Architect Archive
Thread: Ran a test tonight and came out with an 11.8% critical Failure rate
DarythWE wrote:
Did you just go up the Engineeringbranch in Artisan?
If you notice both the Engineering and Domestic Arts branch both grant +10 Artisan Assembly at each level. So if you only did Eng. 1-4 you would have +60 (+20 at Novice). If you did both Eng. and Dom. Arts you would have +100
Artisan assembly shouldn't have anything to do with it, just as high artisan experimentation doesn't allow architects to experiment on their items. It's all about the Structure Assembly bonus from Furniture(which people keep saying is broken, but I can't figure how they know this) and theConstruction Experimentation (for experimentation).
The quality of your tools may or may not have an affect on experimentation; the numbers don't change. But it may affect your overall success rate, it's hard to say until the devs reveal the answer.
In Vegas that is true. Not in this game. Computer random number generators are not ballanced dice. They are notorious for creating strings of similar results. In eq I would go an hour without a failure then fail 4 times in a row. In SWG the attempt immediately following a failure is much more likely to fail. I haven't gathered data to quantify this, but I am positive it is happening, to the point where I never attempt a hard combine immediately after a fail.
Regarding the original post:
What % were your kits? What quality of resources were you using? Were you experementing?
I use a 90%+ kit on 90%+ stations and rarely fail (perhaps around 5% - unscientific). I was using lower % items last week (deleted the wrong kits and didn't catch my mistake) and was failing a lot more. Also, a few posters here say that by using materials with high toughness, you reduce fails. Others say that using experementation lowers the rate.
My opinion:
Maybe the fail rate is a little high, especially when comparing the failure rates for other master-level professions. I do not want to see failures removed from architecture tho. They are a vital part of the experience, imo, and should remain. I like the idea of getting some % of your materials or components back on the big fails tho.
11%.. wait a sec.. you failed way more than that..
look at it this way.. you failed 12 bricks, plus a wall.. BUT, that puts you at making 122 bricks for 10 walls.. 22%.. this is the problem.. the math is imbedded..
If I make a structure with 3 levels of components, the failure rate goes up each time.. you are being taxed a failure rate for all 3 levels.. imagine it this way.. you get hit for 12% charge earning money.. then you get hit for another 12% putting it in the bank, and another 12% taking it out.. you are not getting hit for 12% you are getting hit for way more..
The thing is, the more complex the item, the worse it gets.. 150 modules, you fail 10% and you get 135, and you make 15 more and you fail another, so you make 1 more. Your 10% fail rate costs you 16 modules on 150. So you make 15 walls.. you fail at least one. maybe 2.. you just lost 10 or 20 more modules, and you will lose 1 or 2 more making them back.. so now you have lost 27-38 modules, and if you dont fail the final assembly, you only had a 20-30% failure rate, off of a base failure rate of 10%..
Lets just assume you can do a 3% failure rate.. some people who have not failed 8 modules in a row with good gear and the best materials seem to think that the rate is around 3%.. ( I did this in my post on PA fail)
you need to build 10 large houses. You make 1000 bricks. You fail 30. you now have 970. You make 30 more and fail 1. Now you make 100 walls. You fail 3. That cost 30 more bricks, and you failed one more making those. Now you make 10 houses. At 3 percent, chances are you make them all without fail, but say you get unlucky. You fail one. That cost 100 bricks. You lose 3 more bricks making the 100.
So if you succeed all 10 houses, you lost 62 bricks out of 1000. Thats 6.2%. If you failed one house out of the ten, you lost 165 bricks, and assuming you beat the odds making the last house over and did not fail walls, you only had a 16.5% failure. So a lucky architect at 3% base failure rate actually is failing 6.2% of the time. The unlucky architect is 13% higher in his fail rate..
This is why we need to convince the developers to not include substructures in crit failures. You already took the risk to make them, you should not have to risk them 3 times to make 1 structure.
