To settle a debate; when still having questions left to answer on a test with not enough time to do so - is it more effective to choose a single alternative and answer that for all questions (for example checking "A" on all) than to answer randomly

randtest.py

import random
import time
import sys

def createTest(a, n):
	l = []

	for _ in range(0, n):
		l.append([0 for _ in range(0, a)])
		l[-1][random.randint(0, a - 1)] = 1
	
	return l

def countCorrectRandom(l):
	c = 0

	for li in l:
		a = random.randint(0, len(li) - 1)
		if li[a] == 1:
			c += 1
	
	return c

def countCorrectStraight(l, a):
	c = 0

	for li in l:
		if li[a] == 1:
			c += 1
	
	return c

def testOneList():
	a = random.randint(2, 10)
	n = a * random.randint(3, 10)

	l = createTest(a, n)

	sCorrect = countCorrectStraight(l, random.randint(1, a) - 1) 
	rCorrect = countCorrectRandom(l)

	return (sCorrect, rCorrect)

def runTests(tests):
	sWin = 0
	rWin = 0
	ties = 0

	for _ in range(0, tests):
		s, r = testOneList()

		if s < r:
			sWin += 1
		elif s > r:
			rWin += 1
		else:
			ties += 1
	
	print("Straight wins:", sWin)
	print("Random wins:", rWin)
	print("Ties:", ties)

	print()
	print("Diff:", sWin - rWin, "({}%)".format(abs(sWin - rWin) / tests * 100))

	winner = sorted([
		("Straight", sWin),
		("Random", rWin),
		("No one (tie)", ties)
	], key=lambda tup: tup[1], reverse=True)[0][0]

	print()
	print(winner + " won!")

def main():
	try:
		N = int(sys.argv[1])
	except:
		N = 1

	print("Running", N, "tests...")
	print()

	startTime = time.time()

	runTests(N)

	endTime = time.time() - startTime

	print()
	print("Took %s seconds" % (endTime))

if __name__ == "__main__":
	main()