For S=(0,0,s),S = (0, 0, s), equating distances to AA and CC gives 25+(12s)2=81+s2,25 + (12 - s)^2 = 81 + s^2, so s=113.s = \frac{11}{3}. For R=(0,0,t),R = (0, 0, t), face ABCABC has plane 4y3z+36=04y - 3z + 36 = 0 and face ACDACD has plane 12x+5z=0,12x + 5z = 0, so equal distances require 363t5=5t13    t=11716,\frac{36 - 3t}{5} = \frac{5t}{13} \implies t = \frac{117}{16}, and by the two mirror symmetries this point is equidistant (at distance 4516\frac{45}{16}) from all four faces.

Therefore RS=11716113RS = \frac{117}{16} - \frac{11}{3} =35117648=17548,= \frac{351 - 176}{48} = \frac{175}{48}, which is in lowest terms, so m+n=175+48=223.m + n = 175 + 48 = 223.

13.

Llamamos primos a dos conjuntos finitos de enteros SS y TT si

SS y TT tienen el mismo número de elementos,

SS y TT son disjuntos, y

• los elementos de SS se pueden emparejar con los elementos de TT de modo que los dos elementos de cada par difieran exactamente en 1.1.

Por ejemplo, {1,2,5}\{1, 2, 5\} y {0,3,4}\{0, 3, 4\} son primos. Supongamos que el conjunto SS tiene exactamente 40404040 primos. Halla el menor número de elementos que puede tener el conjunto SS.

Call finite sets of integers SS and TT cousins if

SS and TT have the same number of elements,

SS and TT are disjoint, and

• the elements of SS can be paired with the elements of TT so that the elements in each pair differ by exactly 1.1.

For example, {1,2,5}\{1, 2, 5\} and {0,3,4}\{0, 3, 4\} are cousins. Suppose that the set SS has exactly 40404040 cousins. Find the least number of elements the set SS can have.

Nivel de dificultad: 3370
Pista pequeña:

Cada elemento debe moverse en ±1\pm 1 a un valor fuera de S,S, de forma inyectiva. Un elemento con ambos vecinos en SS queda atascado, así que SS consta de bloques de 11 o 22 enteros consecutivos.

Each element must move by ±1\pm 1 to a value outside S,S, injectively. An element with both neighbors in SS is stuck, so SS consists of blocks of 11 or 22 consecutive integers.

Pista grande:

Los bloques separados por un único entero faltante se encadenan: una cadena de kk bloques unitarios tiene exactamente k+1k + 1 imágenes, y las cadenas separadas se multiplican. Factoriza 40404040 para minimizar el tamaño total.

Blocks separated by a single missing integer chain together: a chain of kk singleton blocks has exactly k+1k + 1 images, and separate chains multiply. Factor 40404040 to minimize total size.

Solución:

Un conjunto primo TT es la imagen de una inyección que envía cada xSx \in S a x1x - 1 o x+1,x + 1, cayendo fuera de S.S. Si x1,x,x+1Sx - 1, x, x + 1 \in S entonces xx no tiene a dónde ir, así que todo bloque maximal de elementos consecutivos de SS tiene tamaño 11 o 2.2. Un bloque doble {a,a+1}\{a, a+1\} está obligado a aplicarse en {a1,a+2},\{a - 1, a + 2\}, mientras que un bloque unitario {a}\{a\} elige a1a - 1 o a+1.a + 1. Dos bloques solo pueden disputarse un valor cuando exactamente un entero los separa, así que agrupamos los bloques en cadenas: bloques consecutivos con huecos de exactamente uno. Dentro de una cadena los únicos patrones consistentes son “los primeros ii bloques se desplazan a la izquierda y el resto a la derecha”, ya que un bloque que elige la derecha y su sucesor que elige la izquierda colisionarían; un bloque doble actúa a la vez como izquierda y derecha, forzando a que el cambio ocurra exactamente en él. Por lo tanto una cadena de kk unitarios produce k+1k + 1 imágenes distintas, una cadena que contiene un doble produce exactamente 1,1, y una cadena con dos dobles produce 0.0. Patrones distintos dan conjuntos distintos T,T, y las elecciones en cadenas diferentes son independientes, así que el número de primos es el producto de (ki+1)(k_i + 1) sobre las cadenas totalmente unitarias.

