Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Brightdance

Get Rid of 50% Bonus

Recommended Posts

24 minutes ago, topgeza said:

frRY2A0.jpg

Whilst carrying out this test I found it easier to achieve a Great Success or Amazing Success when experimenting 1 pip at a time, the more pips used at once reduced the chances of a higher level of success. You would think that by taking a higher risk the rewards when getting a Greater or Amazing success would be beneficial. The table shows this not to be the case as if you take the improvement result for Attack Power from a Great Success @ 1 pip experimentation 0.193 and *2=0.386 *3=0.579 *4=0.772 *5=0.965 *6=1.158 You can clearly see the results match those in the table for pips 2-6 . Given these results I fail to see the benefit of putting anymore than 1 pip in at a time when crafting as I can achieve between Great and Amazing Success when doing this, where as putting all 6 pips in I struggled to get to Great Success and never achieved Amazing success (30 bars were crafted in total and 13 experimentation points were put into each bar). So crafting @ 1 pip at a time I would achieve a bonus on average of 1.06 added to the base stat for the bar of 5= 6.06 Attack Power, If I put 6 pips in the best I could achieve if I hit a Great success would be a bonus of 1.159+base stat of 5 = 6.159 a difference of 0.099. Not a risk worth taking in my opinion especially when I would get a result of Moderate or Good success when experimenting with 6 pips which would give total results of 5.369 Moderate or 6.014 Good success compared to a more consistent average result of 6.06 per bar by experimenting 1 pip at a time.

Are you aware of the bonus for experimenting with 50% risk? It is substantial. If you do 1 pip at a time you can't get that bonus and your item will never reach full potential. By doing from 1 to 6-pip experiments with 13 available, you never reached the risk bonus. At 13 pips you'd need to do a 7-pip experiment to see the bonus.

If you are not reaching 50% risk, then yes, do 1 pip at time. If you want a chance for the best possible result, you should be doing 50% risk.

Edited by Jah

IhhQKY6.gif

Share this post


Link to post
Share on other sites

The source of the risk factor is not just how hard an individual risk is, but how hard accumulating the same results with more risks taken, as opposed to one big risk.

Quote

You can clearly see the results match those in the table for pips 2-6 .

Correct.  The more pips used reduces the chance of a higher level of success, but you are taking less chances.

Let's keep it to two pips.

If my chance of amazing success is 10% with a single pip, the chance of getting two amazing success in a row is (1/10) * (1/10)  = 1/100 or 1%.

So in order to properly scale the risk up if I select two pips at once, the chance to get an amazing success would have to be only 1/100.  If it is better than that, say 1/20, or 50% less chance every pip added,  because 1/100 would "feel" terrible to the user experience, it is far less risk mathematically to chose two pips than one.

Now take that out to 5 pips.  You would have to make that chance (1/10) * (1/10) * (1/10) * (1/10) * (1/10) = (.1^5)  or 1/100,000 (.001%) for the chance of getting 5x amazing success rolls in a row vs 1 amazing success roll worth 5x the value.

Now do the math on 13 pips.  How infrequent would you have to make even a 50/50 (coin flip chance of amazing success) on a 1 pip chance, at the 13 pips of chances.  Answer (.5^13) (1/8192) (.012%)

If your going to be working towards a blueprint, there really is no other option than max pips every time. The 50% bonus that could trigger a value of 1 roll 7.5X the value over 5X value, makes it even worse from an odds point of view, not to mention there is literally no way to get that 7.5x value, other than taking the least absolute risk option.

The 50% risk bonus is completely backwards mathematically.

Edited by KrakkenSmacken

Share this post


Link to post
Share on other sites

Thanks for the feedback guys, I was unaware it was only over 50% I had been informed there was a chance at other risk percentages below 50% hence why I compiled the table. Your information has been valuable.

Share this post


Link to post
Share on other sites

Just the fact that it is so completely non-intuitive shows that something is wrong.  This particular mechanic certainly does not add to my enjoyment.  And no I do not think blueprints and factories will make up for it.

