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

Tweaking the crafting model to be less RNG, more progressive.

Recommended Posts

I was actually thinking of blackjack as I read through your posts (maybe since you had posted before about gambling software).  You start off with two cards, check your results, then hit as you deem necessary while you build toward a perfect item; that 21.  Is it a fair analogy to say the current system might be playing blackjack by drawing X number of cards and hoping for the best? lol

 

I linked this two month old video for someone about a different topic, but I realized that the experimentation phase was a little different in that very early prototype.  It looked like your maximum pip was still limited by your resources, but scoring successes actually increased the maximum number of pips you could spend on the same stat.  Have a peek; I'm curious if you think that implementation would have been relevant to your idea:

 

https://www.youtube.com/watch?v=og_DJoG08T4&t=516s

Yup, that older video is much closer to the system I was describing above.

 

Wonder if they changed the idea, or just couldn't get it in for the current round of testing.  They also had a cosmetic stage that has not made it in yet either.

 

It's still missing the idea that if you succeeded before, you could only slide backward so far on your next experiment.  So the system with more pips if you got amazing success still doesn't quite have the flexibility or as strong a sense of progress and variance as what I proposed, but it is better.

 

If the pips are added (+1) per roll vs (+1 * Pips) per roll, it gives incentive to take it one at a time to get the most chance at the most pips, while on the other hand lowers your odds of getting the Amazing success to apply to larges groups at once.

 

I would have to see it in action to re-run the numbers and model to make a determination if it ended with an improved choice matrix, or still has the same "best option always" problem.

Share this post


Link to post
Share on other sites

Ah I get it now, actually as I read this I too was thinking blackjack haha. 

So as on the previous page, Hyriol, mentioned you have your two cards and try to progress towards 21.


So let me explain it BACK to you and see if I get it.

Lets use a 1-100 scale 1 being the absolute most basic/worst item you can create and 100 being the maximum.

You gather some Purple resources and are fairly leveled in crafting, so lets say you have 6 total pips. 

You roll the first. Its a Success. This moves you up +15 points on the 1-100 scale.
The next is a good success. This moves you up +20 points on the scale. You now sit at +35. Do you press your luck more and spend more pips or lock in a 35?

You press your luck. Failure. You move backwards -5 and end up at +30.

You roll again. Amazing success! Wohoo! It grants +25. You move up to +55 now. You have spent 4 of your 6, but feel pretty good about your Blue Quality item (with a 55 score). So you make the item.

Or... Lets say the guy continued. He fails. Moves back -6 this time. now at 49 - it "falls back to green". 

Ok, lets use the final one! "Success!" +15. Now he sits at +64 (shy of progressing to Purple by 6 points).

All those variables - could be adjusted. Im just trying to paint the "picture" of the system you suggest.

So it would be INCREDIBLY rare to get Amazing Success all 6 times. In fact. Maybe even near impossible. But what you might see instead is someone choosing NOT to spend more points, if they had maybe 4 "Amazing Success" in a row, because that would be a truly great item in likely the top tier. So then you wouldnt even NEED to use all 6.

Or if you had 1 Amazing + 3 Good. You might just lock it in at a purple version because more doesnt always = better.


This is what your talking about? 

It seems your title might be backwards though. The current system might be more progressive. The new would be more RNG but the NEW would be ALOT more fun. I would give you that one. Each item would be different. The "goal" would be to fall into a certain category of "rarity". Each path would be different. You might have a guy end up at +68 on the scale, or +64 on the scale and it makes the same item. however a +80 versus a +76 might be two different rarity depending on thresholds? 

Share this post


Link to post
Share on other sites

Ah I get it now, actually as I read this I too was thinking blackjack haha. 

 

So as on the previous page, Hyriol, mentioned you have your two cards and try to progress towards 21.

 

 

So let me explain it BACK to you and see if I get it.

 

Lets use a 1-100 scale 1 being the absolute most basic/worst item you can create and 100 being the maximum.

