Architect Archive
Thread: Extensive crafting station tests, and my results: (crosspost from weaponsmith forum)
Well, I wanted to find out of "good" crafting stations let me experiment any better than "bad" crafting station. So I ran a bunch of trials.
Executive summary: After 350 units produced, it appears that the quality of a crafting station either has NO effect on the quality of the output, or it has a REVERSE effect (where a good station results inWORSE output).
---------------
Background: I'm a weaponsmith with tech4. My weapon assembly and weapon experimentation skill mods are both +75. I used 4 identical crafting tools, each with an effectiveness rating of 12.31.
I decided to grind and experiment on Gaderiffi batons. Complexity is 20. The primary attribute there is Shock Resistance. I made each one with 28 units of Kiirium steel (SR 558, Mall 63), and 20 units of Kammris Iron (SR 945, Mall 514). Since everytime I complete the initial assembly I got a 'great' or 'amazing' success, this meant that the base quality of every weapon was 18% before experimenting.
To experiment, I allocated all 7 points of experimentation into damage in one fell swoop. The listed risk was 17% (no matter which crafting station or toolI used). In terms of prose descriptions, when experimenting I got lots of 'moderate successes' and 'good successes'. I also got some 'failures', 'moderate failures', 'successes', and 'great successes'. (Of course, when I'm trying to make weapons to sell, I only concern myself with 'great successes'.)
All trials were conducted on Saturday August 23, and Sunday August 24.
THE SUCKY STATION (Functionality rating -12.846494, Day 1)
100 units attempted
3 critical failures on assembly (17.46% average initial assembly quality among all units)
Amongsurviving units, average experimented improvement was 27.1%.
Experiments included9 'great successes', best single unit was experimented to +69% damage.
THE PUBLIC STATION (Rating unknown, next to bank in Keren, Naboo, Day 1)
100 units attempted
4 critical failures (17.28% average initial assembly quality)
Average experimented improvement among survivors: 27.3%
Experiments inluded 15 'great successes', best single unit was experimented to +67% damage.
THE UBER STATION (Rating +21.080246, Day 2)
100 units attempted
12 critical failures (15.84% average initial assembly quality). (3 of these crits followed immediately after a preceeding crit. So if you you want to be charitable, you could say it's only 9 crits)
Average experimented improvement among survivors: 25.9%
Experiments included only 4 'great successes', best single unit was +62% damage.
It's worth noting that the crafting *felt* really bad with this station, due to the high number of crits, and the low number of great successes when experimenting.
At this point, I began to worry that maybe i was going to have bad luck no matter what I did, so I busted out the crap station again:
THE SUCKY STATION (Functionality rating -12.846494, Day 1)
49 units attempted
0 critical failures (18% average initial assembly quality)
Experiments included 4 'great succeses', best unit was +68% damage.
-----
When Istarted this, I had the hypothesis that a +21 station would be 33% better than a -12 station. With the above results, if you compare experimented improvement, the 'good' station is about 4.5% worse than the 'bad' station.
To be honest I have no idea of what to make of the numbers. But I think I'm goinng to avoid the good station for now, if only because of the alarming number of critical failures during the trial.
PS - I only kinda-sorta kept track of 'amazing successes' on the initial build, since they don't have any impact on quality. The number of amazing successes seemed similar between the two stations.
PPS - it would have been useful to test this while my weapon assembly skill (and possibly tool quality) were worse, because if I ever got less than great succeses on initial assembly, it might shed more light on the situation.
finally someone posting some real numbers to look at and.... oh, what a surprise... no impact from the great station.
sorry to hear it, but its really not a surprise to me.
Surely you're mistaken Easy, you aren't implying that something is actually broken or doesn't work according to the way logical minds think are you?
On a serious side, great job and thanks for posting this info 
Well, to be fair, the testing is probably better suited to small items, because with the relatively small crit failure chance, you wouldn't want to do 100 trials using large houses.
Some architect might want to try it with subcomponents or something.
Nice to see some numbers.. I suspected as much. I've been getting stations easily up into the 35-36 functional rating area, and I was noticing no difference in experimentation or number of crit fails (and I use high quality materials).
So its still all comes down to dumb luck on crit failures. Think I'll stop building again.
Nice post Easy, it concurs with my findings so far. I even would assume the high number of crits with the UBER station was just a coincidence. I just wish one of the mighty DEVS would shed some light on this matter. It bothers me to no end they don't even feel obliged to explain why the Functional Rating and Effectiveness were included at some point. Not to mention not putting them in the patch notes in the first place.
Leon
From your data, one could easily conclude that on day 1, you had good results... then the devs changed something in the patch, causing your day 2 numbers to go bad.
We need more testing on this tho, as it's a very worthwhile issue.
The strange part is... I sell 40+ functionality stations, and all my customers swear up and down that they get more amazing and great experiments, with fewer crit and failed experiments.
They don't say much for the actual combine crit rate tho...
But please, if you can... Perform the same(or a similar) test all in one day.
-Loserwan Ka'newbie.
Personally using the same tools and station I have notice there are times when I will crit fail a lot and then an hour later I can do no wrong. The failure rate for me seems to go with server load. Early mornings and late at night things work fine for me, but come prime time the failures seem to be higher. Once you hit the stream of bad crits, you just got to stop and wait for things to settle down and stop criting. There is so much voodoo to this it is hard to tell what is and isn't important.
At least the input hoppers work now. 
Line40 wrote:
I even would assume the high number of crits with the UBER station was just a coincidence. Leon
i really have to disagree...my experience has been that stations seem to INCREASE critical failures....
say structure modules...
I can build 30-40 on my front porch with maybe 1 -2 critcal failures
if i am near a station that seems to go up...i think that there might be some bug in some line of code...maybe a negative or a decimal typo
I haven't experimented as much as far as keeping detailed numbers, but I can tell you two things that seem to ring true in my experience.
I spent all weekend crafting new stations and working with various grades and functionality ratings, and here are my two observations:
Increased function rating definitely seems to improve the upper end of my experimentation, but not by much. With my old station (Function around 6) I crafted a number of master-artisan level items and never got over 92%, and the results were very consistent. I crafted a station with function of 34 and started churning out the same items using exactly the same materials. Suddenly the top end was 94%. However, as the weekend wore on, I noticed my experiments resulted in far fewer "Great Success" messages and far more "Good" or "Marginal" results. I didn't see an apparent increase in critical failures.
I also made myself an architect station at functionality 28, and saw similar results. Higher highs, butless frequently. Anyone else seeing this?
In the meantime, I've started selling these high-end stations on Gorath for anywhere from 4000 (function around 16) to 18000 (Function 34.5). I've sold a few already, but we'll see how they sell I guess.
Is there any correlation between the crafting tool rating and the crafting station rating?
is there any understanding as to which is more important or do they both have equal effect?
I'm pretty sure that station quality and tool quality have equal effect, ie. none at all.