Development Cycle Archive

Thread: Crafting Critical Failure Rates

Zarlor
Mon Jan 26, 2004 6:03 pm
#66

Ooops, I was wrong on the station, The crafting station I've primarily been workign with is a Food & Chem rated at -22.85.


I'm using 3 tools, --1.61, -9.31 and 13.62.


All crafting for me is done at +50 Med Assembly and Experimentation.


Quality of resources, of course, varies depending on what we have.


I'm doing most all of the med stuff I can craft, so the schematics also vary, but more components are being crafted by me than anything else (of course, since those need to go into the final meds, usualy.)


I will experiment usually 1 box at a time. I sometimes do more, however, up and including all 5 of my EPs at once. So there is definite variance.


All of that info is strictly what I, the Medic Correspondent, have been crafting in the lab. Others are doing their own things as well, but as I mentioned most of them are likely crafting at +100 Assembly and Experimentation skill in their respective fields.


I would suggest that the variables being used here are otherwise pretty well varied. From crafting from great OQ & Mal resources (both known to affect failure rates) to crafting from really poor OQ & Mal resources. I would just conisder the variables all over the place on most of that stuff. I think the other stations in the Lab are probably all around the -20 range, though, so that is a semi-constant for those Correspondents who put in time testing there. They may be crafting at their own stations, though, so it;s hard to say.


Any folks using the tools that were in the lab are using stuff likely around the +13 range.


Of note is that while "The Lab" is a building, the script that this information comes from is actually associated onto our characters, so it does not matter where we craft or what we craft or anything, it's always logging the info.


Also if it's any consolation I do believe that the numbers listed by Chrysalda are strictly preliminary figures, and that his script is actually capturing a LOT more information than just crit failures.



Zarlor - Mesric City, Dantooine - Eclipse
Arissi Plains, Dantooine - Test Center
Doum
Mon Jan 26, 2004 6:09 pm
#67

I had 3 times 2 failure in a row in less than 3 hours. And i had a few single ones. I can understand the single ones but 2 in a row, makes it frustrating when you have sub components involved.



_____________________________________________________
Doum Medilya (Tarquinas server)
World Trade Organization
Mos Benzin Citizen Colonel in Rebel Army
"Rebelion is the first step towards Freedom!"
ideas
Mon Jan 26, 2004 6:09 pm
#68

Thank you for looking into this issue!


I believe a flat out # attempts vs crit fails tracking is probably not sufficient to get at the problem. For a Novice crafter, 4% crit fails might be just about right or maybe even pretty good. For a Master it is humiliating. As others have pointed out, the crit fails seem to happen in clusters, 3 or 4 in a row sometimes. Maybe it's just probability and a person's tendency to remember the extreme results, but maybe it's a flaw in your randomizer.


In any case, I'd be curious to know what the developers feel the proper ranges *should be*?


For example, what should the crit fail rate be for:


Novice Artisan with +0 crafting tool.


Master Artisan with +0 crafting tool.


Master Artisan with +10 crafting tool


Master Artisan with +10 crafting tool, +25 crafting station


Master Artisan with +15 crafting tool, +40 crafting station


Master Artisan with +15 crafting tool, +40 crafting station, in a Manufacturing / Research Center




These results should be quantifiable somehow, both from a design perspective (what is the intended curve) and from an empirical test (what is the actual performance in the game).



Since Master crafters typically have a +100 (item) Assembly and Experimentation skills, they intuitively expect 100% success on their crafting and experimentation attempts. As game-players, we can understand that there should always be some small chance of loss, but we'd lke to know what that chance is and how we can get it as small as possible.


Do the tool and crafting station bonuses make a difference? If so, how much?


If the experimentation interface shows a Risk %, how does that % relate to our actual chance of failing the experiment? (i.e. Risk = 0% still crit. fails, so what does Risk = 10% mean?)


Can a non-Master craftsman really experiment effectively? Or is his chance of improvement worse than his chance of success?


Can we get some sort of statistical guarantee that we should only get two crit. fails in a rowevery 0.16%(four percent squared) of our crafting attempts?


Let us know what we should *expect* in our crafting, and then we can report whether the game meets those expectations properly.




If I were to recommend a design out of the blue based on what already exists, it would be something like this:


When assembling the basic item, roll % number - Complexity + Assembly skill + Crafting tool bonus (and player city bonus). End result is a % of "base quality" achieved. Negative result is a crit. fail.

(Example:With 100% skill and +30 in bonuses, Complexity 0-30 items automatically succeed, Complexity 35 has 5% crit. fail, Complexity 50 has 20% crit fail, etc.)

The chance of success should be indicated like the Risk-meter, to let a crafter know what his chances are.


Use of the practice button in crafting acts as a 5-minute crafting buff, allowing you increased assembly/experimentation results for that short period.


When experimenting, roll % number - Complexity + Experimentation skill + Crafting station bonus (and player city bonus). End result is a % of improvement achieved based on the number of experimentation bubbles used. Negative results reduce effectiveness accordingly, but only count as a "crit fail" if the negative result reduces the experimented factor below 0.

