Loot By Dynamic Pricing Using Microsegmentation

When you are hunting/mining/crafting and loot suddenly stops, do you:

  • Keep going, a Global/Hof is coming up!

    Votes: 190 57.2%
  • Stop, this loot is horrible!

    Votes: 142 42.8%

  • Total voters
    332
Joined
Feb 8, 2006
Posts
3,841
Location
Chicago, USA
Society
Project-Asylum Corp.
Avatar Name
Revan The Nightbird Evenstar
Before you read further, please answer the poll question first!


I created this thread to talk about EU using an economic framework. If this interests you, or you have nothing better to read, please continue and thanks for reading. :)

What motivated this thread is hearing frequent complaints from experienced players, such as: player A has to deposit too much per year to play, yet player B plays a lot but loses very little or even profits, player C opened a new account even though he shouldn't have, and finds that it is easier to profit with a young avatar, player D takes a break from EU, and finds loot is better upon returning, player E is a relatively new player, and feels as the months pass loot is slowly decreasing.

What this amounts to in my eyes is dissatisfaction over accounts seemingly being charged different amounts to play. Before I continue, please let me describe what Dynamic Pricing is using a simple example.

Scenario:

Ann is a baker, and just discovered a new recipe for a cookie that she knows is both unique and delicious, so she bakes 10 big cookies and some free samples, sets up a stand, and tries to sell them in one hour to get customer feedback. The cookies cost her 1$ each to make, and she wants to sell them at at least 2$ each since that is the price for similar sized cookies. Still, she thinks she can charge more because she thinks customers will love hers and can't buy them from anywhere else.

Over the course of the hour, 10 people who taste the free sample will be interested in buying a cookie. Of these 10, 3 people think that the cookie is good, but nothing special and are only willing to pay 2$. 5 think that it is very good and are willing to pay 3$. 2 think it is the best cookie they have ever tasted, and are willing to pay 5$.

How shall Ann price these cookies to make the most money when she doesn't know exactly how much each person is willing to pay?

Method 1: She puts up a price tag, and sets the same price for everyone. If she charges 2$, she will make 10$ from selling 10 cookies. If she charges 3$, she will make 11$ from selling 7 cookies. If she charges 4$, she will lose 2$.

Method 2: She doesn't put up a price, and when an interested customer asks for the price, she says 2$ if they describe the free sample as good, and 3$ if they say it is very good or better. She sells all 10 cookies and makes 17$.

As you can see, if Ann can make educated guesses based on information about the customer, in this case the customer's reaction, she can price accordingly to make more money. This is Dynamic Pricing. You see it done by car salesmen, real estate agents, insurances brokers, for airline tickets, even on Amazon.com, where people can pay differently for the same item.


Wait! I hope some of you will think, why not price the cookies exactly at what the customer is willing to pay? Then she could make 21$! The premise of the scenario is that Ann doesn't know the exact price each customer is willing to pay so she has to guess. She can't exactly ask each customer what they'd be willing to pay and then charge them that can she? :)


In Ann's example, 'method 2' splits the customers into 2 segments. In Microsegmentation (or First Degree Price Discrimination), 'method 2' is taken to the extreme and customers are charged on an individual basis due to nearly perfect information on their behavior. Instead of making another example, I will just use EU, which we are all familiar with.

In EU:

Taking the example from above further, it is clear that Dynamic Pricing is more profitable. Additionally, the higher the quality of information, the more precise the pricing will be. However, how does MA obtain such exacting information about each of us? The answer is: Why not from everything we do in EU?

Every player that has participated in the 3 main professions know, there are good and bad periods, and the general understanding is that it is a part of the Dynamic loot system. What it tells MA is, how poor of loot will we endure, how much are we willing to pay for the fun, before we stop. They obviously want to charge us as much as possible, since they are for profit.

For the purpose of this example, let's say good periods = 115% of tt spent returned as loot (since MA only deals in tt), medium periods = 85% of tt spent returned, and bad periods = 55% of tt returned.

For the players that play the same amount during good, medium, and bad periods, why not dynamically increase their medium and bad periods until their play time starts decreasing? If they purchase the same amount of gaming hours whether you charge 1$ or 5$ per hour, as a business, you charge 5$/hr. In fact, making them play less would be good for them too! Better health and social life!

For the players that play the most during good periods, less during medium periods, and stop whenever they hit a bad period, it might serve MA's interests to increase medium periods and remove the bad periods. If this significantly increases play time, MA makes more money.

