All skills are doom to diminish if they are not used. Because skills are diminished over time when they are not in use, this game therefore keeps the player constantly using the skills they learn, so you are correct that it is not a measurement, but more of a relection.
The problem is that if they cannot pass the test in real life, then they cannot pass in the game. If they can pass in the game, then they can pass in real life. Remember that the cutoff of this game is at 96%, which is higher than any schools. Passing in this game is not mere passing, but a proof of mastery of the skill at the near past (you can only test the past, and know of the past, there is no such think as knowing the present when you get into higher science), and that is why I use the word "measure," even though it is not technically correct.
Edit: Sorry, but this design tends to be a very broad topic.
I will make a difference in the economic system. In this game, you can only buy from NPCs, you cannot sell to NPCs. Trading will be allow, and there will be a logarithmic scale that measures the ratio. This scale will show how close it is to market value, where 0 is at market value. The tradings that occur will allow an update on the bulletin to the pricing. Players will therefore control the price of items by their supply and demand.
Items/Equipment/etc. consist of Common, Uncommon, and Rare. Both common and uncommon items will wear out over time. Rare items are not sold by NPCs and last forever. The price of items that NPCs sell also fluctuate base on how many the players already own in proportion to the amount of characters, and the current market value. Common items are sold at every location, and uncommon items are specific to higher level settlements.
What would you say is the range of fairness for pricing?
[Edited by - Platinum_Dragon on April 1, 2009 5:39:57 PM]
Levels of Abstraction
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
- move
- mask
- shift
- move
- mask
- shift
- or
So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
I now understand what you mean. When you mention a percentage like 96% you didn't mean anything related to an actually letter grade. When I say "Pass", I didn't mean getting a "C or better" either.
Actually by PASSING I meant performing at 100% during one test. The reason is that since in our setting, there is much lower administrative cost for the player to take the test (for most of the test that I imagine). Since there is no burden on the system, a player could retake the test n times. You might as well make the player score perfect. For example, when you played the 7Segment puzzle, you scored perfectly. So that would be a PASS. If tests are that short that it would work ok because players would couldn't pass a certain question would restart the module and re-do the easier problems to learn/refresh the knowledge necessary to solve the later problems.
The two methods are equivalent. Sometimes it is easier to implement as one test that you score, and when a player gets "above 96%" you would pass him. Sometimes it is easier to order the question in a sequence so that the players that complete the sequence can pass. In the second method you don't have to consider how to balance the points to make it fair. Although I also understand that you could have mentioned 96% only figuratively, without implying that each test must be a scored test instead of scoreless tests.
Actually by PASSING I meant performing at 100% during one test. The reason is that since in our setting, there is much lower administrative cost for the player to take the test (for most of the test that I imagine). Since there is no burden on the system, a player could retake the test n times. You might as well make the player score perfect. For example, when you played the 7Segment puzzle, you scored perfectly. So that would be a PASS. If tests are that short that it would work ok because players would couldn't pass a certain question would restart the module and re-do the easier problems to learn/refresh the knowledge necessary to solve the later problems.
The two methods are equivalent. Sometimes it is easier to implement as one test that you score, and when a player gets "above 96%" you would pass him. Sometimes it is easier to order the question in a sequence so that the players that complete the sequence can pass. In the second method you don't have to consider how to balance the points to make it fair. Although I also understand that you could have mentioned 96% only figuratively, without implying that each test must be a scored test instead of scoreless tests.
Each question is measure in grade level. As you can see, at the lower education level: see math: you will see that addition is a simple thing when you are at a high level, but to first graders, you will find it at your difficulty level. What the test measures is the Grade Level of the person. So you can say the average person is 6th to 10th grade level performance, and this shows that the average person even though they complete 1 year of collete, has a diminished intelligence that they perform at 8th grade level. Each question can measure the exact grade level of the person, and then we will have automated questions at the particular level of each player. You have to perform at a certain grade level, say 16th grade level, to be able to do certain things. Of course, education has only 20 grade levels (20th grade is doctorate degree, 19th grade is engineer, 18th and 17th grade is master, 16th grade is bachelor, 14th grade is associate/foundation, 12th grade is highschool, and you know the rest). The test have to be able to measure this grade level performance, so that even if you know a lot about something, but if you are unable to do something in the time constraint, then you cannot do anything. In a life-death situation, you will have to have someone who can diagnose the cause of fatal injuries, or you will not be able to say that person's life. To ensure that the players are able to do something, that is why there is a 96% cutoff.
The problem with your 7segment analogy is that if you give a limit amount of turns or actions persay, then perfect would not only be to solve the puzzle, but to solve the puzzle in a limited amount of turns. If you make it so that you allow two actions more than the fastest route as the total actions for each stage, then the amount of players succeeding your puzzle decrease drastically.
In my game, there will be a limit of 50 seconds, and for each seconds pass beyond the time limit, the player loses 1 point toward a complex grade level measure. The longer they take to answer the question, the lower their score will be. I don't know about you, but a Calculus problem takes me 5 minutes, and making the time limit to 50 seconds will ensure that I cannot perform at Calculus level, in other words, my grade level is probably pre-Calculus in this game. The mechanics will have nearly 100% chance of underestimate the player's ability, and this is the goal of the mechanic.
In my mechanics, you can take my test as long as you want, because the entire testing statistics is maintain, and that players have to pass 96% inorder to enter the next grade level questions. Each test is summative to the previous level, so that questions of previous grade levels are implemented within the questions of the higher grade level. As long as the player has higher than 96%, the they can access a higher grade level, But say if someone fails grammar, a second grade level question, that is implemented in a short answer question, then they will have to fall back to second grade level regardless of what grade level they are currently at. Failing to add when you are in Calculus may even bring you back to first grade level questions! And you will have to bring back up to your grade level by increasing back to the 96% barrier.