Maybe a system of successive refinements would be better.  I don't really know, but the crafting quality mechanic is not there yet.

Share this post


Link to post
Share on other sites

I understand that there is some frustration when the RNG fairies are feeling unkind and you produce some piece of crap.

At the same time, I strongly believe that there needs to be a strong element of luck involved here.

  • Take away the luck and everyone is creating (virtually) the same thing all the time...
  • There will be no differentiation and things are going to get really tedious.

To me, the current system feels decent and I think that its going to get a lot better once we have blueprints.  At that point in time, there will be a much more clear distinction between what i call "Production" and "Prospecting"

  • Production is making finished items that people are going to use and will involve blueprints
  • Prospecting is making exceptional items that are worthwhile to turn into blueprints

To some extent, prospecting is going to look a lot like real world prospecting.  You're going to wade through a bunch of dross in order to get to those rare nuggets of gold

Here's the thing...  If we want crafting to be special...  If we want people to be known for the quality of what they produce, then crafting needs to be hard.  If anyone can be a master smith then being a master smith doesn't actually mean anything. 

 

 

 

 

 

 

 


WAZ6Fov.png

"The cinnabar is a lie"

Share this post


Link to post
Share on other sites
2 hours ago, narsille said:

I understand that there is some frustration when the RNG fairies are feeling unkind and you produce some piece of crap.

At the same time, I strongly believe that there needs to be a strong element of luck involved here.

  • Take away the luck and everyone is creating (virtually) the same thing all the time...
  • There will be no differentiation and things are going to get really tedious.

To me, the current system feels decent and I think that its going to get a lot better once we have blueprints.  At that point in time, there will be a much more clear distinction between what i call "Production" and "Prospecting"

  • Production is making finished items that people are going to use and will involve blueprints
  • Prospecting is making exceptional items that are worthwhile to turn into blueprints

To some extent, prospecting is going to look a lot like real world prospecting.  You're going to wade through a bunch of dross in order to get to those rare nuggets of gold

Here's the thing...  If we want crafting to be special...  If we want people to be known for the quality of what they produce, then crafting needs to be hard.  If anyone can be a master smith then being a master smith doesn't actually mean anything. 

The 50% bonus has nothing to do with being able to reach these goals or not.  In fact, the way it is set up now with the 50% bonus, ALL prospecting is going to all pips, all the time.  There is going to be zero variety.   You will have your second bullet point in spades.  

The math is the math, and the risk is backwards so the math always comes out to the same answer.  Your best odds are clearly to put all the pips in at once.

The choice would be much better if the risk went up as you added pips in one at a time.  The more chances in terms of experimentation tests you take, the higher the risk. 

Have this a floating bonus rather than a static 50%, and you give the crafter who has 10 pips a tough choice.  Do I experiment with that final tenth pip on a item that is really good, and risk a critical failure, or keep going.

The interface, the model, and even the feel of the crafting does not need to change, only the math and application of risk timing. 

I might throw together a graph on the risk/reward dynamic this test cycle.  With combine failure punishment gone, I am willing to spend time working with what is currently available to see if there is a simple, mathematical improvement that can be made, and sped up training makes this possible.

Edited by KrakkenSmacken

Share this post


Link to post
Share on other sites

I've been thinking about this issue and ways in which it could be improved. Risk percentage being tied to total # of pips does some good things, but also causes some serious problems.

Pros:

Because risk is a percentage of total pips, the more pips you have, the more success you will have when spending 1 pip at a time. If you have 5 pips, spending 1 pip represents 20% risk. If you have 10 pips, spending 1 pip represents 10% risk.

Allowing crafters the ability to determine the level of risk provides some measure of risk vs. reward, but as currrently implemented, it doesn't work as well as it could.

Cons:

An odd number of pips means you can only get the bonus once, and at higher than 50% risk. Having 11 pips instead of 10, or 13 instead of 12, should be a good thing. It currently isn't.