Basically, if loot is reduced but play time doesn't reduce, then keep the change and lower loot more. If loot is increased and play time increases enough, then keep the change and increase loot some more (to a certain max of course, and I guess there has to be a minimum.. somewhere).

MA's ability to dynamically change loot distribution to see its effects on player participation, and collect information about the player, is virtually limitless. This gives them the ability to maximum their profit, on a player by player basis.

Dynamic!
 
Last edited:
Well, I read everything and I have to say I understand what you are trying to say..


But, most of what I've seen, people spend and spend, doesn't really matter if they have a bad period or not.

And as I have heard and think to be right, it's mostly about luck. Yes it is a dynamic system. But I really don't think that bad periods can't become really good periods. It's just a thing of...


Being in the right place at the right time.

Just my thought
 
Hmmm...the poll choices are too narrow for me to vote, would usually depend on more factors.
 
You seem to be presuming that the good/average/bad periods cycle based on time... they could cycle based on ped spent or a combination of the two.

Of course, that's presuming there is some sort of cycle set up at all. :)
 
Well I don't let the casino gimmics pull me in... If I got money to play, I play... if I don't have money set aside for EU then I don't play. I let my real life financial budget dictate my in game time.

Some months I have more money to play and some months I don't. I think I am dynamic and EU should deal with it...
 
Hmmm...the poll choices are too narrow for me to vote, would usually depend on more factors.

likewise...

it depends what you mean by loot stopping... if loot for a particular mob/area has been bad, I switch mobs or areas...

If i've been losing alot overall, i switch to a cheaper activity...

my budget and activity level remains relatively the same, and if I profit... i use the proceeds for upgrades.
 
When I started it seemed to be based on ped spent, while hunting in a team we would global every 70 to 80ped spent ammo used.

Its difficult to tell now, many hunt for several hours spending several thousand ped so it could be ped spent or by time or both.

I've had globals and a hof from switching from one mob to another but that was in the past, I usually just give up now and log out.

I've followed soc members out to areas after they've reported good loots from the mobs there only to get some of the worst results I've had for weeks so I just go with my gut, sometimes it works - globalled on the first atrox to attack me after tping out, hof'd on umbranoids on a spur of the moment switch and globaled on my first kill on atrax at which point I stopped as I knew I'd lose the profit if I continued.

I've tried camping mobs and after a week of low maturity daikiba trying to loot gremlin feet, in the end I gave up, it was getting a little too tedious to continue and it was cheaper to buy them.

There are some areas I know are just bad for me so I don't go there, west of Jason for instance is a just a bad loot area for me.
 
Hmmm...the poll choices are too narrow for me to vote, would usually depend on more factors.

agree , so i didnt vote
 
I think they have 2 simple slot machines system which NEVER pays out more then its put in. One is your personal and one is global slot machine and all your returns are paid thru those 2 at variable percentages.

Tweaking jackpot size is simple for a slot machine just as a slider for crafting. Also they may make loot more "dynamic" by making payouts more from global then your own slot machine system.

;)

I.

P.S.

Allways try to lean to most simple of explanations because it is most probable.
 
Allways try to lean to most simple of explanations because it is most probable.

Doesn't apply to business The plan that gathers the most profit is the most probable :)
 
Doesn't apply to business The plan that gathers the most profit is the most probable :)

Simple things are more robust and take less time to prepare therefore - usually most profitable in the long run due to low maintenance.

Loot mech is highly optimised machine because it has to be pinpoint precise.
I highly doubt they have the power to actually micromanage it to the point you described.

I.
 
Hmmm...the poll choices are too narrow for me to vote, would usually depend on more factors.

Yup, thats it :)
 
Simple things are more robust and take less time to prepare therefore - usually most profitable in the long run due to low maintenance.

Loot mech is highly optimised machine because it has to be pinpoint precise.
I highly doubt they have the power to actually micromanage it to the point you described.

I.

Dynamic pricing typically results in 50% to 200% more profit over simple pricing, and the ratio of additional maintenance cost to additional revenue is extremely low. Do a search of some examples, you'd be surprised :)
 
Last edited:
My feeling is that MA has a few personas(profiles) defined. Every avatar is assigned one of these personas, based on their actions. This way, they can increase the sense of "randomness" or "avatar luck" and optimize the income.

We all pay in one way or another for our globals/hofs/aths. It's just a matter of paying in advance or after we get it, paying faster or slower. This is why I think these profiles are used.
 
