Problem 84: Monopoly odds

(see projecteuler.net/problem=84)

In the game, Monopoly, the standard board is set up in the following way:

board

A player starts on the GO square and adds the scores on two 6-sided dice to determine the number of squares they advance in a clockwise direction.
Without any further rules we would expect to visit each square with equal probability: 2.5%.
However, landing on G2J (Go To Jail), CC (community chest), and CH (chance) changes this distribution.

In addition to G2J, and one card from each of CC and CH, that orders the player to go directly to jail, if a player rolls three consecutive doubles,
they do not advance the result of their 3rd roll. Instead they proceed directly to jail.

At the beginning of the game, the CC and CH cards are shuffled. When a player lands on CC or CH they take a card from the top of the respective pile and,
after following the instructions, it is returned to the bottom of the pile.
There are sixteen cards in each pile, but for the purpose of this problem we are only concerned with cards that order a movement;
any instruction not concerned with movement will be ignored and the player will remain on the CC/CH square.

- Community Chest (2/16 cards):
Advance to GO, Go to JAIL

- Chance (10/16 cards):
Advance to GO, Go to JAIL,
Go to C1, Go to E3, Go to H2, Go to R1,
Go to next R (railway company), Go to next R, Go to next U (utility company),
Go back 3 squares

The heart of this problem concerns the likelihood of visiting a particular square. That is, the probability of finishing at that square after a roll.
For this reason it should be clear that, with the exception of G2J for which the probability of finishing on it is zero, the CH squares will have the lowest probabilities,
as 5/8 request a movement to another square, and it is the final square that the player finishes at on each roll that we are interested in.
We shall make no distinction between "Just Visiting" and being sent to JAIL, and we shall also ignore the rule about requiring a double to "get out of jail",
assuming that they pay to get out on their next turn.

By starting at GO and numbering the squares sequentially from 00 to 39 we can concatenate these two-digit numbers to produce strings that correspond with sets of squares.

Statistically it can be shown that the three most popular squares, in order, are JAIL (6.24%) = Square 10, E3 (3.18%) = Square 24, and GO (3.09%) = Square 00.
So these three most popular squares can be listed with the six-digit modal string: 102400.

If, instead of using two 6-sided dice, two 4-sided dice are used, find the six-digit modal string.

My Algorithm

Disclaimer: this was one of my least favorite problems so far.

The main idea is to run a Monte-Carlo simulation (see en.wikipedia.org/wiki/Monte_Carlo_method):
roll the dice five million times and count how often you land on each field.
The longer the simulation runs, the closer you get to the true statistical value.

Because I hate this problem so much, there is no further explanation of my code ...

Modifications by HackerRank

You can't solve this problem with my approach: only a true mathematical analysis will give the correct order of all fields
(yes, some of their test cases ask for all 40 fields in descendingly sorted order).
Right now I am not interesting in rewriting my code. Apparently, Markov chains are the way to go (see en.wikipedia.org/wiki/Markov_chain).

Note

I reduced Rolls from the original 5 million to 1 million otherwise my live tests would time out.
It still solves the original problem but has a lower score at Hackerrank.

My code

… was written in C++11 and can be compiled with G++, Clang++, Visual C++. You can download it, too.

The code contains #ifdefs to switch between the original problem and the Hackerrank version.
Enable #ifdef ORIGINAL to produce the result for the original problem (default setting for most problems).

#include <vector>
#include <map>
#include <algorithm>
#include <iomanip>
#include <iostream>
#include <cmath>
 
int main()
{
// four-sided dices ?
unsigned int diceSize = 4;
std::cin >> diceSize;
// print the 3 most frequent fields
unsigned int showFields = 3;
std::cin >> showFields;
 
// number of simulated moves
const unsigned int Rolls = 1000000;//50000000;
 
// always the same random numbers
srand(0);
 
// board setup
const unsigned int NumFields = 40;
// special field positions
const unsigned int Go = 0;
const unsigned int Jail = 10;
const unsigned int GoToJail = 30;
const unsigned int Community [3] = { 2, 17, 33 };
const unsigned int Chance [3] = { 7, 22, 36 };
const unsigned int NextRailway[3] = { 15, 25, 5 }; // index x corresponds to Chance[x]
const unsigned int NextUtility[3] = { 12, 28, 12 }; // index x corresponds to Chance[x]
 
// random setup of community and chance cards, give them IDs 0..15
std::vector<unsigned int> chance, community;
for (unsigned int i = 0; i < 16; i++)
{
chance .push_back(i);
community.push_back(i);
}
std::random_shuffle(chance.begin(), chance.end());
std::random_shuffle(community.begin(), community.end());
 
// Monte-Carlo simulation
unsigned int current = Go;
unsigned int doubles = 0;
std::vector<unsigned long long> count(NumFields, 0);
for (unsigned int rolls = 0; rolls < Rolls; rolls++)
{
// roll the dice ...
unsigned int dice1 = (rand() % diceSize) + 1;
unsigned int dice2 = (rand() % diceSize) + 1;
// next field
unsigned int next = (current + dice1 + dice2) % NumFields;
 
// double ?
if (dice1 == dice2)
doubles++;
else
doubles = 0;
 
if (doubles == 3)
{
next = Jail;
doubles = 0;
}
 
// chance
if (next == Chance[0] || next == Chance[1] || next == Chance[2])
{
// ID of current chance field (needed to figure out next railway / utility)
int id = 0;
if (next == Chance[1])
id = 1;
if (next == Chance[2])
id = 2;
 