 

You gather some Purple resources and are fairly leveled in crafting, so lets say you have 6 total pips. 

 

You roll the first. Its a Success. This moves you up +15 points on the 1-100 scale.

The next is a good success. This moves you up +20 points on the scale. You now sit at +35. Do you press your luck more and spend more pips or lock in a 35?

 

You press your luck. Failure. You move backwards -5 and end up at +30.

 

You roll again. Amazing success! Wohoo! It grants +25. You move up to +55 now. You have spent 4 of your 6, but feel pretty good about your Blue Quality item (with a 55 score). So you make the item.

 

Or... Lets say the guy continued. He fails. Moves back -6 this time. now at 49 - it "falls back to green". 

 

Ok, lets use the final one! "Success!" +15. Now he sits at +64 (shy of progressing to Purple by 6 points).

 

All those variables - could be adjusted. Im just trying to paint the "picture" of the system you suggest.

 

So it would be INCREDIBLY rare to get Amazing Success all 6 times. In fact. Maybe even near impossible. But what you might see instead is someone choosing NOT to spend more points, if they had maybe 4 "Amazing Success" in a row, because that would be a truly great item in likely the top tier. So then you wouldnt even NEED to use all 6.

 

Or if you had 1 Amazing + 3 Good. You might just lock it in at a purple version because more doesnt always = better.

 

 

This is what your talking about? 

 

It seems your title might be backwards though. The current system might be more progressive. The new would be more RNG but the NEW would be ALOT more fun. I would give you that one. Each item would be different. The "goal" would be to fall into a certain category of "rarity". Each path would be different. You might have a guy end up at +68 on the scale, or +64 on the scale and it makes the same item. however a +80 versus a +76 might be two different rarity depending on thresholds?

Yep, you got it.

 

The only thing you missed was here.

 

"It seems your title might be backwards though. The current system might be more progressive."

 

"progressive: happening or developing gradually or in stages; proceeding step by step."

 

The current system reinforces always taking maximum risk for maximum results, rather than step by step.  If you look at the math I provided above, you can see just how far the odds would have to be adjusted to even get in the same ball park, where 1 multi pip amazing success out preforms orders of magnitudes in the millions (1/20) vs (1/100000000) on 8 pips.

 

It is also progressive in that each step is more likely to be better than the last, rather than each step being purely individual random chance.

 

It can also create a sense of the gamblers dilemma or gamblers fallacy in participants.

The gambler's fallacy, also known as the Monte Carlo fallacy or the fallacy of the maturity of chances, is the mistaken belief that, if something happens more frequently than normal during some period, it will happen less frequently in the future, or that, if something happens less frequently than normal during some period, it will happen more frequently in the future

For example, I really want a high minimum and maximum damage, but don't care so much about my durability.  

 

So I first select one pip in minimum damage.  Amazing Success.  Humm, I don't think it's likely to get two of those in a row, so I select durability thinking I can waste a "bad" roll on something I don't care about.  Normal Success, yup guessed right, I say stupidly to myself, and switch maximum damage.

 

Oh no, another normal success,  That can't happen again, I think, and try to double down on my minimum damage again because I'm "due" a good roll.

 

It's the same mentality that makes a gambler switch back and forth between red and black on the roulette table.

 

By making it both progressive, as in each pip always gets a roll, with a general up tendency based on skill, the advantage to selecting four pips at once is only a reduction in complexity, for the risk of not stopping at a good roll and going backwards from an amazing success result. 

 

It's still 100% RNG, it's just more stable/predictable in its results, and harder to achieve those incredibly lucky moments when you get two or more amazing successes in a row.

 

----------------------------------- New Approach ----------------------------

 

I was also thinking given the "Amazing success = more pips" from the video, and the fact pips and such are already built, an approach that uses the existing tools rather than a full sliding scale may be easier and faster to implement.

 

So the idea was a bit like how bowling is scored.

 

The rules are simple.

  • Each pip uses the total value of itself and the previous two (or alternately next two) pips, if the pips are selected one at a time.
  • Critical failure clears two previous roll values, so every roll could result in a zero, as well as reducing previous values you had.