I do understand what are you saying and I will try myself to be also.First nobody folow you in any way to know somehow what are you doing in game.The game are based on mathematical analysis and combinations and if I knowing well is never changed from the day I start,but over the time I was the one who`s changed,getting older,loosing somewhere a little bit patience and according with the game status I'm the main factor who slowing down.
Most of the time not really the people spend their money based on the mathematical algoritms,analysis and combination because they said that is a game,right is a game,where I should make something fun,to enjoy the game and I'm sick of all mathematics,I want to play it for fun and now is the point that they are forgot that,because they spent a lot of money on it and they make a simple mathematics plans on how they spent,how they get back,etc.
Let me know if somebody is mad,exhausted or embarasing when he make some excercise of IQ test,all are relaxed,the mind are clear and ready for strike.When the mind strike everybody know`s that is most of the human kind pleasure and all of us are enjoyied about it.
Should you make an simple plan even you got peds or not in game,log in,make the plan for today:I go hunt alone-I'm lvl 34 in handgun-for this I need first of all the creature where I will like to go,for example daikiba,I will use an gun at the lvl of the mob,even I'm 34 an I will NOT use other one that will fix my powers for this lvl,because daikiba is much little that my powers,at my disposal.When you hunt a little mob don't even think to use an armour,the armour is only for creatures who can meet your kind of level,your kind of power that can be an response,reply from it.
When the balances are not equal that will happened what you name the bad time of play,is yours because you created yourself and nobody do it for you for decreasing your loot,or make you felling bad,or bad luck,bad day,etc.You should reflect on this and make some time for you before you start the game,don't blame anyone,make your plan from start,be patient only 5 min for yourself and should the game will be a pleasure,for sure you will enjoy it much than ever.;)
Are more to say and argue but my duty is calling me fast and I must go now, be the joy and smiling the shining star on your face.:)
 
And here I was thinking there was NO WAY someone could do a better job at driving costumers away than MA is doing already.

Making people pay more if they keep playing in a bad time?

Great... so those people who just keep playing in hopes of finally getting that big loot which may compensate their losses a bit will get even WORSE returns?

That's just stupid.
 
Even if MA can't set a constant price - people play and contribute different after all, there should be open information on exactly how much is going into MA's coffers from each person. If only for our peace of mind as customers. Hiding it is pretty much evidence that something like this is occurring.
 
I didnt get your intention exactly.
Do you think MA does it or should do? IMO the more profit MA makes the crappier the game becomes.

On the subject.. well it seems more money from the same customers with the same work done is a no brainer. But what if you pushed it too much and the customer leaves and no chance for you to lower the price again.
Also MA needs people to profit to sell their product to others and keep the community somewhat happy. I'm sure big profiters keep on playing when loose 25% profit (but still profit).

I think since you can play this game on many levels, people charge themselves already. If was willing to spend 1000 dollar each month i would be crafting big shit, use big L weapons on big mobs etc.
 
I just wanted to discuss price and the addictive nature of this game in an economic framework that culminates in profit for MindArk. This is just how they make money. I think I wrote the first post well enough to illuminate the process, and the proof is self-evident.


The question in the poll is simply something I've noticed in myself. The addictive nature of the game whispers: "push through the bad loot, and you'll hit a big one!" What I hear now is: "Hahahahaha (evil laugh), you have a gambling problem and I know it, so I takes your moneyz!!!!!!!" , so I log off. Now, I try to play in a way that values time in the game appropriately based on the entertainment I derive from it.

As a miner, I play with 45$ a run, which takes about 3 hours. I get 83% back in tt on average, let's just say 80% and MA nets 9USD, or 3 USD per hour. Over a bad period, I only get 50% back, which increases the cost to me to 7.5USD per hour. To me, that's twice the cost of a movie ticket, so I go do something else instead (like poke around EF, write in this thread, until I think of something better).

Everyone obviously values their entertainment differently, the above is just for me, so the intent of the thread is a FYI, For Your Interest :)
 
My own actual experience has been: Deposit; Spend what I've chosen to spend per month; Run out quickly, but see the myriad stacks of loot too small to auction get a tiny bit bigger; sweat for a month, repeat.

Initially, I saw all 3 loot periods, randomly mixed. After a couple months, I began to notice that when I return from a month of sweating, I see horrid loot until I have spent barely more than I earned from sweat - and then semirandom loot, but worse each month.

All that being said, I see no difference between this game's payout pattern and that of your average slot machine. Over time, it just gets more random and more frustrating.

I do see a huge correlation between avatar skill and success at any particular activity - just no correlation at all between that success and the value of its accompanying loot.

fwiw (gak I should get some sleep),
Medb
 
I'm rather surprised that everyone is so agreeable with my first post.