Necesitamos (ki+1)=4040=235101\prod (k_i + 1) = 4040 = 2^3 \cdot 5 \cdot 101 minimizando a la vez el número de elementos ki\sum k_i (las cadenas con dobles solo desperdician elementos). Reemplazar un factor compuesto f=ghf = gh por los dos factores g,h2g, h \ge 2 reduce estrictamente el costo, porque (g1)+(h1)<gh1.(g - 1) + (h - 1) \lt gh - 1. Así que el óptimo usa la factorización en primos: (f1)=1+1+1+4+100=107, \begin{aligned} &\sum (f - 1) \\ &= 1 + 1 + 1 + 4 + 100 \\ &= 107, \end{aligned} realizada por cinco cadenas de 1,1,1,4,1001, 1, 1, 4, 100 unitarios —series de enteros alternos— colocadas muy separadas.

El menor número posible de elementos es 107.107.

A cousin TT is the image of an injection sending each xSx \in S to x1x - 1 or x+1,x + 1, landing outside S.S. If x1,x,x+1Sx - 1, x, x + 1 \in S then xx has nowhere to go, so every maximal block of consecutive elements of SS has size 11 or 2.2. A double block {a,a+1}\{a, a+1\} is forced to map to {a1,a+2},\{a - 1, a + 2\}, while a singleton {a}\{a\} chooses a1a - 1 or a+1.a + 1. Two blocks can fight over a value only when exactly one integer separates them, so group blocks into chains: consecutive blocks with gaps of exactly one. Within a chain the only consistent patterns are “the first ii blocks shift left and the rest shift right,” since a block choosing right and its successor choosing left would collide; a double block acts as both left and right, forcing the switch to happen exactly at it. Hence a chain of kk singletons produces k+1k + 1 distinct images, a chain containing one double produces exactly 1,1, and a chain with two doubles produces 0.0. Distinct patterns give distinct sets T,T, and choices in different chains are independent, so the number of cousins is the product of (ki+1)(k_i + 1) over the all-singleton chains.

We need (ki+1)=4040=235101\prod (k_i + 1) = 4040 = 2^3 \cdot 5 \cdot 101 while minimizing the element count ki\sum k_i (chains with doubles only waste elements). Replacing a composite factor f=ghf = gh with the two factors g,h2g, h \ge 2 strictly lowers the cost, because (g1)+(h1)<gh1.(g - 1) + (h - 1) \lt gh - 1. So the optimum uses the prime factorization: (f1)=1+1+1+4+100=107, \begin{aligned} &\sum (f - 1) \\ &= 1 + 1 + 1 + 4 + 100 \\ &= 107, \end{aligned} realized by five chains of 1,1,1,4,1001, 1, 1, 4, 100 singletons — runs of every-other integer — placed far apart.

The least possible number of elements is 107.107.

14.

Para enteros aa y b,b, definimos ab=aba \circ b = a - b si aa es impar y bb es par, y ab=a+ba \circ b = a + b en caso contrario. Halla el número de sucesiones a1,a_1, a2,a_2, a3,a_3, ,\ldots, ana_n de enteros positivos tales que a1+a2+a3++an=12a_1 + a_2 + a_3 + \cdots + a_n = 12 y a1a2a3an=0,a_1 \circ a_2 \circ a_3 \circ \cdots \circ a_n = 0, donde las operaciones se realizan de izquierda a derecha; es decir, a1a2a3a_1 \circ a_2 \circ a_3 significa (a1a2)a3.(a_1 \circ a_2) \circ a_3.

For integers aa and b,b, let ab=aba \circ b = a - b if aa is odd and bb is even, and ab=a+ba \circ b = a + b otherwise. Find the number of sequences a1,a_1, a2,a_2, a3,a_3, ,\ldots, ana_n of positive integers such that a1+a2+a3++an=12a_1 + a_2 + a_3 + \cdots + a_n = 12 and a1a2a3an=0,a_1 \circ a_2 \circ a_3 \circ \cdots \circ a_n = 0, where the operations are performed from left to right; that is, a1a2a3a_1 \circ a_2 \circ a_3 means (a1a2)a3.(a_1 \circ a_2) \circ a_3.

Nivel de dificultad: 3370
Pista pequeña:

El valor acumulado siempre tiene la misma paridad que la suma acumulada, así que un término se resta exactamente cuando es par y la suma de los términos anteriores es impar.

The running value always has the same parity as the running sum, so a term is subtracted exactly when it is even and the sum of the terms before it is odd.

Pista grande:

Los términos restados deben sumar 6.6. La paridad de la suma prefija solo cambia en los términos impares, así que divide la sucesión en sus términos impares (hay 2,2, 4,4, o 66) y cuenta cada caso.