Having more pips than the item has experimentation slots also means you can only get the bonus once. In some cases, you can't get the bonus at all. If I am crafting a green quality elven leg and I have 15 pips, it is impossible for me to reach the 50% threshold since the leg only has 1 stat and green quality has a total of 7 experimentation slots.

 

Proposed solution:

Create 5 "bonus" levels. Level 1 is the base level. It's what you get when you spend 1 pip at a time. Level 5 is the current 50% bonus (or some amount lower than that, if the proposed changes would go beyond the overall stat budget). Levels 2, 3 and 4 would be incremental bonuses between those 2 numbers. Crafters will only need to spend 5 points to roll for the highest bonus. This means that players with the maximum number of pips (15) will have the chance to get the highest bonus 3 times. This will also make crafter choice much more meaningful than the current system.

Add a display to the UI that shows the current bonus level. The bonus level would be dependent on the number of pips spent, up to 5. This would give the crafter a much clearer indication of how the bonuses work than the current system where there is no indication of the 50% bonus. Spending more than 5 pips at a time should not be allowed.

Calculate the total risk as a percentage of total pips + 3 per pip spent, up to a maximum of 100%. The numbers would look like this:

jCxz1JJ.png

 

With this system, it won't matter how many experimentation lines an item has or what quality it is.

What does everyone think? Can you see any issues with this system?

 

Share this post


Link to post
Share on other sites
9 hours ago, Arkade said:

<Snip>

 

So if I am reading this correctly, you want to base risk on the line total potential, not the players total potential. 

That does have some merit to solving the problem of better crafters making worse gear at the low end, but it does not solve what I think is the real problem. That it must ALWAYS be better odds wise to go "all in", in your case spend 5 points if possible, to trigger the bonus.  In short, it does not overcome the [(Odds) To The Power Of ^ (Tries)] problem.

E.G. The odds of getting a 1/10 chance event (if amazing success is that high when selecting one pip at a time), 5 in a row would still only happen 1/100,000 (1/(0.1^5)) times, while there is no way at all any individual try is going to have that low of odds when you have selected 5 at once or almost every try is going to need result in a fail to balance against the O^T math. Even the least likely success would only be possible 1/3 times if the 1 pip at a time had an 80% success or better rate.

Here is a copy of what I posted for my solution in the developers forums.

It can be summarized in three changes. No new tables, no tricky math, just two calculation changes about how risk is determined and the bonus received, and two small features.

  1. Change the calculation for the risk value from the amount of pips you are spending when you push "Run Experiment", to the amount of pips spent in previous experiments on the item and previous gains in the row.
  2. Add an "Increase Risk" button. This button spends pip(s), but does not perform any experiments.
  3. Add a "restart experiment" option.  This option would unset all currently applied pips and set earned values to 0, but the pips would still be spent.

 

 

On 5/18/2018 at 6:35 AM, KrakkenSmacken said:

I tried very hard not to re-engineer the wheel with these, like I have with previous suggestions.  I am hoping that by sticking mostly to math, and two small possible feature additions, that these fall in the realm of possibility. It could seem complicated because of the math in the explanation, but the system changes really seem quite minor.

Suggestion 1: Change the calculation for the risk value from the amount of pips you are spending when you push "Run Experiment", to the amount of pips spent in previous experiments on the item and previous gains in the row.

In essence you are not considered to be risking what you are spending now, as what you have already have used up and earned. This is the most critical change, and I hope the simplest to implement.  The others I think would make the game more interesting.

This solution converts the game from being like a single pull slot machine with the only way to get to the "bonus" being betting max amount on all lines, into being more like blackjack where you are guessing when to stop.  The first experiment is done at 0% risk, no matter how many pips are spent. People in a rush, or wanting to take the least risk/time, or produce the simplest items that are made the fastest, simply go all in on the first throw, with no chance of a risk bonus.