At the very least, I expected someone to say, "there's no way MA identifies people that spend spend spend and give them worse loot to profit more from their reckless behavior" :D


DOh! I know why, most people don't read to be end :/
 
Last edited:
I work for what is probably the leading company on the planet for developing algorithms and applications for Dynamic Pricing.

I am not sure that the behavior that you are identifying meets the definition of Dynamic Pricing.

First and foremost, it is considered illegal under discrimination laws in most countries to offer goods at a price determined upon the ability of a customer to pay. You cannot give cookies to poor people for $2.00 and then charge $3.00 to those who fit the profile of "wealthy". It would be patently illegal for MA to skew loot profiles on the basis of either deposit levels, time in game, or account values. You may collect such information to profile individuals who may be elligible for offers (think credit cards) but you must offer the same deal to all who qualify.

Secondly, Dynamic Pricing requires strict data on purchasing patterns at various prices as they relate to demand. I do not believe the correlation between in ingame payouts and deposits is anything other than linear. MA lets a certain percentage (probably 99+%) decay and item purchase prices roll right back into the loot pool. Depositing only has an inicintal effect once it is spent, and no direct effect on the depositor other than to increase the lootpool size for everyone.

Thirdly, Dynamic Pricing could be applied, but not in the framework you describe. MA could design an algorithm for NPC's to purchase and sell things like oil at an ever-changing market price, where high demand items would see their prices increase where low-demand items would decrease. But there would be no point, we already accomplish that through actual market mechanisms through the auction. There would be no benefit to simulate it.

While there is no question that MA collects a good deal of data on spending and playing patterns of it's users, I have no doubt it is for legal demographics and marketing purposes. It is not used for some malicious tinfoil black-hat algorith not even related to Dynamic Pricing that determines who gets loot and how much. That would be dishonest, and also completely illegal.

Additionally, anyone who keeps records will quickly learn that loot distribution has nothing to do with good times or bad times. It is exactly like the ball on the roulette table. Work out the standard deviation for yourself. I think you will be shocked. Go kill as many foul youngs as you like. As you kill them you will start accumulating bones. Match up your number of bones to the number of fouls you killed. You will find that the more you kill, the closer you come to the ratio of one bone per 4 fouls. 25%. Loot profiles are generated for critter upon spawn based upon a loot profile and then jiggered for the total lootpool size. It's pretty straight forward, and the easiest way to code it. Been done in Vegas for decades.

Nothing dynamic about it, other than the ever-changing lootpool size.
 
Last edited:
First and foremost, it is considered illegal under discrimination laws in most countries to offer goods at a price determined upon the ability of a customer to pay.

Secondly, Dynamic Pricing requires strict data on purchasing patterns at various prices as they relate to demand.

Thirdly, Dynamic Pricing could be applied, but not in the framework you describe. MA could design an algorithm for NPC's to purchase and sell things like oil at an ever-changing market price

Additionally, anyone who keeps records will quickly learn that loot distribution has nothing to do with good times or bad times. It is exactly like the ball on the roulette table. Work out the standard deviation for yourself. I think you will be shocked. Go kill as many foul youngs as you like. As you kill them you will start accumulating bones. Match up your number of bones to the number of fouls you killed. You will find that the more you kill, the closer you come to the ratio of one bone per 4 fouls. 25%. Loot profiles are generated for critter upon spawn based upon a loot profile and then jiggered for the total lootpool size. It's pretty straight forward, and the easiest way to code it. Been done in Vegas for decades.

Nothing dynamic about it, other than the ever-changing lootpool size.

Well,

1. It is not illegal to price discriminate, as per the examples I listed and more if you search.

2. We are purchasing hours of game play, and they charge based on % of amount spent returned in the long run. This is a perfectly good dynamic pricing scheme.

3. That would not be dynamic pricing, since dynamic pricing is based upon the customer's desire for the product. What you described is standard supply-demand pricing.

4. There is no loot pool.

5. I tracked my tt returns over 4 years, and it has been steadily decreasing. When I look at new players' tt returns, it is the same as what I had before, so the decrease, which is a fall from 90% to 80% on average, is due to some cause. My guess is it is due to my recklessness during long periods of game play on my part, but I can't claim to know MA's system as well as you. ;)

While there is no question that MA collects a good deal of data on spending and playing patterns of it's users, I have no doubt it is for legal demographics and marketing purposes. It is not used for some malicious tinfoil black-hat algorith not even related to Dynamic Pricing that determines who gets loot and how much. That would be dishonest, and also completely illegal.
 
I would just tell you the cookie was ok and maybe was worth $1.50.