BTW, artisan assembly has nothing to do with structure assembly. And having 10 fails in a row is possible.. I have done it.. with 900+ resources.. architects get 55 structure assembly.. tailors get 140.
Perpetual_Dark wrote:
11%.. wait a sec.. you failed way more than that..
look at it this way.. you failed 12 bricks, plus a wall.. BUT, that puts you at making 122 bricks for 10 walls.. 22%.. this is the problem.. the math is imbedded..
If I make a structure with 3 levels of components, the failure rate goes up each time.. you are being taxed a failure rate for all 3 levels.. imagine it this way.. you get hit for 12% charge earning money.. then you get hit for another 12% putting it in the bank, and another 12% taking it out.. you are not getting hit for 12% you are getting hit for way more..
This is incorrect. We are simply talking about the failure rate as a percentage of attempts, not the amount of resources lost to failures. If you make 11 attempts at crafting (ten blocks and a wall), and you failon oneblock and then on the final wall, that's still less than 20% (2/11). Of course your resource loss was 100%, but all professions worry about this. We just use so many more resources that it is far more costly.
However, since we can theoretically sell our items for far more than other crafters (I know, the current situation doesn't always support that) it should balance out. It's like gambling at the high-stakes table as opposed to the $2 table. Your risk is higher, but potentail rewards are too. I can sell a medium house for a *profit* of 50k or more. What does the weaponsmith have to match that? If I fail, it's a big loss, but if I succeed, it's a big gain.
If the WS fails on his 10k gun, it's a smaller loss, but he has to succeed 7-8 times to match my profit.
Daynn wrote:
Regarding the "gambler's fallacy:"
In Vegas that is true. Not in this game. Computer random number generators are not ballanced dice. They are notorious for creating strings of similar results. In eq I would go an hour without a failure then fail 4 times in a row. In SWG the attempt immediately following a failure is much more likely to fail. I haven't gathered data to quantify this, but I am positive it is happening, to the point where I never attempt a hard combine immediately after a fail.
I think this statement assumes that each person has his own little random number generator that's only spewing out numbers for their individual crafting efforts. In between the time that you failed once and the time that you failed a again, the generator has produced 5000 other numbers for other people.
I think the evidence is in the fact that many other people here have said the exact opposite from your experience. I've heard plenty say you should craft small blocks until you fail, then immediatelymake your bigitem, since two failures in a row is rare.
No matter how much any of us crafts, unless you have been keeping exacting stats on every aspect of it since launch, then our anecdotes and personal experience aren't worth much. We *need* the devs to come out and tell us what affects these things, or we are all just shooting in the dark.
Yes, my data is anecdotal. But the phenomenon happens. Three years of gaming and it still bites me. I wish it was the gambler's fallacy, but this isn't craps in Vegas. And btw, who says the random number generator isn't on the client side? Most of the game is. I think it is likely that I DO have my own random number generator.
You are right though, we will never have a reliable model of how the system works until they come out and tell us. I doubt they will ever do that, so all we are left is a sea of speculation. So that is what we do.
The number of failures is irrelavent.. if you fail a wall, you dont have those 10 bricks.. they failed too.. That is what I am trying to tell you.. you didnt just fail 1 wall, you failed 1 wall with the resources for that wall, and you failed the 10 clocks that went with it. I want the devs to make it so that if you successfully make something, you dont have to gamble it over and over again until you have a deed. One large house failure IS 10 wall failures, which IS 100 module failures.. You still clicked 100 modules and still lost them.. Thats the problem.. 1 failure results in another 115 failures for a large house. You still have to build them again and you still will fail along the way AGAIN. If you have to make 1350 modules to make 10 houses, which require 1000 modules, your failure rate is 35%. You will have to make those 350 again.
And yes, you are losing the resources at a 35% rate instead of a 10% rate. Any way you do the math, a 10% failure rate is NOT 10%.
