Optimal Selling price for stores

Guest · Post by **Guest** » 18.03.2007, 18:22

Well, I think you can easily make an argument that it may not be worth the effort, but I still disagree that you would have to waste time with empty stores. Personally, when I have things coordinated, none of my shops ever sits empty. Also, with the correct formula, you wouldn't need to play with the prices in the shop to see what gives you a good selling time or 35-55% happiness; these would be results of the analysis, not inputs.

As for the formula, the equation wouldn't change at all once you figured it out (the hard part). If you had the correct formula, then all you would have to do is take 2 minutes to punch in all the current numbers from the game and see the new optimum result.

But I think it is all just a thought excercise since I doubt any of us will actually try to make the equation

Guest · Post by **Guest** » 18.03.2007, 22:18

Im just gonna vote for PhoenixOils method. This early in the game micromanaging won't "win" the game for me.

Guest · Post by **Guest** » 19.03.2007, 01:18

I use excel sheet just to look at my cost, total production time, and if i have the sufficient resources to build it thats all.

For me, i just set a time when i (most probably) log in next, i.e. 25 hrs from now, then i will just set my production + selling stores to finish their tasks at ~ 25 hrs from now. I know it might not be the optimum price, but then, it works for me that i log in as i want not the other way round

Guest · Post by **Guest** » 19.03.2007, 08:23

PolymerTim wrote:Well, I think you can easily make an argument that it may not be worth the effort, but I still disagree that you would have to waste time with empty stores. Personally, when I have things coordinated, none of my shops ever sits empty. Also, with the correct formula, you wouldn't need to play with the prices in the shop to see what gives you a good selling time or 35-55% happiness; these would be results of the analysis, not inputs.

As for the formula, the equation wouldn't change at all once you figured it out (the hard part). If you had the correct formula, then all you would have to do is take 2 minutes to punch in all the current numbers from the game and see the new optimum result.

But I think it is all just a thought excercise since I doubt any of us will actually try to make the equation

The thing is, the formula will change, because the market changes. There are way too many factors to determine with supply, demand, average price, average quality, etc. All of this will effect the formula and where that 'optimum' point is at any given time. It will virtually change by the hour, and no one here has time to keep up with all the various factors that would change it.

With the '35-55' method, you can do the same thing, and it saves a lot of time. When I say 35-55, let me clarify, every product is different, some products, (like gas for instance) will sell most optimally at 34-38, other things, (like wardrobes and leather jackets) it may be 47-50, but once you determine what range or rate is best for that product, you know it forever and don't need a spreadsheet.

When you think about it, this makes sense... the number of people who think your product is too expensive, is an indication of the very formula you are talking about... in other words, Kapilands has already done your spreadsheet, and it is giving it to you in the form of this number. Of course, it's not a 'static' number, you can't say the 'optimal' rate is when 47 of 100 think your product is too expensive, it will vary for different products, but essentially, this is the determined by the same 'formula' or data compiled in a spreadsheet, it's just expressed in a more simple way.

As I say, I'm not criticizing those who want to waste time with spreadsheets, if that's what they like and they think they do better with them, by all means, go for it! I have just found an easier way to do the same thing, and I guarantee you, my profit margins will run just as high, or higher, than those with the fancy spreadsheets.

Guest · Post by **Guest** » 19.03.2007, 09:01

guys chill y u doin so much calculations its a game i like to have fun

well wat i do is i put gas in my gas station and sell it for 25 as it takes 24hrs so i just go check it after a day and again put gas in it sel for 25 and it takes 24hrs so its very simple instead of doin those calculations

Guest · Post by **Guest** » 19.03.2007, 09:27

Oops, HB you are so confident.

To Richbay:
Even if I think your effort should be encouraged, I am sorry to say that your hypothetical formula is a little bit too far from the right. Because if I sell a relatively higher Q product for the price relatively lower than the average, the time is 10 min. from your calculations, but it can't be. In addition, the X's square is useless, because the X in your formula is just a coefficient, not a variable. However, it is not a blame or criticism, because I also can't tell a right one.

As far as I know, the function should be:
T ((-)D, (+)S, (-)deltaQuality, (+)deltaP, (+)quantity, (-)Employee, (-)Ad.)

But I don't know how the exact formula should be, too, so I can't criticize anyone makes the mistakes by tryting to find one. Very sorry I didn't learn more in statistics area.

