sir thank you for ur effort but i dont understand why we multiply the perfect information by their probablity? if it is the pefect information why we need calculating probablity and i dont get the concept of perfect information/

It is because when we pay for the information we do not know what answer we are going to be told. We will be told the correct outcome (and then then make the best decision) but it could be any of the possible ones and we know what the probability of each is of occurring.

I don’t understand why we have to calculate the fees to pay to the Market researcher by basing it on the difference between the expected value of perfect information with that of without one? Why do we calculate the “expected value” of perfect information in the first place – ie the calculation using probability he will bring this answer or that answer to pay someone for perfect information that he is going to bring either way if you pay him -and which will thus eliminate the need for the use of any unnecessary probabilities ? Why thus is the expected value of perfect information a necessity in the first place for calculating the fees of perfect information?

But why do we need to study expected return of perfect information in order to pay this fees ? Why study probability when you want to pay info for not having probabilities

I understand the logic behind the $300 as the maximum amount we are willing to pay for research, on the basis that the decision maker is risk-neutral and expects for a profit of $4,500. My question is, why do we have to compare the “average” guaranteed profit of $4,800 to the expected value? For instance, if the decision maker is a risk-avoider, so he decides to sign contract for 800 units for a profit of $4,400 (based on Maximin), then wouldn’t the maximum amount for research in this case be $400? ($4,800 – $4,400)

What you say is correct. However for the exam, we always calculate the maximum to pay for perfect knowledge based on the expected values (and I do state effectively what you are saying in my lecture).

I have a doubt about expected values. The customer contract is a fixed quantity so there no required to use the probability. So, isn’t it correct that we use the full contract value plus the probability of the expected market demand?

for eg: At contract 300 and demand 400, contract profit is 300×3, and demand profit is 400×0.2×5, so the expected value will be 1300?

The problem is that they will. not always be able to supply the expected normal demand. For example, if they have signed the contract to supply 500 units, then the most they can sell to the normal demand is 700 units (because the capacity is limited to 1,200 units) regardless of the size of the normal demand.

Hi John, im abit unsure on how we get to putting 400,500,700,900 at the top and 300,500,700,800 on side, is there a particular ruling which needs to be followed when labelling each side and then making the calculations, such as when we do maximin and minimax we go sideways but when looking at minimax regret we go downwards? im just confused as I always label the sides wrong and because of that all my workings are wrong,

It does not matter which way round you draw the table. What matters is that you identify what it is that is uncertain and what it is that we have to make a decision on. Don’t learn it just as rules – make sure you really understand the basis on which we are making the decision under each of the methods.

Well you are completely wrong – I do all of my teaching in non-English speaking countries in Eastern Europe 🙂

I am puzzled that you only say this for this lecture. Have you not watched all of the previous lectures – if you have then you have presumably got used to my speaking? There is no point in watching a lecture out of order – the lectures are a complete free course and cover everything needed to be able to pass the exam well (and, of course, this lecture makes no sense on its own because it follows on from the earlier chapters).

Hi John, I didn’t understand how you came up with expected value of 4500, if we signed the contract of 700, as I figured, profit is calculated like: 700*3+(1200-700)*5 = 4600, did I miss the point?

Hi John, I don’t understand why do we must use this way to calculate the EV if we using market research? How do we know to if this is the way to calculate the EV if we using market research to obtain perfect information?

hermela says

sir thank you for ur effort but i dont understand why we multiply the perfect information by their probablity? if it is the pefect information why we need calculating probablity and i dont get the concept of perfect information/

John Moffat says

It is because when we pay for the information we do not know what answer we are going to be told. We will be told the correct outcome (and then then make the best decision) but it could be any of the possible ones and we know what the probability of each is of occurring.

Asif110 says

Greetings.