If like I did and you have 8 pips to spend, and you spend one pip on your first try, your second experiment is now done at 12.5% risk. With the current risk bonus model, to get a bonus you have to have already spent or got lucky with close to 50% of your pips to trigger it. I would also increase risk partly on the success of previous attempts in addition to spent pips. What is working against this is the increasing level of risk, so the reduction in odds of a positive outcome. 

The more pips you have to work with, the less pips you have have to spend if you use the compounding model below. Players therefore make choices, and try to plan the best risk solution as they move forward.  The more they have put in, the higher the risk, the more incentive to stop "hitting" and asking for another card (when thinking in terms of blackjack) until they bust out. At some point the risk will trigger fear/caution, and they will all in on the remaining pips hopping for a big score or stop to prevent the "Critical Failure" chance from climbing too high.

Players choose with each pip spent if they are going to risk what they have already earned, or stand pat. I see some of this thinking with the critical failure mechanic, so I know it was part of the thinking in the original design. This also lets players pick what secondary feature they want to put the highest risk/reward outcomes into. Do you build up risk on durability, and then really go for it with grade, or spread the risk out by alternating between options?

If you had 37.5% risk from spending three pips, and those three pips had generated a great success (19.32%), your current risk would be (37.5+19.32)=56.82 and a single pip success would be worth (1.05 * 1.5682)=1.64 while a good success would be 5.64*1.5682 = 8.84.

I personally really like this option, because it varies both the bonus and risk based on previous successes, and gives players choice about when to stop pushing their luck and to go for all the remaining pips at the current risk and bonus, or keep trying to compound previous successes and increase risk for more bonus, and really push for the maximum. If anything will make items unique by not having standard best choice. 

I ran some numbers on how that would look with each pip given each possible outcome after 8 pips compounded 1 pip at a time. Any pair or multiple pip selections after the 50% bonus trigger point, would result in significantly lower final numbers. There is obviously nothing stopping you from putting a 100% risk cap into place, with at simple 2x bonus total cap, to keep compounding from getting insane.

vEfo86n.jpg

(I can give you the spreadsheet and formulas I used to work this out if your interested.)

Suggestion 2:  (Dependent on suggestion 1 being implemented.) Add an "Add Risk" button.  This button spends a pip, but does not perform any experiments.  This allows you to raise your risk and risk bonus before experimenting with any pips.  This also would partly solve issue 6 (above) that better trained players are usually making worse items with lesser quality materials.  A player with 8 pips available, and only 2 pips worth of experimentation due to material quality, could "Add Risk" 6 times to 75% before trying the 2 pip experiment with their last two pips. A player with only 5 pips could hit 60% risk before starting on the same item.


Suggestion 3: Add a "restart experiment" option.  This option would unset all currently applied pips and set earned values to 0, but the pips would still be spent. Basically the player could start over as if they only had the remaining pips to begin with. Again this is mostly for items that have less experimental options than a player has pips. So in the example of a player with 8 pips and only 2 pips worth of experiment to work with, they could effectively try to experiment on the same item 4 times, with each time having less pips. They could keep spending 1 pip at a time, and then restart until they got that perfect first pip and tried to follow it with another. Combined with "Add risk", all sorts of additional options and choices to play with open up.

Well there it is.  I hope it was worth your time to read.

Cheers


 

 

Edited by KrakkenSmacken

Share this post


Link to post
Share on other sites
On 4/27/2018 at 1:28 AM, Jah said:

Are you aware of the bonus for experimenting with 50% risk? It is substantial. If you do 1 pip at a time you can't get that bonus and your item will never reach full potential. By doing from 1 to 6-pip experiments with 13 available, you never reached the risk bonus. At 13 pips you'd need to do a 7-pip experiment to see the bonus.

If you are not reaching 50% risk, then yes, do 1 pip at time. If you want a chance for the best possible result, you should be doing 50% risk.

 

That is why when I listen to these guys on youtube/twitch say 2 pips at a time, I just laugh.  2 pips and a great success will not increase the amount gained per pip but will increase the risk.  