That is how I shop at places with dynamic pricing. I would never come back from a test drive in a car and tell the car salesman "OMG that is the most awesome car ever! It's exactly what I want! Even the colour is perfect! So, what is your best price?"

The same with EU. I learned soon after I started that if I wasn't careful, $20 does not sound like much, but too often in a month could ruin the budget. Now, I deposit what I can and that is that. If I run out before it's time to deposit again, Tough tittie, I play for free (sweat, run around scanning things, etc.) or I sell crap in storage, or just don't play until the next payday and I can depo again. By your logic, MA should just stop giving me "bad" loot cycles, and I should have "good" cycles more often, perhaps even profit slightly from time to time. Then I could and would play more often and buy better gear.

hmmmm. You might have something there ......
 
Woosh, something flew over my head.
I had you until you finished with the cookie sales.
Then I just got hungry, :umn:
 
By your logic, MA should just stop giving me "bad" loot cycles, and I should have "good" cycles more often, perhaps even profit slightly from time to time. Then I could and would play more often and buy better gear.

hmmmm. You might have something there ......

It would work more like this, when MA gives you 80% tt back per month, you depo 20$ per month. When MA gives you 90% tt back, you play more and end up depoing 30$ per month, then it might be worthwhile for MA to let you have 90% back. (i.e. more good periods)

Woosh, something flew over my head.
I had you until you finished with the cookie sales.
Then I just got hungry, :umn:

:) I was hungry when I wrote this thread I guess.
 
5. I tracked my tt returns over 4 years, and it has been steadily decreasing. When I look at new players' tt returns, it is the same as what I had before, so the decrease, which is a fall from 90% to 80% on average, is due to some cause. My guess is it is due to my recklessness during long periods of game play on my part, but I can't claim to know MA's system as well as you. ;)
Which basically boils down to:

Better dump your avatar from time to time

?

Tussi
 
Which basically boils down to:

Better dump your avatar from time to time

?

Tussi

Well, this wasn't meant to be a "this is the way you should play" thread, but rather a guess at MA's business plan. I've always wondered why they are so vague about how they charge players. Combine this with the experience of veteran players and I put together a business model that takes advantage of detailed player behavior information, results in different average returns, and maximizes profit for MA.

It made sense for MA to price dynamically. If players are randomly assigned to good, average, and bad loot, then bad things would happen.

1. When risk averse players (people that don't like to gamble) get bad loot, they deposit less, play less, and might even leave the game.

2. Risk-loving people will play more if they get good loot, but they will deposit less for the duration as well.

Both these cases will result in less income for MA, something they don't have to risk (haha) if they make use of each player's risk tolerance.



Now all this talk is pretty worthless if nothing practical comes out of it. To be honest, nothing much I think can be done about it. If MA keeps track of information by name rather than avatar, then a new avatar wouldn't change anything since you will need to register a card. If they didn't, you'll get a reset but still lose your old avatar.

I guess you can also fudge your own behavioral information by playing like a very risk averse person. This is unfortunately very difficult to do, and you would have to keep it up for a long time. Myself, I have 5 years of bad behavior I would need to overcome if I start now LOL :)

Here is my old mining journal, compare it with my current diary in my signature. What a difference!

https://www.planetcalypsoforum.com/forums/mining/82690-nightbirds-mining-journal.html


Sorry, I can't be of help. Maybe someone else knows of a surefire way but I don't. ;)
 
Interesting theory ... but completely wrong, IMO.

The main flaw is the notion that returns are the only factor influencing play-time. Whereas in reality many other things play a much larger part (RL commitments, RL finances, enjoyment/boredom of EU, other leisure activities, access to PC etc. etc.)

Your theory has no room for these other factors and therefore MA's whole loot allocation algorithm would be based on incorrect data.

It sounds like a very overcomplicated scenario in any case, and I dont see any benefit to MA in something like this over a much simpler one, like 'the loot pool', for example.

Also, I believe Falkao's testing on average returns (BIG thread on this somewhere) would disprove this theory. If I remember right, average returns followed a normal distribution for all tested avatars.

EDIT:
Where I DO see some use for MA in this theory is in setting decay levels for new items, based on extensive user profiling of existing items.

ie If we introduce a new high end shortblade, what do we know about how skilled shortblade users currently use their equipment? How long do they hunt for? What do they like to hunt? etc. etc.

2nd EDIT:
Which if you think about it, is closer to what you were intending to discuss anyway, which is dynamic pricing.

I believe decay falls into the pricing category, but looting does not.


(sorry for the multi-edits)
 
Last edited:
Back
Top