I don’t understand why we have to calculate the fees to pay to the Market researcher by basing it on the difference between the expected value of perfect information with that of without one? Why do we calculate the “expected value” of perfect information in the first place – ie the calculation using probability he will bring this answer or that answer to pay someone for perfect information that he is going to bring either way if you pay him -and which will thus eliminate the need for the use of any unnecessary probabilities ? Why thus is the expected value of perfect information a necessity in the first place for calculating the fees of perfect information?

John Moffat says

Because with perfect knowledge the expected return in higher than without the information.

Asif110 says

But why do we need to study expected return of perfect information in order to pay this fees ? Why study probability when you want to pay info for not having probabilities

John Moffat says

You pay the fee and then you receive better information which enables you to make better decisions.

ty0311 says

Hi John,

I understand the logic behind the $300 as the maximum amount we are willing to pay for research, on the basis that the decision maker is risk-neutral and expects for a profit of $4,500. My question is, why do we have to compare the “average” guaranteed profit of $4,800 to the expected value? For instance, if the decision maker is a risk-avoider, so he decides to sign contract for 800 units for a profit of $4,400 (based on Maximin), then wouldn’t the maximum amount for research in this case be $400? ($4,800 – $4,400)

Many thanks in advance!

Regards,

Tim

John Moffat says

What you say is correct. However for the exam, we always calculate the maximum to pay for perfect knowledge based on the expected values (and I do state effectively what you are saying in my lecture).

ty0311 says

Thank you John! Just to check my understanding.

christafar says

Hello John,

I have a doubt about expected values. The customer contract is a fixed quantity so there no required to use the probability. So, isn’t it correct that we use the full contract value plus the probability of the expected market demand?

for eg: At contract 300 and demand 400, contract profit is 300×3, and demand profit is 400×0.2×5, so the expected value will be 1300?

John Moffat says

The problem is that they will. not always be able to supply the expected normal demand. For example, if they have signed the contract to supply 500 units, then the most they can sell to the normal demand is 700 units (because the capacity is limited to 1,200 units) regardless of the size of the normal demand.

pateladam says

Hi John, im abit unsure on how we get to putting 400,500,700,900 at the top and 300,500,700,800 on side, is there a particular ruling which needs to be followed when labelling each side and then making the calculations, such as when we do maximin and minimax we go sideways but when looking at minimax regret we go downwards? im just confused as I always label the sides wrong and because of that all my workings are wrong,

Azka. says

I have the same question.

John Moffat says

It does not matter which way round you draw the table. What matters is that you identify what it is that is uncertain and what it is that we have to make a decision on. Don’t learn it just as rules – make sure you really understand the basis on which we are making the decision under each of the methods.

tugcem says

Thanks a lot John. Very clear again.

Does the syllabus include sensitivity analysis and simulation?

Shivangi says

Isn’t that part of the FM syllabus? (Where simulation isn’t something we’re required to actually carry out in exam)

elenaakhra says

Oh dear… You speak so fast, I didn’t understand anything. It looks like you teach for students from english speaking countries only…

John Moffat says

Well you are completely wrong – I do all of my teaching in non-English speaking countries in Eastern Europe 🙂

I am puzzled that you only say this for this lecture. Have you not watched all of the previous lectures – if you have then you have presumably got used to my speaking? There is no point in watching a lecture out of order – the lectures are a complete free course and cover everything needed to be able to pass the exam well (and, of course, this lecture makes no sense on its own because it follows on from the earlier chapters).

marblserockin says

Hi John, I didn’t understand how you came up with expected value of 4500, if we signed the contract of 700, as I figured, profit is calculated like: 700*3+(1200-700)*5 = 4600, did I miss the point?

marblserockin says

Sorry John, Just figured I missed one lecture.

ervlin says

Hi John, I don’t understand why do we must use this way to calculate the EV if we using market research? How do we know to if this is the way to calculate the EV if we using market research to obtain perfect information?

alie2018 says

Thank you John.

John Moffat says

You are welcome 🙂