Edited by Mr.Kurtz

Share this post


Link to post
Share on other sites
On 5/20/2018 at 12:06 AM, KrakkenSmacken said:

So if I am reading this correctly, you want to base risk on the line total potential, not the players total potential. 

No, risk would still be dependent on how many total pips you have (plus how many you are spending at once), but the bonus wouldn't be at 50%. The max bonus would only occur when spending 5 pips and the overall risk would be higher or lower based on how many pips you have.

Let's say we have two crafters. One has 6 pips and the other has 12 pips, and they are both crafting the same white quality item. The 6 pip crafter could spend 5 points at once and try for the max bonus, but his risk would be 98.33%. The 12 pip crafter could do the same thing, but his risk would only be 56.67%. The 6 pip crafter is very unlikely to be successful in this scenario, so instead of spending 5 pips at once, he decides to spend 3 pips at a time. Now his risk is only 59%, but he needs to get 2 good rolls and he will only get the level 3 bonus instead of the level 5 bonus. Or he could do 2 pips at a time for the level 2 bonus, at 39.33% risk, but he would need 3 good rolls.

In this system, having more pips is always an advantage because there is no arbitrary risk threshold that needs to be met. Crafters simply spend 1, 2, 3, 4 or 5 pips and the risk is based on how many total pips they have and how many they are spending. Bonus levels are dependent on how many pips are spent at a time.

I get what you are saying with regards to the odds of someone rolling 1 pip at a time and getting an amazing on each. Yeah, that's extremely unlikely, but I don't think that's a problem. I think it fits with the risk/reward model. If you are only spending 1 pip at a time, you aren't risking much. Your final result won't be the best, but it won't be the worst either. If you are risking more by spending more pips at a time, you have the potential to produce something better, but also something worse. 

Share this post


Link to post
Share on other sites

I don't think this discussion is ready to be had until mass production comes online. Blair stated that the bonus system will make more sense once we get factories in game. I cant remember if he said it in a stream of a post in the Dev forum, but it was recent (this year).


Share this post


Link to post
Share on other sites
23 hours ago, Arkade said:

No, risk would still be dependent on how many total pips you have (plus how many you are spending at once), but the bonus wouldn't be at 50%. The max bonus would only occur when spending 5 pips and the overall risk would be higher or lower based on how many pips you have.

Let's say we have two crafters. One has 6 pips and the other has 12 pips, and they are both crafting the same white quality item. The 6 pip crafter could spend 5 points at once and try for the max bonus, but his risk would be 98.33%. The 12 pip crafter could do the same thing, but his risk would only be 56.67%. The 6 pip crafter is very unlikely to be successful in this scenario, so instead of spending 5 pips at once, he decides to spend 3 pips at a time. Now his risk is only 59%, but he needs to get 2 good rolls and he will only get the level 3 bonus instead of the level 5 bonus. Or he could do 2 pips at a time for the level 2 bonus, at 39.33% risk, but he would need 3 good rolls.

In this system, having more pips is always an advantage because there is no arbitrary risk threshold that needs to be met. Crafters simply spend 1, 2, 3, 4 or 5 pips and the risk is based on how many total pips they have and how many they are spending. Bonus levels are dependent on how many pips are spent at a time.

I get what you are saying with regards to the odds of someone rolling 1 pip at a time and getting an amazing on each. Yeah, that's extremely unlikely, but I don't think that's a problem. I think it fits with the risk/reward model. If you are only spending 1 pip at a time, you aren't risking much. Your final result won't be the best, but it won't be the worst either. If you are risking more by spending more pips at a time, you have the potential to produce something better, but also something worse. 

This is true as far as it's concerned, but its the scale a factor of the difference with each successive pip.

I ran a small 11 blank rune stone test to get a sense of the numbers (Skill 10, difficulty 20).

Using all pips are the target result "or better" way of looking at it, this is the odds table for 6 in a row selected individually based on those results.

