This is super interesting. As you mention, it seems really strange though that your own payoff is not taken into account! Trying to wrestle with that for a game I play somewhat regularly
So basically unless you choose a certain point on the spectrum, as to how much percentage you choose one over the other, the opponent can always choose one move and stick to that one move only, to maximize his win and minimize yours. The more extreme the percentage of our choice is, away from the balance point, the more the opponent can stick with the counter move. The closer to the balance point, the less the opponent can take advantage of by just sticking with one move. At the center point, regardless of what move he stick to, the damage to you is the same. Any other point, will lead the opponent to damage you more. Wow… I was watching a course on Wondrium (The Great Courses) and was so freaking confused. I love the explanation here much better. The college professor other there didn’t even take the time to explain why the 50/50 point for the opponent is the best for us.
Is there any psychological studies, where human agents will overshoot their next strategy in response to the poor pay-off of the previous strategy? For example If Player 1 keeps going left, will they likely and consistently overshoot to the right in response? Or is this dependent on the individual and context of the game?
This assumes a random distribution between options, however in lots of adversarial games (for instance like a pitcher+catcher choosing pitches, versus a batter sitting on certain options) the sequences are rarely uniformly random. Is there any way to exploit human pattern recognition/psychology to change the sequence of choices in order to maximize favorable outcomes? In a lot of cases you are playing against someone trying to maximize value rather than being truly unexploitable For example, if p1 goes l,l,l,l,... How many times before p2 perceives an opportunity to adjusts strategy for optimal exploitation?
Thank you for your very clear explanation. But there is one thing that I do not understand. Here you say that p =3/7 but here th-cam.com/video/S49oKtISkqI/w-d-xo.html instead is p=2/3. Am I missing something ? Thank you
The optimal value for p should be 2/3. A little mistake made by Professor Hodgson. Just replace the 3/7 with 2/3. 2/3 will be the optimal value for q. However, the important part of this video is the logic and the intuition provided, which is excellent!
Hi Ashley, thanks for your enlightening videos. I have an observation. In the other videos solving the same problem solution is p=2/3; q=3/7.
This is super interesting. As you mention, it seems really strange though that your own payoff is not taken into account! Trying to wrestle with that for a game I play somewhat regularly
So basically unless you choose a certain point on the spectrum, as to how much percentage you choose one over the other, the opponent can always choose one move and stick to that one move only, to maximize his win and minimize yours. The more extreme the percentage of our choice is, away from the balance point, the more the opponent can stick with the counter move. The closer to the balance point, the less the opponent can take advantage of by just sticking with one move. At the center point, regardless of what move he stick to, the damage to you is the same. Any other point, will lead the opponent to damage you more. Wow… I was watching a course on Wondrium (The Great Courses) and was so freaking confused. I love the explanation here much better. The college professor other there didn’t even take the time to explain why the 50/50 point for the opponent is the best for us.
Madam, you are amazing. Thank you for this.
Extremely interesting lecture with the good explanation. Thanks ma'am.
u ar lovely and ur teaching is love, i wish my ecomics teacher are as understandable as you....
hi, isn't p equal to 2/3 instead of 3/7
Is there any psychological studies, where human agents will overshoot their next strategy in response to the poor pay-off of the previous strategy? For example If Player 1 keeps going left, will they likely and consistently overshoot to the right in response? Or is this dependent on the individual and context of the game?
Excellent video. Mixing games theory and probabilities is amazing.
This is really cool. I am really looking forward to the next video.
I was literally in need of this vedio. Thank you ❤️ mam
Now I understand, thank you for the clear explanation :)
Plz upload a vedio on repeated games.....urgently needed
very nice explanation, thanks a lot
I m waiting for the 3/7 demonstration it could be very instructive
My hope is to have it up for next week.
This assumes a random distribution between options, however in lots of adversarial games (for instance like a pitcher+catcher choosing pitches, versus a batter sitting on certain options) the sequences are rarely uniformly random.
Is there any way to exploit human pattern recognition/psychology to change the sequence of choices in order to maximize favorable outcomes? In a lot of cases you are playing against someone trying to maximize value rather than being truly unexploitable
For example, if p1 goes l,l,l,l,... How many times before p2 perceives an opportunity to adjusts strategy for optimal exploitation?
how do you get 3/7
Thank you for your very clear explanation. But there is one thing that I do not understand. Here you say that p =3/7 but here th-cam.com/video/S49oKtISkqI/w-d-xo.html instead is p=2/3. Am I missing something ? Thank you
The optimal value for p should be 2/3. A little mistake made by Professor Hodgson. Just replace the 3/7 with 2/3. 2/3 will be the optimal value for q. However, the important part of this video is the logic and the intuition provided, which is excellent!
I think p=2/3 not 3/7