Guest · Post by **Guest** » 19.03.2007, 11:01

Richbay wrote: (^is probably missing something or just plain wrong!)

As an earlier poster mentioned, you may need to factor in population levels, employment and the fact that you are after (at least) 3 different co-efficients. This is before we factor in that the game creators may have put in random market fluctuation.

All of this is not important of course, we are not necessarily after the exact formulae the game uses but merely as good a representation of it as we can come up with. A 'best fit' line if you will. If you can get a reliable estimate then that is almost as useful as perfect information.

Guest · Post by **Guest** » 19.03.2007, 16:27

Oops, HB you are so confident.

Confidence is a Virtue!

Guest · Post by **Guest** » 20.03.2007, 00:56

Halee-Burton wrote:With the '35-55' method, you can do the same thing, and it saves a lot of time...

I agree, it is probably a very similar result for much less work. That's why I mentioned before that you could easily argue the regression method isn't worth it. The regression just gives you more specific information, which may or may not be useful to you.

Halee-Burton wrote:The thing is, the formula will change, because the market changes. There are way too many factors to determine with supply, demand, average price, average quality, etc. All of this will effect the formula and where that 'optimum' point is at any given time. It will virtually change by the hour, and no one here has time to keep up with all the various factors that would change it.

I disagree here, if the formula (model) is properly created, then it will not change with time. If it does change, then the model was never correct to begin with and only coincidentally gave the same results for the set of data it was modeled on. With enough data, an accurate representation of the true formula can be constructed that will not change as the inputs change. That is the whole point of modelling. As I said before, there will be inputs that must be updated each time, but this would simply be punching in a few numbers. I believe that the formula we are discussing would require no more than about 5 to 10 pieces of information that are easily available in the game (avg price, avg quality, population, supply, demand, etc...).

Halee-Burton wrote:As I say, I'm not criticizing those who want to waste time with spreadsheets,...

I would suggest that by using the word 'waste' you are criticizing

But again, I'm not arguing (and never did) that it would be worth anyone's time to perform the aforementioned regression analysis. It was only a comment that it is possible and my thoughts on how difficult it might be and what benefits it could provide. I myself am not planning to do it either.

Guest · Post by **Guest** » 20.03.2007, 00:57

Guest · Post by **Guest** » 20.03.2007, 01:53

HB - you've quoted PolymerTim's statement, but didn't nothing to address it. He said that production can still happen while you're number-crunching, which is clearly true, and you've again ignored it entirely. Those who do the math will have the advantage over those who do (unless you do it wrong)

And you seem to have no idea how the math will actually work. A linear regression would potentially produce a formula to determine sale time for any and all cases, so while the price will fluctuate, assuming someone ends up finding the correct formula, it's just a matter of plugging in the numbers and crunching them, and voila - increased efficiency forever.

I don't understand the air of anti-intellectualism in this game. And sure, you can get by and do well based on intuition and instinct alone, but I don't understand how anyone with a quarter of a brain can argue that doing the math will actually hurt you.

Guest · Post by **Guest** » 20.03.2007, 02:07

cygnus - you're correct in saying that price and time is on a straight line.

What you've calculated is revenue/time, which does increase as price lowers, because you sell faster and faster.

However, it's important to take into account that it costs money to buy or produce the materials to sell. What you've calculated is revenue/time, not profit/time, which is what matters.

Consider these numbers (I'm going to make them up for the purposes of this example, but trust me, this is how it works)

Sell price Sold/hr Revenue/hour
20 10 200
19 12 228
18 14 252
17 16 272
16 18 288
15 20 300

But now, take account that you paid a price to obtain the product:

Price paid Sell price Profit/unit Sold/hr Revenue/hr PROFIT/HR
10 20 10 10 200 100
10 19 9 12 228 108
10 18 8 14 252 112
10 17 7 16 272 112
10 16 6 18 288 108
10 15 5 20 300 100

So, in this simulation, the optimal sell price for this product is somewhere between 17 and 18.

This is the key to what this thread is trying to find. There's an optimal profit/hr rate to sell at based on sale time, and the price you paid for your product.

Guest · Post by **Guest** » 20.03.2007, 03:37

JuliusC wrote: What you've calculated is revenue/time, which does increase as price lowers, because you sell faster and faster.

