0:00:06.412,0:00:10.558
You and a fellow castaway[br]are stranded on a desert island

0:00:10.558,0:00:13.610
playing dice for the last banana.

0:00:13.610,0:00:15.604
You've agreed on these rules:

0:00:15.604,0:00:17.146
You'll roll two dice,

0:00:17.146,0:00:21.069
and if the biggest number [br]is one, two, three or four,

0:00:21.069,0:00:23.353
player one wins.

0:00:23.353,0:00:28.326
If the biggest number is five or six, [br]player two wins.

0:00:28.326,0:00:30.154
Let's try twice more.

0:00:30.154,0:00:33.247
Here, player one wins,

0:00:33.247,0:00:35.971
and here it's player two.

0:00:35.971,0:00:37.741
So who do you want to be?

0:00:37.741,0:00:42.207
At first glance, it may seem [br]like player one has the advantage

0:00:42.207,0:00:46.222
since she'll win if any one [br]of four numbers is the highest,

0:00:46.222,0:00:47.236
but actually,

0:00:47.236,0:00:53.619
player two has an approximately[br]56% chance of winning each match.

0:00:53.619,0:00:57.527
One way to see that is to list all[br]the possible combinations you could get

0:00:57.527,0:00:59.527
by rolling two dice,

0:00:59.527,0:01:02.674
and then count up [br]the ones that each player wins.

0:01:02.674,0:01:05.308
These are the possibilities[br]for the yellow die.

0:01:05.308,0:01:07.784
These are the possibilities [br]for the blue die.

0:01:07.784,0:01:13.214
Each cell in the chart shows a possible[br]combination when you roll both dice.

0:01:13.214,0:01:15.269
If you roll a four and then a five,

0:01:15.269,0:01:17.445
we'll mark a player two [br]victory in this cell.

0:01:17.445,0:01:22.496
A three and a one gives [br]player one a victory here.

0:01:22.496,0:01:24.817
There are 36 possible combinations,

0:01:24.817,0:01:28.091
each with exactly the same [br]chance of happening.

0:01:28.091,0:01:31.236
Mathematicians call these[br]equiprobable events.

0:01:31.236,0:01:34.801
Now we can see why [br]the first glance was wrong.

0:01:34.801,0:01:37.466
Even though player one [br]has four winning numbers,

0:01:37.466,0:01:39.560
and player two only has two,

0:01:39.560,0:01:43.704
the chance of each number [br]being the greatest is not the same.

0:01:43.704,0:01:48.681
There is only a one in 36 chance[br]that one will be the highest number.

0:01:48.681,0:01:52.857
But there's an 11 in 36 chance [br]that six will be the highest.

0:01:52.857,0:01:55.586
So if any of these [br]combinations are rolled,

0:01:55.586,0:01:57.473
player one will win.

0:01:57.473,0:01:59.668
And if any of these [br]combinations are rolled,

0:01:59.668,0:02:01.397
player two will win.

0:02:01.397,0:02:03.719
Out of the 36 possible combinations,

0:02:03.719,0:02:09.819
16 give the victory to player one,[br]and 20 give player two the win.

0:02:09.819,0:02:12.163
You could think about it this way, too.

0:02:12.163,0:02:14.359
The only way player one can win

0:02:14.359,0:02:18.639
is if both dice show [br]a one, two, three or four.

0:02:18.639,0:02:21.596
A five or six would mean [br]a win for player two.

0:02:21.596,0:02:26.705
The chance of one die showing one, two,[br]three or four is four out of six.

0:02:26.705,0:02:30.556
The result of each die roll [br]is independent from the other.

0:02:30.556,0:02:33.869
And you can calculate the joint [br]probability of independent events

0:02:33.869,0:02:36.386
by multiplying their probabilities.

0:02:36.386,0:02:40.822
So the chance of getting a one, two, [br]three or four on both dice

0:02:40.822,0:02:46.279
is 4/6 times 4/6, or 16/36.

0:02:46.279,0:02:48.467
Because someone has to win,

0:02:48.467,0:02:54.502
the chance of player two winning[br]is 36/36 minus 16/36,

0:02:54.502,0:02:57.303
or 20/36.

0:02:57.303,0:03:01.409
Those are the exact same probabilities[br]we got by making our table.

0:03:01.409,0:03:04.045
But this doesn't mean [br]that player two will win,

0:03:04.045,0:03:09.413
or even that if you played 36 games[br]as player two, you'd win 20 of them.

0:03:09.413,0:03:12.624
That's why events like dice rolling [br]are called random.

0:03:12.624,0:03:15.903
Even though you can calculate [br]the theoretical probability

0:03:15.903,0:03:17.415
of each outcome,

0:03:17.415,0:03:22.070
you might not get the expected results[br]if you examine just a few events.

0:03:22.070,0:03:26.417
But if you repeat those random events[br]many, many, many times,

0:03:26.417,0:03:30.357
the frequency of a specific outcome,[br]like a player two win,

0:03:30.357,0:03:33.418
will approach its theoretical probability,

0:03:33.418,0:03:36.372
that value we got by writing down [br]all the possibilities

0:03:36.372,0:03:39.039
and counting up the ones for each outcome.

0:03:39.039,0:03:42.994
So, if you sat on that desert island[br]playing dice forever,

0:03:42.994,0:03:46.913
player two would eventually [br]win 56% of the games,

0:03:46.913,0:03:49.995
and player one would win 44%.

0:03:49.995,0:03:53.564
But by then, of course, the banana[br]would be long gone.