Talk about gambling? would you play a game where you had to put 100 dollars on the table, and they flip a coin, and if you lose they take the 100 dollars and if you win, you have to take another 100 and do it again. and when you have done it 10 times without losing, you have to take the 1000 dollars you got to keep and flip a coin. If you lose you lost the 1000, and if you win you keep it. Now do this whole thing 10 times, and once you have 10,000 dollars, you flip a coin one more time for the whole amount. If you win, you get to keep your money, and if you lose, they take it all. Now a coin flip has a 50% win/lose ratio right? So by your logic, your odds of having your 10,000 dollars are 50%. I would like to play this game with you, and I will make is so that if you get to that 10,000 dollar mark and win, I will pay you an extra 10,000 dollars, and we have to play the game 10 times too..
You can say that I am doing math on resources, but I am telling you, you will have to make 35% more blocks to make 10 large houses if the base failure rate is 10%.
Now take a profession that only uses 3 or 4 sub-assemblies in 2 layers instead of 3. They lose way less components. Its pretty simple math.
Now tell me again how a weaponsmith that uses 500 resources and sells his gun for 20k is having a harder time making money than us? I can sell a medium house for 60k, and I will fail about 1 in 10, along with 1 in 10 walls, and 1 in 10 structural modules. I risk losing 20-30% of my materials, then sell them for 3 cr per. The weaponsmith risks the same percentage but sells for 40 cr per. He uses 500 resources to make 20k, and even buying them at 6cr per on the bazaar, can turn out a gun faster than I can make a wall. Given 50k credits to start, the weaponsmith can make a few hundred guns, and I can make 1 house. He sells 10, makes 200k, and has about 5k into those. I burn up close to 20k resources and assuming I dont fail on final assembly, I sell the medium for 60k.. who came out ahead again?
btw, I have not met a fighter that would not pay 20k for a good weapon, but I have met fighters that wont buy a house..
Oh, and the guns will deteriorate, so in a few weeks, that weponsmith has customers again..
The only upside we have is PA halls. There is still a market that can net us upwards of 10 cr per resource. And since it is severely resource intensive, we make lots of money. So we have a huge advantage if we make master to make money, however, how many PA halls do you think you will sell in a month? in 3 months? Matter of fact, I think that at least on Tarquinas, by weeks end, over half the PA's that want a hall will have one. After that, the last PA's will have saved enough for one, and I can make 2 million a week for another week. How about the rest of the novices here? What will they do when they hit master in 2 weeks and there are no more customers looking for PA halls? Sure there will be large and medium house sales for a couple months, but at 3cr per resource profit, and losing an additional 6-40% of those resources (depending on who you talk to) on fails, your margins are pure crap, and you wont be making 2 million a week, unless you have 200k ore per day at your disposal..
Meanwhile, the weaponsmith will still be selling good guns at 10-20k that cost less than 1k resources to make.. and I dunno if he has 3 layers of buildingto do, but heck.. I will accept one PA hall failure in ten if I can make 1.5 million a piece on them.. But I failed 1 in the first 4, and at 1/3rd that cost/sell ratio. That hurt..
What does expirimenting the durability of a crafting station have to do with succes rates? Is there some empirical evidence that supports expirimenting with tools and stations as actually doing anything?
I'm an aspiring architect with very minimal help in lots and in about 3 days I've failed on 2 medium houses on final combine. I'm also a wookie so I have the added (or so they say) assembly bonus. I've calculated that over the last few days I've lost about 60k resources not to mention the sales and customers that question my ability now and have gone elsewhere.
(and whats Sony's issue making us have to repeatedly log in to everything??)
OK, I RARELY critical fail. You guys are talkin 10%??? I made 10 wall modules last night and failed on 3 structure parts. And *I* was thinking my crits were high that night.
I'm not a master-I've got buildings 1, and installations 4. Thats it. I AM a master artisan.
I think some people need to go back and get some of their basic artisan skills completed.
If I failed that much I;d be hating life. seriously.
And master architecs-WTH are you thinking? Crank up that ol factory-it doesn;t fail. leave it running making identical structural modules 24/7.