However, it's important to take into account that it costs money to buy or produce the materials to sell. What you've calculated is revenue/time, not profit/time, which is what matters.

OK, JuliusC I see your point.

I

Guest · Post by **Guest** » 20.03.2007, 07:36

JuliusC wrote:HB - you've quoted PolymerTim's statement, but didn't nothing to address it. He said that production can still happen while you're number-crunching, which is clearly true, and you've again ignored it entirely. Those who do the math will have the advantage over those who do (unless you do it wrong)

And you seem to have no idea how the math will actually work. A linear regression would potentially produce a formula to determine sale time for any and all cases, so while the price will fluctuate, assuming someone ends up finding the correct formula, it's just a matter of plugging in the numbers and crunching them, and voila - increased efficiency forever.

I don't understand the air of anti-intellectualism in this game. And sure, you can get by and do well based on intuition and instinct alone, but I don't understand how anyone with a quarter of a brain can argue that doing the math will actually hurt you.

First of all, I'm not arguing that math will hurt you, I am saying you don't need math to make money and have fun. And I did indeed address PT's point. In order to use any "formula" you come up with, each time you go to put stock in a store, you will need to determine about 10 different variables which change by the hour. Some of these variables will alter the formula itself, and render it invalid. This process can not be done while the store is still selling, because all the variables will change the minute you restock, it has to be done precisely when you restock, to have any change of any "formula" performing as intended.

My point was, while you are busy trying to find the data to input into your formula, I am busy selling my products. Maybe I am off by a few percent here or there, but so what, I am selling product while you are playing with Excel. (...Not that either of us are actually earning a real income!)

A linear regression would potentially produce a formula to determine sale time for any and all cases.

Not at all. The linear regression might potentially produce a formula to determine sale time for a single product at one moment in time, but that is about it. Every hour, all the variables change, and any formula you presumed accurate 2 hours ago, is no longer valid. As population grows, there is more demand, as new players join the game, there is more supply, some people sell at higher than average market rate, some sell for less, this also changes constantly, and you have no way of determining what every player is selling every product for in every area. All of these factors go into the determination of an "optimal rate" at any given point in time, and this rate changes with the various conditions. It's virtually impossible to determine a set formula to calculate this 'optimal rate'. You can come relatively close, and if you want to go to all the trouble to be 'relatively close', that's fine, but I can be relatively close by looking at one statistic, my customer price satisfaction percentage.

I don't understand the air of anti-intellectualism in this game.

I could say the same for the 'air of intellectualism' in this game. It IS, after all, a GAME! Every person here is playing it because we enjoy the game in our own ways, and I am not trying to convince you that your way is not enjoyable for you, that is entirely up to you. I have merely stated what I have observed, and presented it for consideration here.

I don't think it's possible to produce a 'perfect' formula to determine the best selling price for all products all the time. There are way too many factors involved, and they are all changing constantly. Besides, the "formula" has already been established by the gamemaster, and the "result" is expressed in the form of a customer satisfaction percentage. So, what you are doing has already been done for you, if you understand how to interpret the result.

Guest · Post by **Guest** » 20.03.2007, 08:30

Sorry to keep going after you, HB, it's nothing personal, just a point I'm trying to get across.

Yes, a linear regression will produce a formula to determine optimal price at all times. You said that the optimal price changes all the time - this is true - but a formula would be found precisely to account for those factors.

The formula for sales time is something along the lines of

Z - Y(advertisement level) - X(average price - your price) + W(average quality - your quality) + V(% demand satisfied)

And the trick is finding V, W, X, Y, and Z. The goal is precisely to determine how all those factors affect sales time. Is it really hard to find a formula involving 5 or more factors? You bet, but it's possible, and that's part of the challenge that some people enjoy.

And if you'll observe the conversation I'm having with cygnus, it's more complicated that your customer satisfaction rate. When you take account the cost of purchasing or making a product, the optimal selling point becomes parabolic, and is very hard to determine using customer satisfaction.

When I talk about the "air of anti-intellectualism," I'm certainly not criticizing those who refuse to do math. You're right, it is a game, and people should have fun in whatever way they choose. The only thing that bothers me is when people criticize the extra effort that some choose to make, and say that it doesn't do any good. It DOES help, it DOES give an advantage, whether you care or not - just don't argue that it doesn't. And it's not you. I hear over and over again that you shouldn't bother with math, and it irks me the more I hear it.