The subtracted terms must total 6.6. Prefix-sum parity flips only at odd terms, so split the sequence at its odd terms (there are 2,2, 4,4, or 66) and count each case.

Solución:

Como aba+b(mod2),a - b \equiv a + b \pmod 2, el valor acumulado tras kk pasos tiene la misma paridad que a1++ak.a_1 + \cdots + a_k. Así que el término aka_k se resta exactamente cuando aka_k es par y la suma prefija a1++ak1a_1 + \cdots + a_{k-1} es impar, y el valor final es 1212 menos el doble del total de los términos restados. Debemos contar las composiciones de 1212 en las que los términos pares situados donde la suma prefija es impar suman exactamente 6.6. La paridad prefija cambia exactamente en los términos impares, así que los términos impares vienen en número 2m2m (el total es par), y los términos restados son precisamente los términos pares que están entre el (2i1)(2i-1)-ésimo y el 2i2i-ésimo términos impares; estos mm “tramos impares” deben contener términos pares que sumen 6,6, mientras que los otros m+1m + 1 tramos contienen términos pares que suman 6A,6 - A, donde AA es la suma de los términos impares.

Sea fr(t)f_r(t) el número de maneras de llenar rr tramos ordenados con sucesiones de términos pares que suman 2t.2t. Un tramo es una composición de 2t2t en partes pares, es decir, de t:t: f1(t)=2t1f_1(t) = 2^{t-1} para t1t \ge 1 y f1(0)=1;f_1(0) = 1; al convolucionar se obtienen los valores necesarios más abajo: fr(0)=1,f_r(0) = 1, fr(1)=r,f_r(1) = r, f2(2)=5,f_2(2) = 5, y f1(3),f2(3),f3(3)=4,12,25.f_1(3), f_2(3), f_3(3) = 4, 12, 25. Las composiciones de AA en 2m2m partes impares son en número (A2m2+2m12m1).\binom{\frac{A - 2m}{2} + 2m - 1}{2m - 1}.

Análisis por casos según mm y A:A: para m=1:m = 1: A=2,4,6A = 2, 4, 6 dan 14f2(2)=20,1 \cdot 4 \cdot f_2(2) = 20, 24f2(1)=16,2 \cdot 4 \cdot f_2(1) = 16, y 341=12.3 \cdot 4 \cdot 1 = 12. Para m=2:m = 2: A=4,6A = 4, 6 dan 112f3(1)=361 \cdot 12 \cdot f_3(1) = 36 y 4121=48.4 \cdot 12 \cdot 1 = 48. Para m=3:m = 3: A=6A = 6 da 1251=25.1 \cdot 25 \cdot 1 = 25. El total es 20+16+12+3620 + 16 + 12 + 36 +48+25=157.+ 48 + 25 = 157.

Since aba+b(mod2),a - b \equiv a + b \pmod 2, the running value after kk steps has the same parity as a1++ak.a_1 + \cdots + a_k. So term aka_k is subtracted exactly when aka_k is even and the prefix sum a1++ak1a_1 + \cdots + a_{k-1} is odd, and the final value is 1212 minus twice the total of the subtracted terms. We must count compositions of 1212 in which the even terms sitting where the prefix sum is odd total exactly 6.6. The prefix parity flips exactly at odd terms, so the odd terms come in 2m2m (the total is even), and the subtracted terms are precisely the even terms lying between the (2i1)(2i-1)st and 2i2ith odd terms; these mm “odd stretches” must hold even terms totaling 6,6, while the other m+1m + 1 stretches hold even terms totaling 6A,6 - A, where AA is the sum of the odd terms.

Let fr(t)f_r(t) be the number of ways to fill rr ordered stretches with sequences of even terms totaling 2t.2t. One stretch is a composition of 2t2t into even parts, i.e. of t:t: f1(t)=2t1f_1(t) = 2^{t-1} for t1t \ge 1 and f1(0)=1;f_1(0) = 1; convolving gives the values needed below: fr(0)=1,f_r(0) = 1, fr(1)=r,f_r(1) = r, f2(2)=5,f_2(2) = 5, and f1(3),f2(3),f3(3)=4,12,25.f_1(3), f_2(3), f_3(3) = 4, 12, 25. Compositions of AA into 2m2m odd parts number (A2m2+2m12m1).\binom{\frac{A - 2m}{2} + 2m - 1}{2m - 1}.