 

Lets say your first pip was Success, so +1.05.  

You throw again and get moderate success +2.10.  Instead of the pip being just 2.10 in value, it also add in the previous success value of 1.05, making the full total 3.15 for the second pip, and 4.20 total.

Next is an Amazing Success, at 14.75 or whatever it is. (Can't quite remember)

Now your third pip is worth 14.75 + 2.10 + 1.05, total 17.90.  New full total is 22.10.

Fourth pip is a normal success again, but still gets to add in 14.75 and 2.10, for a total of 17.90 again and new improvement total of 40.00.

 

If you get to the last pip and throw an amazing success, it will automatically give you another pip, and if you again get amazing success, a third.  So just like bowling and strikes, an amazing success on that that last pip is ultra important. Equally important is getting an amazing success on your first pip, because that will carry you for three more.

 

The alternative is if you select multiple pips in one go, ALL those pips count once, and have the value of the single roll. 

 

I'm not 100% sure on the math, but at first glance that appears to be a meaningful choice as individual pips selected that way can be more valuable than pips selected together. I suspect that it could be balanced so that each choice has a proximate equal value.

 

Individually a single pip stringing 3 amazing successes together (1/10000 chance or worse), would be worth 44.25, where it is much easier (1/13.75) to get lucky for one roll and get three pips without overlap worth the same together.

 

That takes away the need for the sliding scale, as progression is built into the system directly, all with mostly the same as existing mechanics.

Edited by KrakkenSmacken

Share this post


Link to post
Share on other sites

Here is a link to another thread with some math in the economy forum.

 

With some more experimentation data provided, and a possibility I had not considered, the model they have just needs a couple of tweaks to the formula to bring risk/reward closer in line.

 

If they happen to want to use any ideas here great, but all they need to do really is isolate the desired best 100% risk results for amazing success, and then trickle down the results/odds through a pip/odds algorithm.

 

With more data points I could probably reverse engineer the actual formula being used, and offer some suggestions to the value inputs, but as it is the current model can create a more progressive system, than the all or nothing that currently exists.

 

I'm rather happy to have figured that out and confirmed it actually.

 

I think it is like this:
1 pip 11% risk:?, most common result (median) Great Success?

2 pips 18% risk?, most common result Good Success?

3 pips 25% risk?, most common result Moderate Success?

4 pips 50% risk?, most common result Success?

5? pips 75% risk?, most common result Fail?
 

 

Also I should say sorry that I missed understanding what you were saying here Frykka. That is exactly what the sample numbers someone gave bears out, to a degree I had not even considered as an option.  Probably because of my background.  Never had to build a game where the jackpots are the most likely outcome, which is what it seems to be when you only select 1 pip at a time.

Edited by KrakkenSmacken

Share this post


Link to post
Share on other sites

The missing loop is the time.

 

In theory, the larger risk should negate the time, not have a chance to net a better result. 

 

Using pips one at a time vs all at once should not effect how much the max possible percentage.

 

Using them all at once should incur the highest risk of not getting an amazing, not the magnitude of the amazing.

 

Using 1 pip at a time should have a much higher chance of an amazing but each one increases the total time.

 

So 8 pips of amazing all at once and 8 pips of amazing one at a time should equal the same total increase, the difference should be the one at a time amazing chance should be relatively high but increase the time to finish the item a huge amount. While the all at once chance to get an amazing be extremely small but create the item very quickly.

 

 

How it is now, the more 'risk' the higher magnitude of the amazing success. So you get the item quicker and at a higher max percent, theres no downsides as long as you have enough material.  Theres no weighted choice, its 100% yolo to get the best.

 

In Markeedragons video with blair, he clearly stated its supposed to work with risk vs time, the implementation of it actually turned out to be different. Risk not only nets a shorter time but a better magnitude. 


CfWBSig.png

Share this post


Link to post
Share on other sites

The missing loop is the time.

