Due: Wednesday, February 9 at 10:59pm (submit via Gradescope).
Policy: Can be solved in groups (acknowledge collaborators) but must be written up individually
Submission: Your submission should be a PDF that matches this template. Each page of the PDF should
align with the corresponding page of the template (page 1 has name/collaborators, question 1 begins on page
2, etc.). Do not reorder, split, combine, or add extra pages. The intention is that you print out the
template, write on the page in pen/pencil, and then scan or take pictures of the pages to make your submission.
You may also fill out this template digitally (e.g. using a tablet.)
First name
Last name
SID
Collaborators
For sta↵ use only:
Q1. Expectimax Yahtzee /40
Total /40
1
Muskaan
Goyal
3032666782
MIA
This study source was downloaded by 100000840858457 from CourseHero.com on 07-06-2022 03:01:12 GMT -05:00
https://www.coursehero.com/file/130827698/cs-188-spring-2022-written-homework-2-2pdf/
Q1. [40 pts] Expectimax Yahtzee
Consider a simplified version of the game Yahtzee. In this game, we have 3 regular tetrahedral dice with 4 sides each
(numbered 1-4) and the game begins by rolling all 3 dice. At this point, a player can make a decision: pick one of
the 3 dice to reroll, or don’t reroll anything. Then, points are assigned as follows:
• A reward of 10 points is given for two-of-a-kind (for example, 4-4).
• A reward of 15 is given to three-of-a-kind (for example, 4-4-4).
• A reward of 7 points is given for rolling a series (1-2-3 or 2-3-4).
• Otherwise (or if the sum is higher than the special reward), the score is equal to the sum of all 3 dice.
(a) We will formulate this problem as an expectimax tree.
(i) [3 pts] The resulting tree for the problem is drawn below. Given a specific initial roll, the branching factor
(of the player’s decision) from the root node is . The branching factor at the chance nodes is
. What do those chance nodes represent? (There are multiple solutions, you only need to write
down one solution)
• Chance node 1:
• Chance node 2:
• Chance node 3:
2
4
4
neroll dice I
renew dice 2
renew dice 3
max
special
reward
I 2 3 4 I 2 3 4 I 2 3 Y
This study source was downloaded by 100000840858457 from CourseHero.com on 07-06-2022 03:01:12 GMT -05:00
https://www.coursehero.com/file/130827698/cs-188-spring-2022-written-homework-2-2pdf/
(ii) [7 pts] Given a starting roll (1,2,4) (corresponding to the outcomes of die rolls 1, 2, and 3 respectively),
what move should you take? Fill in the values of the expectimax tree below to justify your answer.
3
no recall
poor
died
puog.IT Refs
9 as 7 ( 1,214)
Suma 7
7 10 9 10 10 7 8 10 10 10 7 7
Reroddiu reward Rerotldrcel :
Reynard ① ( 41,4)
1 : ① 41214 )
f | ② ( " " "I 7
② (2,214 ) 1°
③ ( 1,314)
⑨ ( 114,4) 10
③ 1312,4) 9
④ (4/44)
10
"
"%→"""""
As revolting any
dice
gives better rewards,
RuoUDñYpqand so we should
choose sending over
① ( 112,1) 10
not no - revolt .
② ( 1,2 , 2) 10
③ 11 , 2,3)
⑨ 1112,4) / 7
Average → 3414--85 /
This study source was downloaded by 100000840858457 from CourseHero.com on 07-06-2022 03:01:12 GMT -05:00
https://www.coursehero.com/file/130827698/cs-188-spring-2022-written-homework-2-2pdf/
Now suppose the human player does not understand how to play the game, and as a result, they choose any action
with uniform probability, regardless of the initial roll. Moreover, we assume that the human’s choice will be carried
out by a ”somewhat helpful” robot called Albertbot: given a configuration of dice and the desired action from the
human, this robot either actually implements the human’s action (with probability 1
Category | exam bundles |
Comments | 0 |
Rating | |
Sales | 0 |