Casework on mm and A:A: for m=1:m = 1: A=2,4,6A = 2, 4, 6 give 14f2(2)=20,1 \cdot 4 \cdot f_2(2) = 20, 24f2(1)=16,2 \cdot 4 \cdot f_2(1) = 16, and 341=12.3 \cdot 4 \cdot 1 = 12. For m=2:m = 2: A=4,6A = 4, 6 give 112f3(1)=361 \cdot 12 \cdot f_3(1) = 36 and 4121=48.4 \cdot 12 \cdot 1 = 48. For m=3:m = 3: A=6A = 6 gives 1251=25.1 \cdot 25 \cdot 1 = 25. The total is 20+16+12+3620 + 16 + 12 + 36 +48+25=157.+ 48 + 25 = 157.

15.

Halla el número de 77-tuplas ordenadas (a1,a2,a3,,a7)(a_1, a_2, a_3, \ldots, a_7) que tienen las siguientes propiedades:

ak{1,2,3}a_k \in \{1, 2, 3\} para todo k.k.

a1+a2+a3+a4a_1 + a_2 + a_3 + a_4 +a5+a6+a7+ a_5 + a_6 + a_7 es un múltiplo de 3.3.

a1a2a4a_1a_2a_4 +a2a3a5+ a_2a_3a_5 +a3a4a6+ a_3a_4a_6 +a4a5a7+ a_4a_5a_7 +a5a6a1+ a_5a_6a_1 +a6a7a2+ a_6a_7a_2 +a7a1a3+ a_7a_1a_3 es un múltiplo de 3.3.

Find the number of ordered 77-tuples (a1,a2,a3,,a7)(a_1, a_2, a_3, \ldots, a_7) having the following properties:

ak{1,2,3}a_k \in \{1, 2, 3\} for all k.k.

a1+a2+a3+a4a_1 + a_2 + a_3 + a_4 +a5+a6+a7+ a_5 + a_6 + a_7 is a multiple of 3.3.

a1a2a4a_1a_2a_4 +a2a3a5+ a_2a_3a_5 +a3a4a6+ a_3a_4a_6 +a4a5a7+ a_4a_5a_7 +a5a6a1+ a_5a_6a_1 +a6a7a2+ a_6a_7a_2 +a7a1a3+ a_7a_1a_3 is a multiple of 3.3.

Nivel de dificultad: 3500
Pista pequeña:

Trabaja módulo 33 con 303 \equiv 0 y 21.2 \equiv -1. Las siete ternas de índices {i,i+1,i+3}\{i, i+1, i+3\} forman un plano de Fano: cada par de posiciones está en exactamente una terna.

Work modulo 33 with 303 \equiv 0 and 21.2 \equiv -1. The seven index triples {i,i+1,i+3}\{i, i+1, i+3\} form a Fano plane: every pair of positions lies in exactly one triple.

Pista grande:

Clasifica según el conjunto ZZ de posiciones que tienen un 3:3: un producto sobrevive solo si su terna evita Z,Z, y cada producto que sobrevive es ±1\pm 1 según cuántos 22 contenga su terna.

Classify by the set ZZ of positions holding a 3:3: a product survives only if its triple misses Z,Z, and each surviving product is ±1\pm 1 according to how many 22’s its triple holds.

Solución:

Trabajamos módulo 3:3: las entradas 33 son 00 y las entradas 1,21, 2 son ±1.\pm 1. Como las diferencias de {0,1,3}\{0, 1, 3\} alcanzan cada residuo no nulo módulo 77 exactamente una vez, las siete ternas {i,i+1,i+3}\{i, i+1, i+3\} son las rectas de un plano de Fano sobre las posiciones: cada par de posiciones está en exactamente una recta, y dos rectas cualesquiera se cortan en exactamente un punto. Sea ZZ el conjunto de posiciones que tienen un 33 y k=Z.k = |Z|. Un término producto sobrevive exactamente cuando su recta evita Z.Z. Si tLt_L es el número de entradas iguales a 22 en esa recta, su contribución es (1)tL,(-1)^{t_L}, y la condición lineal obliga a que los 7k7 - k valores ±1\pm 1 sumen 00 módulo 3.3.