Questions are broken down to the nearest 0.1 grade level to ensure accuracy of your skill.
[Edited by - Platinum_Dragon on April 2, 2009 6:29:39 PM]
The problem with your 7segment analogy is that if you give a limit amount of turns or actions persay, then perfect would not only be to solve the puzzle, but to solve the puzzle in a limited amount of turns. If you make it so that you allow two actions more than the fastest route as the total actions for each stage, then the amount of players succeeding your puzzle decrease drastically.
In my game, there will be a limit of 50 seconds, and for each seconds pass beyond the time limit, the player loses 1 point toward a complex grade level measure. The longer they take to answer the question, the lower their score will be. I don't know about you, but a Calculus problem takes me 5 minutes, and making the time limit to 50 seconds will ensure that I cannot perform at Calculus level, in other words, my grade level is probably pre-Calculus in this game. The mechanics will have nearly 100% chance of underestimate the player's ability, and this is the goal of the mechanic.
In my mechanics, you can take my test as long as you want, because the entire testing statistics is maintain, and that players have to pass 96% inorder to enter the next grade level questions. Each test is summative to the previous level, so that questions of previous grade levels are implemented within the questions of the higher grade level. As long as the player has higher than 96%, the they can access a higher grade level, But say if someone fails grammar, a second grade level question, that is implemented in a short answer question, then they will have to fall back to second grade level regardless of what grade level they are currently at. Failing to add when you are in Calculus may even bring you back to first grade level questions! And you will have to bring back up to your grade level by increasing back to the 96% barrier.
Questions are broken down to the nearest 0.1 grade level to ensure accuracy of your skill.
[Edited by - Platinum_Dragon on April 2, 2009 6:29:39 PM]
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
- move
- mask
- shift
- move
- mask
- shift
- or
So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
What do you mean by a range of fairness of prices?
[Edited by - Wai on April 1, 2009 10:10:53 PM]
The fairness of price is what percentage above the market value are you willing to pay. That way, I will have a logarithm scale that shows up the "fairness" of the trade when players trade with each other. At the 0 point is the current market value. log_unknownBase(marketValueOfOtherParty/marketValueOfYourGoods). I want to have a particular base for the logarithm.
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
- move
- mask
- shift
- move
- mask
- shift
- or
So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
I think that number depends on the item.
Suppose I only want to buy something when it is on sale, say 10% off.
Then my fairness value would be:
log10( 100/90 ) = 0.045
How do you use this number?
Suppose I only want to buy something when it is on sale, say 10% off.
Then my fairness value would be:
log10( 100/90 ) = 0.045
How do you use this number?
Well, won't 10 be too large of a base? I think that 10 is too large for a base.
If you say 10% then it's more like this:
log_1.1(110/100) = 1
log_1.1(100/110) = -1
So we can say that if the absolute value of the number is less than 1, then the trade is fair. If the number is positive, then you get the gain, but if it is negative, then you will lose in the trade, but it is still within fair range. When the absolute value is larger than 1, and the other player force you to trade, then you will have the right to put a bounty on the head of that player's character.
If you say 10% then it's more like this:
log_1.1(110/100) = 1
log_1.1(100/110) = -1
So we can say that if the absolute value of the number is less than 1, then the trade is fair. If the number is positive, then you get the gain, but if it is negative, then you will lose in the trade, but it is still within fair range. When the absolute value is larger than 1, and the other player force you to trade, then you will have the right to put a bounty on the head of that player's character.
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
- move
- mask
- shift
- move
- mask
- shift
- or
So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
What if someone wants to buy my item at 120% fair market price but I
don't want to sell? Will he be granted a permission to kill me?
don't want to sell? Will he be granted a permission to kill me?
During the trade, there is one player who accept the condition before the other player, so if they buy it at 120% of market value, then it is their fault to make a lost as long as they accept the condition first, so make sure the condition is within fair value. The market value also change based on the series of recent trades that occur throughout the game by other players, and the buying from NPCs.
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
- move
- mask
- shift
- move
- mask
- shift
- or
So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
In this game, players can steal from other players at anytime, but to protect the weaker players, there is a bounty system.
In this bounty system, there are four ways to increase your bounty
1) Steal = immediate increase
2) Unfair Trade = player with lost must report, and pay the fee
3) General declare war without warning = bounty is based on the settlement value.
4) Kill someone with an honor stat = bounty increase is equal to that player's honor stat.
Honor stat
1) Kill someone with bounty = honor increase equal to bounty
2) Declare war on Generals with bounty, and kill them in battle
Honor = -Bounty
There is a bounty/honor stat that is visible for all players. The more bounty a player collects, the more honor they have. The more bounty a player has, the more valuable it is to take his/her head.
In this bounty system, there are four ways to increase your bounty
1) Steal = immediate increase
2) Unfair Trade = player with lost must report, and pay the fee
3) General declare war without warning = bounty is based on the settlement value.
4) Kill someone with an honor stat = bounty increase is equal to that player's honor stat.
Honor stat
1) Kill someone with bounty = honor increase equal to bounty
2) Declare war on Generals with bounty, and kill them in battle
Honor = -Bounty
There is a bounty/honor stat that is visible for all players. The more bounty a player collects, the more honor they have. The more bounty a player has, the more valuable it is to take his/her head.
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:
- move
- mask
- shift
- move
- mask
- shift
- or
So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
Recommended Tutorials
Advertisement