Success or better:1/3.33
Moderate: 1/138.96
Good: 1/7256.31
Great: 1/155,527.00
Amazing: 1/1,291,467,969

When I ran the same numbers as 6 pips at once, I got 5 of my 11 tries (45%) at the Good results, something that if running one at time has only a 1 in 7256 chance of happening.

6 Pip Amazing one at a time is even worse odds than the lottery, because of the way lottery math odds work.

6/49 odds are calculated as (6/49)*(5/48)*(4/47)*(3/46)*(2/45)*(1/44)=13,983,816

Based on my test of 2 Amazing success in 66 tries, the math is.

(1/33)*(1/33)*(1/33)*(1/33)*(1/33)*(1/33)=1,291,467,969

It's easy to think that 6 tries in a row at 1/33 is much more likely than winning a 6/49 draw because the 33 number is so much lower than 49, but because you have the chance at 6 possible numbers to hit on the first try, and 5 on the second, and so on, it is actually over 92 times more likely to win a lottery than see 6 pips in a row of Amazing success with each pip having a 3% chance of landing Amazing.

To get to 1 in a million 6 pips one at a time, you need to have 1/10 (10%) chance of getting an Amazing Success on each pip.  The numbers actually get tolerable at 1/5 20% chance per pip, with a final result of 1/15,626, but then you have to think, how fun would the game be if you only had a 1/16k chance of ever seeing a 6 pip amazing result when going all in.

There is literally no way to balance this math with a "risk" factor that gives a 50% bonus, so that one at a time is worthwhile, and still make there be any effective chance of getting even regular moderate results going all in.  

17 hours ago, Tark said:

I don't think this discussion is ready to be had until mass production comes online. Blair stated that the bonus system will make more sense once we get factories in game. I cant remember if he said it in a stream of a post in the Dev forum, but it was recent (this year).

Maybe this discussion isn't ready yet, but I also know from the last live stream they are starting on the next round of crafting workflow changes, so maybe right now is exactly the correct time for them to make an adjustment or two. 

I get that factories and such are coming, but I also know odds from my professional background, and there is no way to balance these odds with reasonable choice the way it is laid out.  I also personally suspect those systems are not going to make the system less prone to problems and show how it works, but rather highlight and multiply the math issues related to the risk bonus.

Fortunately I suspect the suggested changes are not going to be hard to do,  before or after factories.

Edited by KrakkenSmacken

Share this post


Link to post
Share on other sites
On 4-4-2018 at 4:53 PM, Brightdance said:

Yup, I didn't think I'd say this but boy am I sick of the half pip bonus. It is utterly infuriating to basically only craft for a lucky roll. This should not be the case. 

Crafting should be an art and if done right a great piece of equipment should be the result. This is not the case right now. Currently, you carefully craft all the pieces to get to the final assembly and have a chance at a worthwhile piece. Roll a Success and throw it away, start over.

At first, this didn't bother me so much but as someone who basically logs on just to craft daily, it is getting to the point that I just don't want to bother.

Don't know what else to say except something here isn't right. I'm a big supporter of this crafting system... I'll go as far as to say I totally love it. Except for this one part of it because it really takes it COMPLETELY out of your hands and puts it on a lucky roll. 

BD

 

Totally agree with you. 

Share this post


Link to post
Share on other sites
On 5/21/2018 at 2:49 PM, Tark said:

I don't think this discussion is ready to be had until mass production comes online. Blair stated that the bonus system will make more sense once we get factories in game. I cant remember if he said it in a stream of a post in the Dev forum, but it was recent (this year).

He said it in a stream after I asked about it. But honestly, that's bull and a crappy go-to-answer when it comes to crafting. We are smart enough and have seen similar systems in the past to know what factories will do. I get it.

This 50% exp bonus has nothing to do with factories. YES, it will be easier to deal with when you can craft 100 of each item needed and combine away until you get a 50% amazing or great result on all the pieces including final combine. That doesn't fix the issue it actually makes it worse by leaving crafters without a factory (new players) to experiment at 50% and ruin their hard earned work... or have run around with poorly made socksty gear never to be able to compete.