// check card ID of the topmost card
switch (chance.front())
{
case 0: next = Go; break;
case 1: next = Jail; break;
case 2: next = 11; break; // C1
case 3: next = 24; break; // E3
case 4: next = 39; break; // H2
case 5: next = 5; break; // R1
// go back 3 squares (might land on GoToJail, too)
// awkward formula needs to take care of CC1 which redirects to H2
case 6: next = (next + NumFields - 3) % NumFields; break;
case 7: // two railway cards
case 8: next = NextRailway[id]; break;
case 9: next = NextUtility[id]; break;
default: break;
}
 
// move card to the bottom
std::rotate(chance.begin(), chance.begin() + 1, chance.end());
}
 
// community
if (next == Community[0] || next == Community[1] || next == Community[2])
{
// check card ID of the topmost card
switch (community.front())
{
case 0: next = Go; break;
case 1: next = Jail; break;
default: break;
}
 
// move card to the bottom
std::rotate(community.begin(), community.begin() + 1, community.end());
}
 
if (next == GoToJail)
next = Jail;
 
count[next]++;
current = next;
}
 
// normalize (actually not needed, just to ease debugging)
unsigned long long sum = 0;
for (auto x : count)
sum += x;
 
// use std::map's implicit sorting
std::multimap<double, unsigned int> sorted; // multimap in case of identical probabilities
for (unsigned int i = 0; i < count.size(); i++)
sorted.insert(std::make_pair(count[i] * 100.0 / sum, i));
 
// extract the field positions
std::vector<unsigned int> result;
for (auto x : sorted)
result.push_back(x.second);
 
// and show the most frequent ones (they are at the end of the container)
auto i = result.rbegin();
 
#define ORIGINAL
#ifdef ORIGINAL
 
for (unsigned int j = 0; j < showFields; j++)
std::cout << std::setw(2) << std::setfill('0') << *i++;
 
#else
 
// field names
const char* names[] = { "GO", "A1", "CC1", "A2", "T1", "R1", "B1", "CH1", "B2", "B3",
"JAIL", "C1", "U1", "C2", "C3", "R2", "D1", "CC2", "D2", "D3",
"FP", "E1", "CH2", "E2", "E3", "R3", "F1", "F2", "U2", "F3",
"G2J", "G1", "G2", "CC3", "G3", "R4", "CH3", "H1", "T2", "H2" };
for (unsigned int j = 0; j < showFields; j++)
std::cout << names[*i++] << " ";
 
#endif
 
return 0;
}

This solution contains 25 empty lines, 25 comments and 10 preprocessor commands.

Interactive test

You can submit your own input to my program and it will be instantly processed at my server:

Input data (separated by spaces or newlines):

This is equivalent to
echo "6 3" | ./84

Output:

(please click 'Go !')

Note: the original problem's input 4 3 cannot be entered
because just copying results is a soft skill reserved for idiots.

(this interactive test is still under development, computations will be aborted after one second)

Benchmark

The correct solution to the original Project Euler problem was found in 0.05 seconds on a Intel® Core™ i7-2600K CPU @ 3.40GHz.
(compiled for x86_64 / Linux, GCC flags: -O3 -march=native -fno-exceptions -fno-rtti -std=c++11 -DORIGINAL)

See here for a comparison of all solutions.

Note: interactive tests run on a weaker (=slower) computer. Some interactive tests are compiled without -DORIGINAL.

Changelog

March 17, 2017 submitted solution
May 9, 2017 added comments

Hackerrank

see https://www.hackerrank.com/contests/projecteuler/challenges/euler084

My code solves 11 out of 21 test cases (score: 55%)

I failed 10 test cases due to wrong answers and 0 because of timeouts

Difficulty

35% Project Euler ranks this problem at 35% (out of 100%).

Hackerrank describes this problem as hard.

Note:
Hackerrank has strict execution time limits (typically 2 seconds for C++ code) and often a much wider input range than the original problem.
In my opinion, Hackerrank's modified problems are usually a lot harder to solve. As a rule thumb: brute-force is rarely an option.

Links

projecteuler.net/thread=84 - the best forum on the subject (note: you have to submit the correct solution first)

Code in various languages:

Python: www.mathblog.dk/project-euler-84-monopoly-popular-squares/ (written by Kristian Edlund)
Java: github.com/nayuki/Project-Euler-solutions/blob/master/java/p084.java (written by Nayuki)
Scala: github.com/samskivert/euler-scala/blob/master/Euler084.scala (written by Michael Bayne)

Heatmap

green problems solve the original Project Euler problem and have a perfect score of 100% at Hackerrank, too.
yellow problems score less than 100% at Hackerrank (but still solve the original problem).
gray problems are already solved but I haven't published my solution yet.
blue problems are solved and there wasn't a Hackerrank version of it at the time I solved it or I didn't care about it because it differed too much.
red problems are solved but exceed the time limit of one minute or the memory limit of 256 MByte.

Please click on a problem's number to open my solution to that problem:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175
176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200
201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300
301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350
351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375
376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425
426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450
451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475
476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500
501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525
526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
The 233 solved problems (level 9) had an average difficulty of 29.0% at Project Euler and
I scored 12,983 points (out of 15100 possible points, top rank was 17 out of ≈60000 in August 2017) at Hackerrank's Project Euler+.
Look at my progress and performance pages to get more details.

My username at Project Euler is stephanbrumme while it's stbrumme at Hackerrank.

more about me can be found on my homepage, especially in my coding blog.
some names mentioned on this site may be trademarks of their respective owners.
thanks to the KaTeX team for their great typesetting library !