 

In theory, the larger risk should negate the time, not have a chance to net a better result. 

 

Using pips one at a time vs all at once should not effect how much the max possible percentage.

 

Using them all at once should incur the highest risk of not getting an amazing, not the magnitude of the amazing.

 

Using 1 pip at a time should have a much higher chance of an amazing but each one increases the total time.

 

So 8 pips of amazing all at once and 8 pips of amazing one at a time should equal the same total increase, the difference should be the one at a time amazing chance should be relatively high but increase the time to finish the item a huge amount. While the all at once chance to get an amazing be extremely small but create the item very quickly.

 

 

How it is now, the more 'risk' the higher magnitude of the amazing success. So you get the item quicker and at a higher max percent, theres no downsides as long as you have enough material.  Theres no weighted choice, its 100% yolo to get the best.

 

In Markeedragons video with blair, he clearly stated its supposed to work with risk vs time, the implementation of it actually turned out to be different. Risk not only nets a shorter time but a better magnitude. 

Agreed.

 

From the sample data I have in the other thread, the odds are both better, and take shorter to produce all at once vs pips at a time.

 

4 pips has a 60% less likely chance at double amazing success as 8 pips at once. I didn't look lower simply because below 50% risk you lose so much potential it's simply not worth it.

 

A "simple" fix, besides the obvious adjustments needed to the percentages, would be to add % improvement based on complexity, regardless of results.

 

So when you select 1 pip, you add 1 complexity, and you add 1% to your bonus, regardless of the outcome.  Picking 8 individual pips would automatically give you 8% bonus, in addition to your experiment results.

 

 It would also increase the time it takes to build, so that's the balance/payoff point.

Edited by KrakkenSmacken

Share this post


Link to post
Share on other sites

Would it make sense in this system if you "consume" multiple pips in any 1 experiment, it would give a higher probability of success?

So this enables someone who has 8 pips from being a bad ass crafter and using high quality materials to consume 2 pips each experiment to get a higher % chance of amazing success? 

So rather than using 2 pips to do TWO experiments, rather you use two pips on ONE experiment to merely have a higher % chance at positive results? 

So to use blackjack as the analogy, using 1 pip is basically sitting down at the table fresh. Using 2 pips would be aking to counting cards and knowing more "10s and face cards" are expected to come, tpy of thing.

This way if someone has a good result after 4 individual experiments, they can use their 3 remaining pips to give the highest success rate

Share this post


Link to post
Share on other sites

Would it make sense in this system if you "consume" multiple pips in any 1 experiment, it would give a higher probability of success?

 

So this enables someone who has 8 pips from being a bad ass crafter and using high quality materials to consume 2 pips each experiment to get a higher % chance of amazing success? 

 

So rather than using 2 pips to do TWO experiments, rather you use two pips on ONE experiment to merely have a higher % chance at positive results? 

 

So to use blackjack as the analogy, using 1 pip is basically sitting down at the table fresh. Using 2 pips would be aking to counting cards and knowing more "10s and face cards" are expected to come, tpy of thing.

 

This way if someone has a good result after 4 individual experiments, they can use their 3 remaining pips to give the highest success rate

Short answer, no if I understand what you are saying.

 

Let me expand the vocabulary a bit, because "PIP" is currently holding three values in the current process, and I think you are describing a system that needs to separate the values, where the current system is entirely 1-1.

 

Stake = How much of something your putting down in exchange for a chance.

Outcome = One random event that leads to a result.

Result = One multiple of a positive outcome.

 

If I read you correctly, you would like to stake two pips, on one chance, and get 1 result with better odds, where as now you stake two pips, on one chance and get 2 results.

 

If that is the case, then the answer is definitely a no, simply because you are reducing the number of results possible.  Basically if you stake 2, outcome once, and get 1 result, that result has to be always worth the multiple of the stake (2), or it is better to only stake 1, outcome 1 and get 1 result. Even if the odds of a positive outcome double, you still are better off with two bets because you can get that positive out come twice.

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...