AIME Flashcards
(2012 AIME I #10)
Let S be the set of all perfect squares whose rightmost three digits in base 10 are 256. Let T be the set of all numbers of the form {x-256}/{1000}, where x is in S. In other words, T is the set of numbers that result when the last three digits of each number in S are truncated. Find the remainder when the tenth smallest element of T is divided by 1000.
What are the first steps?
- Replace x with n^2
- Convert to equation n^2-256=1000k (n, k are integers)
- Diophantine, so factor (n-16)(n+16)=2^35^3k
Given (n-16)(n+16)=2^35^3k, what do you do next?
Since the 2 factors on the left by 32, exactly one of them must be a multiple of 125, and both of them must be multiples of 4. Therefore, one of them is a multiple of 500. Therefore, the solutions for n are n=16, 500-16, 500+16, 1000-16, 1000+16,… Therefore the answer to 2012 AIME I #10 is [(2500-16)^2-256]/1000=6170 –> 170
When something is changing in a complex manner, what do you do?
Look for something that doesn’t change (an invariant), e.g., 2012 AIME I #11 x+y varies by 3m, x-y changes by 5n
What are the (complex) solutions to z^n=1?
cos(2kpi/n)+isin(2kpi/n), k=1,2,3, … n-1
If you have angle trisectors, what do you also have?
Angle bisectors, which you can use the angle bisector theorem with
When looking for the sin, cosine, and tangent of an angle (especially in a triangle), what should you do?
Look for law of sines/cosines, drop altitudes, and look for right triangles with the values you need in them.
A point on a figure, when a transformation applies, needs to map onto another figure. What do you do?
Apply the transformation to the first figure. Note that not all points will be able to satisfy this.
You need to count the numbers that are not multiples of j or k between m and n. What do you do?
Count how many are not between 1 and gcd(j, k), count how many multiples of gcd(j, k) are between m and n, multiply those numbers, and add the extra.
What are Grant Sanderson’s 9 tips for problem solving?
- Use the defining features of the setup.
- Give things meaningful names.
- Leverage symmetry
- Describe 1 object 2 ways
- Draw a picture (use numbers as coordinates)
- Ask simpler versions of the problem
- Read a lot, and think about problems a lot
- Always sanity check
- Learn a bit of programming
What is 1/log2(100!)+1/log3(100!)+…+1/log100(100!)?
Two methods:
- Using logb(a) = 1/loga(b), we get log100!(2)+log100!(3)+…log100!(100) = log100!(23…*100) = 1
- Using change of base logb(a)=log(b)/log(a), we get log(2)/log(100!)+log(3)/log(100!)+…+log(100)/log(100!) = log(23…*100)/log(100!) = 1
What do you do if you see a strange restriction?
Forget about it if you can - incorporate it later.
Suppose you have a solution that satisfies some, but not all of the constraints, what can you do?
Try to establish a relationship between some of the properties of your solution. If they hold, you may be able to find the answer without the actual solution.
For example: You need to find the length of a hypotenuse, and you know one of the legs. If you find a solution that isn’t perfect, but can scale appropriately, and you know the ratio between the leg and hypotenuse, you can find the hypotenuse.
You have a+b+c = 0. When you have a value of a+b, what should you call it?
-c. It will simplify things.