(Example: With 100% experimentation and +30 in bonuses, a Complexity 30 item will yield 10%-20% improvement per bubble -- if there are 10 bubbles across a row, then each is worth 10%, the roll of 1-100 + 100 + 30 - 30 yields 101 - 200%of the bubble value, or 10-20% improvement.)

Again, the Risk-meter should accurately reflect the chance of losing quality with the experiment.



In a system like this, Master Artisans could create many low-complexity items safely in the field without fear of crit fails, but would recognize the benefits of a high-quality crafting station when making anything of high complexity or that needs major experimentation.










So, let me get this straight: To advance my character, I have to give up my current abilities?

Flurry: Ikeya Ibye (Master Droid Engineer, Master Artisan, Master Merchant)

IKEYA Grand Mall - Naboo, Moenia - Waypoint 5000 -4000



SithLizard
Mon Jan 26, 2004 6:20 pm
#69

4.49% is still a bit high for a Master.


Why not give Masters a break? Lower that rate by half for Masters.




RNK
Mon Jan 26, 2004 6:35 pm
#70

When gaining xp, smugglers are known, from time to time, to fail opening a container or trying to improve a piece of equipment through slicing.


However, upon reaching master, they NEVER fail.


Why, then, should master crafting classes fail when putting things together? Smugglers are doing something potentially dangerous to a weapon, piece of armor, etc.that could cause it to not work properly. Yet when we, the crafters, are assembling the item (with choice top-cut resources), we have a chance to screw it up to the point where it doesn't get put together?


I understand this is a type of time/money sink, but let's be realistic! Would you request a vase maker to throw one together for you if you knew he'd been in business forover twentyyears, but every 1 out of 20 he makes fall apart when people take them home?


As far as experimentation goes, I think 90% of the time, masters should get the "great" or "amazing" success. Then, that other 10% of the time they can get something worse.



***Jozu Methusela. Master Weaponsmith. Mastered Carbineer, now onto Teras Kasi Artist.***
Bamboozle
Mon Jan 26, 2004 7:10 pm
#71

What is the point of critical failures, anyway? I'm thinking it's something put in by someone who forgot what a game is all about: having fun. Having a critical failure on a PA hall or a RIS armor segment is not fun.

Remove critical failures, or at least remove them for masters.



The Kitten's Diary, Day 781: I have discovered a most delightful way of making life miserable for my captors. Yesterday, I ate the woman's precious begonia, and today I have dismembered the amaryllis in the bedroom window. She is furious over the loss of her darling house plants! I find it highly gratifying.

Shai - Lieutenant Colonel of the Imperial army, FK division
Resource Vendor in the BlueDog Mini-Mall, Haven Island, Corellia -2010 -4670

Sooty
Mon Jan 26, 2004 7:12 pm
#72

Asa bio-engineer, myself and a lot of my fellow professionals are seeing much higher critical fail rates than these AND harsher penalties for those critical fails.


During experimentation;


All our experimentations are done on a paired stat - 2 numbers being increased at the same time. A critical fail on either stat drops that stat AND another stat not in the pair being experimented on. E.g a failure on Psychology drops Dependibility (from the Psychology pair being experimented on) AND drops Cleverness (from the Mental pairnot being experimented on). This lowers the potential maximum for both stats.


Now, given your own test data critical failure rate of 4.2%, this means that a Master BE has less than a 40% chance of completing ONE creature withoutone critical fail - I simply cannot believe this is as intended. Particularly with the double hit we get on stats with any fail.


However - I personally could probably deal withrelatively high levels of failure if, once the template was made, we didn't have a chance of critical failing the final creature construction and losing potentially hours of personal time in both crafting and resource gathering on one single die roll that certainely seems to be a lot worse than 4.2%.


For me, asa BE, 2 adjustments must be made to avoid 'keyboard breaking frustration' that happens far too often;



  • Stop reducing a second, unrelated, stat on a crit fail in experimenting on a pair

  • Eliminate the critical fail chance of the final creature creation once the template is made
Abominable-TCO
Mon Jan 26, 2004 7:43 pm
#73






Chrysalide wrote:


Assembly attempts: 1448
Assembly critical failures: 65
Assembly critical failure rate: 4.49%


Experimentation attempts: 1284
Experimentation critical failures: 54
Experimentation critical failure rate: 4.21%






Ok now lets see the stats on how many "moderate sucesses" you got whilst experimenting! I am a MWS and I cannot make a weapon with just one of these in the experimentation stage! A "good success" is marginal at best but realistically anything less that all "great successes" is garbage and ends up being destroyed! Maybe some clarification on what is meant by moderate and the rest etc as I fail to see why any success makes a weapon worse than it already is/was! One moderate success means that typically you cannot recover the item to be salable


I could honestly give 2 hoots about crits as they happen rarely to me like your stats suggest sure they are annoying but I can live with those when they happen (I have yet to try using enhancers tho, so I can't say wether or not they effect the critical aspect). I understand it is fustrating for architects and the like but they use so much more resources.