I supposed Factories could have a way to do what we are asking for and allow players to mitigate the risk when using a factory. That'd be neat but still leaves people without factories out in the rain. Doomed to experiment on countless swords to have them all turn out like crap.

@thomasblair any insight into this? I think enough has been asked that warrants some kind of response. You have an amazing system here with one huge issue that makes it so painful! Not sure how much crafting you've done but the combined experience of the people speaking up is well into the 100's of thousands of crafted items.

 

Maybe someone can ask ing the money forums?

BD

Share this post


Link to post
Share on other sites
1 hour ago, Brightdance said:

Maybe someone can ask ing the money forums?

BD

Already there.  

The posts by me on this page are pulled almost word for word from that Forum.

 

Share this post


Link to post
Share on other sites

FWIW, I am going to provide a detailed breakdown of some crafting experiments that I did today.  

Background information

  • All crafting was done using a L10 Nethari rune crafter with 125 INT
  • With buffs I had 9.6 experimentation pips and a 42.5 experimentation skill
  • All of crafting involved making skinning knives
  • Each knife was constructed with common quality parchment, gold, silver, and stone

I started by crafting 25 knives and assigned one pip at a time when making the sigil, the rhinestone, and the final assembly.  Next, I crafted a second batch of 25 knives and, this time, I rolled 6 pips for the sigil, the rhinestone, and the final assembly.  I recorded the results of each and every roll and summarized the information in the following chart.

  Diff 25 / One Pip Diff 45 / One Pip Diff 25 / Six Pips Diff 45 / Six Pips
Crit Fail 0.3% 2% 0% 7%
Failure 5.1% 3% 10% 21%
Success 0.0% 34% 21% 32%
Moderate Succes 37.7% 27% 21% 18%
Good Success 28.7% 15% 21% 14%
Great Success 18.7% 10% 19% 4%
Amazing Success 9.5% 8% 8% 4%

Next, I used this data to run some simulation and "crafted" 10,00 items using each methods and plotted the results.  (Note, for convenience I ran a kernel smooth over the results of the simulation which is why the probability density function extends below zero)  You can see the plot at

https://nofile.io/f/zDBNWuXErkf/Rplot19.jpeg

When you look at the chart

  • Red shows a simulation where I rolled one pip at a time
  • Blue shows a simulation where I rolled 6 pips at a time

 

  • The solid line shows a 25 assembly roll
  • The lines constructed of circles shows a 45 assembly roll

Here's what I am taking away

  1. I don't see an excessive number of failures
  2. On average, for both the component assembly and final assembly, on average one pip assembly produces much better items that rolling all your pips at once.  (The center of mass of the distribution is much further to the right)
  3. Rolling all your stuff at once has the potential of producing exceptional goods, but this happens very rarely.  Most of the time you end up with something mediocre.

 

Personally, I see nothing wrong with this system.  Changing things to make the two distributions closer to one another will decrease the set of options available to people.

  • You can create average items if you want
  • And you have the option to be more risky and try to create something exceptional knowing that you might end up with something bad

From my perspective, the current round of complaints boils down to "make it easier / less risky for us to produce the best goods".

I don't think that this part of the game should be trivialized.

 


WAZ6Fov.png

"The cinnabar is a lie"

Share this post


Link to post
Share on other sites
13 minutes ago, narsille said:

Personally, I see nothing wrong with this system.

You don't think its a bit strange that it is possible for a trained crafter to have too many experiment points available, making it impossible to get the risk bonus on an item? So his product can never be as good as a less-trained crafter making the same item?

Or how gaining an experimentation point that brings you up to an odd number is generally a bad thing because you can't do two 50% rolls?

Edited by Jah

IhhQKY6.gif

Share this post


Link to post
Share on other sites
38 minutes ago, narsille said:

 

  1. On average, for both the component assembly and final assembly, on average one pip assembly produces much better items that rolling all your pips at once.  (The center of mass of the distribution is much further to the right)

From my perspective, the current round of complaints boils down to "make it easier / less risky for us to produce the best goods".