Análisis por casos según k.k. k=7:k = 7: la tupla con todos 33 funciona: 1.1. k=6:k = 6: un único ±1\pm 1 no puede sumar 0:0: ninguna. k=5:k = 5: ninguna recta sobrevive; las dos entradas no nulas deben ser un 11 y un 2:2: (72)2=42.\binom{7}{2} \cdot 2 = 42. k=4:k = 4: tres ±1\pm 1 suman 00 solo si son todos iguales, y las tres posiciones no nulas no deben formar una recta, pues de lo contrario su producto es ±1:\pm 1: (357)2=56.(35 - 7) \cdot 2 = 56. k=3:k = 3: cuatro ±1\pm 1 deben repartirse dos y dos; si ZZ no es una recta, exactamente una recta la evita (arruinando la suma), mientras que si ZZ es una recta ninguna recta la evita: 7(42)=42.7 \cdot \binom{4}{2} = 42. k=2:k = 2: cinco ±1\pm 1 deben ir cuatro y uno; exactamente dos rectas evitan Z,Z, se cortan en un punto pp y cubren las cinco posiciones, y sus productos se cancelan exactamente cuando el único valor minoritario evita p:p: (72)24=168.\binom{7}{2} \cdot 2 \cdot 4 = 168. k=1:k = 1: seis ±1\pm 1 suman 00 si son todos iguales o tres de cada; las cuatro rectas que evitan ZZ se cortan por pares en las seis posiciones no nulas, y como el producto de los cuatro productos de recta es +1,+1, necesitamos exactamente dos rectas negativas. Todos iguales da 00 o 44 rectas negativas; para tres 22, viendo las posiciones como aristas de K4K_4 sobre las cuatro rectas, una recta es negativa exactamente cuando tiene grado impar en el conjunto de 33 aristas elegido, y exactamente los 1212 caminos de tres aristas (de los (63)=20\binom{6}{3} = 20 subconjuntos) dan dos grados impares: 712=84.7 \cdot 12 = 84. k=0:k = 0: siete ±1\pm 1 necesitan dos o cinco 22, que hacen 44 o 33 rectas negativas respectivamente, pero 72t0(mod3)7 - 2t \equiv 0 \pmod 3 necesita t2(mod3):t \equiv 2 \pmod 3: ninguna.

El total es 1+42+561 + 42 + 56 +42+168+84=393.+ 42 + 168 + 84 = 393.

Work modulo 3:3: entries 33 are 00 and entries 1,21, 2 are ±1.\pm 1. Because the differences of {0,1,3}\{0, 1, 3\} hit every nonzero residue mod 77 exactly once, the seven triples {i,i+1,i+3}\{i, i+1, i+3\} are the lines of a Fano plane on the positions: every pair of positions lies on exactly one line, and any two lines meet in exactly one point. Let ZZ be the set of positions holding a 33 and k=Z.k = |Z|. A product term survives exactly when its line avoids Z.Z. If tLt_L is the number of entries equal to 22 on that line, its contribution is (1)tL,(-1)^{t_L}, and the linear condition constrains the 7k7 - k values ±1\pm 1 to sum to 00 mod 3.3.

Casework on k.k. k=7:k = 7: the all-33s tuple works: 1.1. k=6:k = 6: a single ±1\pm 1 can’t sum to 0:0: none. k=5:k = 5: no line survives; the two nonzero entries must be a 11 and a 2:2: (72)2=42.\binom{7}{2} \cdot 2 = 42. k=4:k = 4: three ±1\pm 1s sum to 00 only if all equal, and the three nonzero positions must not form a line, else its product is ±1:\pm 1: (357)2=56.(35 - 7) \cdot 2 = 56. k=3:k = 3: four ±1\pm 1s must split two and two; exactly one line avoids a non-line ZZ (spoiling the sum), while a line ZZ is avoided by no line: 7(42)=42.7 \cdot \binom{4}{2} = 42. k=2:k = 2: five ±1\pm 1s must go four and one; exactly two lines avoid Z,Z, meeting at a point pp and covering the five positions, and their products cancel exactly when the lone minority value avoids p:p: (72)24=168.\binom{7}{2} \cdot 2 \cdot 4 = 168. k=1:k = 1: six ±1\pm 1s sum to 00 if all equal or three of each; the four lines avoiding ZZ pairwise meet in the six nonzero positions, and since the product of all four line-products is +1,+1, we need exactly two negative lines. All-equal gives 00 or 44 negative lines; for three 22’s, viewing positions as edges of K4K_4 on the four lines, a line is negative exactly when it has odd degree in the chosen 33-edge set, and exactly the 1212 three-edge paths (of the (63)=20\binom{6}{3} = 20 subsets) give two odd degrees: 712=84.7 \cdot 12 = 84. k=0:k = 0: seven ±1\pm 1s need two or five 22’s, which make 44 or 33 lines negative respectively, but 72t0(mod3)7 - 2t \equiv 0 \pmod 3 needs t2(mod3):t \equiv 2 \pmod 3: none.

The total is 1+42+561 + 42 + 56 +42+168+84=393.+ 42 + 168 + 84 = 393.