--------------------------------------------------------------------------------
Quagga's TCO Weapons Depot

Based in the Player City of Draicco, Talus
{I am on the map and have a shuttleport as a doorstep(2980 1331)}


Dharmon
Mon Jan 26, 2004 8:13 pm
#74

Try that in Coronet on Chilastra on a busy night and you will get 25-40% failure rates due to server lag. Same thing for repair attempts.. I have done enough city crafting and repairs to notice a DIRECT relationship between failure rates and server lag/high traffic areas. I always take people outside of town for repairs now and do not hand craft armor for people in cities anymore.





First complete RIS set made for myself on Aug 1, 2004 with stats of approx 75% Kinetic/4% Stun/5% Acid/65% Base/vuln to Lightsaber. (unsliced)

Diwo (Master Armorsmith/Master Smuggler) - Semi-retired but still making RIS (email me for a guide on what you need for RIS)
Dharmonn (Swordsman)

DCORP Armorcrafting (Armor Division partially operating with 80% Kinetic Composite and 75% Kinetic Ubese up for sale)
Cavalieri Neri, Naboo (Merchant Tent west of our shuttleport at 2588 4804)
Zarlor
Mon Jan 26, 2004 8:18 pm
#75






MasterNerfSlayer wrote:
Using a rating 9 tool (not the best I know) in front of a 40 rating crafting station when vehicles came out, I crit failed on my first 2 attempts at making vehicles. I dunno what kinds of testing you're doing, but it doesn't match player experience. My tools werent the best, but there is no way I should crit fail twice in a row with even half decent tools. I have experienced many more crit fails since and have since dropped artisan.






As one of the folks doing some of the testing (well ginding out stuff, at any rate) I have definitely ahd several instances of doubl-fails like that. So it still happens within that percentage being shown.



Zarlor - Mesric City, Dantooine - Eclipse
Arissi Plains, Dantooine - Test Center
jol69
Mon Jan 26, 2004 8:23 pm
#76

Chrysalide said:


I do have a minor change to crafting in development somewhat reducing the rate of critical failure. I believe this solution will be amicable to everyone, but I am definitely open to hearing any feedback about the existing system that anyone might have. While we may not be able to implement every suggestion, we do take them to heart and will do what we can to make it the best system for all parties involved.

No offense Chrysalide, but so far the Devs have proved remarkably poor judges of what changes the community as a whole will like. Rather than surprise us, go ahead and tell us what this intended change is.This policy of vagueness, recently witnessed in the Changes to the Jedi System announcement is simply unnacceptable.



~ Eseex Aptopack: Master Image Designer, Aspiring Carbineer~
cadof
Mon Jan 26, 2004 8:31 pm
#77

4% is critical failures high for a master. Also lag in the game has a grate deal with critical failures. Try you test during peek hours of game playing. Ability, crafting tools and stations Quality's does not make a deference is that what you are trying say. If some one can in lighting me on how this works I though it was this way. Or is it just luck of the draw is that what soe is saying to us all.
ra1f
Mon Jan 26, 2004 8:34 pm
#78

During beta I distinctly remember a specificarmorsmith crafting thread. It was relating to discussions of whether to use just 1 experimentatin point at a time or more than one. This was when they put in the display that showed the 'risk' and chance of failure by adding essentially 'too many' points at a time.


Someone asked why not just add them 1 at a time instead of all at once since there was 'zero' risk per experimentation.


The response was simply that everytime you experiment it increases the complexity of the final item (as you can clearly see yourself when crafting to this day.) This was important since at the time insurance costs were based upon item complexity. The more complex the item the more expensive it would be to insure it.


Also in that response was the Dev comment that besides the increased complexity (and subsiquent increased insurance costs, making less complex items more in demand by customers since it was cheaper to insure) that there was no other reason to do it, since many armorsmiths would likely choose the 'safe' route to produce high quality armor "without failures."


"Without Failures" .. many things have changed since the good ol days of beta, insurance costs just being one of them.


Any armorsmith that has been paying attention can tell you that there certainly is an 'optimum' number of experimentation points to use at one time (there is no magical number since it's based on too many things to list) but if done correctly you can even end up with an extra point or two at the end. Anyway.. Since if you put in 8 points at a time in a single bar (Hypothetically in this case) filling it up then experiment you will have to put more points in the same bar and re experiment since each 'dot' is 10% but each experimentation point does not result in a 10% increase in most cases (besides kenetic on RIS) forcing you to at least do 2 or more experimentatinos on a single item... increasing the 4.9% chance of experimenatation failure.. and that 4.9% chance of falure bases on the math comes into play everytime you hit 'experiment' .. so on any given piece or armor you are looking at up to a 15% or more chance of receiving a failure during experimentatino alone.


As for Armor Assembly critical failures. I have +15 Armor Assembly skill and several weeks ago tracked the failures throught a huge order of 15 full composite suite (all hand crafted no factory made). I had a 19.5% critical failure assembly rate (yes maxed out equppment).. including one very very very painful RIS Boot assembly to complete the quest.



Ralf Belefonte
Master Armorsmith - Quality Saves Lives
Tarquinas: TwinLoch, Talus; Coronet, Corellia

Cuyler Belefonte -- Retired
Jack of All Trades - Gave up chasing the Jedi Dream
Page 6 of 49