I don't think that this part of the game should be trivialized.

 

I don't see any final results math here.  What was the average and best final values of every item determined by each model?

What were the cases where one pip at a time produced better final results?  With the risk bonus and trying to get repeated successes, that is far different than what % chance you have for any given pip.

Also, did you apply the total pips earning the %, or the total tries?  For example

Quote

I rolled 6 pips for the sigil, the rhinestone, and the final assembly.  I recorded the results of each and every roll and summarized the information in the following chart.

 

Good Success 28.7% 15% 21% 14%

 

Does 28.7% represent the chance of an individual roll coming out positive, or was that by pips spent. If it was by individual roll, although you are more likely to get a "Good Success" roll per individual pip, given your chart you will have only a 3.39% chance of stringing together 6 Good Success or better on the 25 Diff, while you show a 14% chance of the same on a 45 Diff.

That does not even include the value of the 6 pip representing a 50% better final score due to the risk bonus.

It has zero to do with "make it easier", and everything to do with odds math related to stringing together successive wins on low odds.

Now consider, that 4% 45 Diff amazing success just made you a blueprint you can crank out (est) 100 copies of the same results. spending 20% of total resource vs 5% is a very small price to pay for experimentation that results in a guaranteed 1.5X better item final value.

It has zero to do with "make it easier", and everything to do with odds math related to stringing together successive wins on low odds. In fact, I am certain with my suggested changes that the average results will be lower, not higher, and the really good items will be even more rare and difficult to obtain, not more frequent and easy.

In case you missed my suggestions above in the wall of text.

Quote

Suggestion 1: Change the calculation for the risk value from the amount of pips you are spending when you push "Run Experiment", to the amount of pips spent in previous experiments on the item and previous gains in the row.

Suggestion 2:  (Dependent on suggestion 1 being implemented.) Add an "Add Risk" button.  This button spends a pip, but does not perform any experiments.  This allows you to raise your risk and risk bonus before experimenting with any pips.

Suggestion 3: Add a "restart experiment" option.  This option would unset all currently applied pips and set earned values to 0, but the pips would still be spent. 

I want to see a better and more engaging crafting game, not less or worse, or easier.

If anything, your numbers prove the point I was trying to make, not the opposite.

Edited by KrakkenSmacken

Share this post


Link to post
Share on other sites

Krakken, if you are unable to look at PDF and extract the center of mass, then its not worth my time dealing with you.

With this said and done, here is the code.  It should answer your questions

 

setwd("~/Documents")

foo_cf = rep(-100,27) 
foo_f = rep(0,515)
foo_m = rep(1.422, 3767)
foo_good = rep(3.91, 2873)
foo_great = rep(4.468, 1870)
foo_amazing = rep(5.585, 949)

foo = c(foo_cf, foo_f, foo_m, foo_good, foo_great, foo_amazing)

bar = sample(foo, 60000, replace = TRUE, prob = NULL)

bar = matrix(bar, 6, 1000)

test = colSums(bar)

test[test < 0] <- 0

mean(test)
max(test)

plot(bkde(test, kernel = "normal", canonical = FALSE,
         gridsize = 401L, truncate = TRUE), main = "Kernel Smoothed Assembly Rolls", 
     type="p",col="red",xlab = "Frequency", ylab = "Quality")


###########

foo_f = rep(0,10)
foo_success = rep(6.552,21)
foo_m = rep(12.978, 21)
foo_good = rep(35.19, 21)
foo_great = rep(40.212, 19)
foo_amazing = rep(50.265, 8)

foo = c(foo_f, foo_success, foo_m, foo_good, foo_great, foo_amazing)

bar = sample(foo, 10000, replace = TRUE, prob = NULL)

points(bkde(bar, kernel = "normal", canonical = FALSE,
          gridsize = 401L, truncate = TRUE), type="p",col="blue")

mean(bar)
max(bar)

 


WAZ6Fov.png

"The cinnabar is a lie"

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...