Architect Archive
Thread: How well can a 0-4-x-x dabbler make harvesters versus a master?
Page 1 of 2
Pawlin
Tue Feb 22, 2005 12:02 am
#1
Ok, I decided to grind up architect with my character Oprolan on Sunrunner for no good reason.
While I was at it I decided to take the opportunity to test out how well a dabbler can make stuff compared to a master and if you can hit maximum BER with only 9 experiment points.
Oprolan is a 'dabbler' as a 0-4-2-1 architect Wookiee which gives him 90 structure experimentation and 20 structure assembly.Pawlin is a master architect Mon Calamari which gives him 100 structure experimentation and 110 structure assembly (without clothing mods but +10 structure assembly racial bonus). The building and harvester trees don't give any stat bonuses so aren't relevant to quality of experimentation.
The test: Oprolan and Pawlin both made 20 light ore mining units (LOMU) and then used the 'good' parts to make as many mineral mining installations (MMI) as possible.
For Oprolan I did two tests with the LOMU's one test with with a basic quality crafting station that I've got in my house and the other was with a +41 station I found in a public room in a shopping mall. I only ran one LOMU test with Pawlin. I did my first test on Oprolan with my 8.8 station I made myself but then thought that I'd better test with a high rating station to make sure that doesn't have a significant impact.
LOMU's are really only useful if you can get them into at least the 85-90% minimum range.
Oprolan: LOMU's at +8.8 crafting station
29%, 91, 90, 90, 76, 90, 77, 76, 86, 90, 48, 86, 45, 45, 94 (double amazing), 41, 76, 90, 62, 34, 62
== 9/20 usable > 85%
29%, 91, 90, 90, 76, 90, 77, 76, 86, 90, 48, 86, 45, 45, 94 (double amazing), 41, 76, 90, 62, 34, 62
== 9/20 usable > 85%
Average experiment result was 71% out of max 94%.
Oprolan: LOMU's at +41 crafting station
76%, 90, 86, 77, 90, 81, 90, 76, 90, 77, 90, 49, 86, 90, 76, 90, 77, 86, 90, 76
== 11/20 usable >85%
76%, 90, 86, 77, 90, 81, 90, 76, 90, 77, 90, 49, 86, 90, 76, 90, 77, 86, 90, 76
== 11/20 usable >85%
Average experiment result was 81% out of 94% max.
So note that the +41 station might be up to 10% better results than the +8.8 station for someone with low stat points. The sample size is small so its hard to know for sure the impact but its not a huge one.
Pawlin: LOMU's at +44 crafting station
95, 95, ... 95 (all came out maximum)
== 20/20 usable >85%
95, 95, ... 95 (all came out maximum)
== 20/20 usable >85%
Average experiment result was perfect 95% out of 95% max.
So this gave me a total of 20 usable LOMU's for Pawlin and Oprolan. But for Oprolan I made a couple of them into schematics and my tool ate another (I found it later). So minus those 3 left me 17 LOMU's for Oprolan to make MMI's from.
So I then tried making mineral installations out of 20 LOMU's for Pawlin and 17 for Oprolan.
Oprolan: MMI's from 90% and 86% LOMU's (percent / BER rate)
90's: 90/10, 42/8, 92/9, 76/9, 0/6 (crit fail), 49/8, 90/10, 90/10, 76/9, 85/10, 65/9, 90/10
86's: 90/10, 90/10, 48/8, 62/8, 76/9
== 7/17 BER10, 5/17 BER9, 4/17 BER8, 1/17 BER6
Oprolan used all points to get BER. No points were left for storage hopper size.
90's: 90/10, 42/8, 92/9, 76/9, 0/6 (crit fail), 49/8, 90/10, 90/10, 76/9, 85/10, 65/9, 90/10
86's: 90/10, 90/10, 48/8, 62/8, 76/9
== 7/17 BER10, 5/17 BER9, 4/17 BER8, 1/17 BER6
Oprolan used all points to get BER. No points were left for storage hopper size.
Pawlin: MMI's from 95% LOMU's (percent / BER rate)
83/10, 91/10, 83/10, 83/10, 82/10, 83/10, 83/10, 83/10, 91/10, 34/8, 41/8, 55/9, 83/10, 83/10, 83/10, 83/10, 91/10, 76/9, 83/10, 84/10
== 16/20 BER10, 8/20 BER9, 2/20 BER8
83/10, 91/10, 83/10, 83/10, 82/10, 83/10, 83/10, 83/10, 91/10, 34/8, 41/8, 55/9, 83/10, 83/10, 83/10, 83/10, 91/10, 76/9, 83/10, 84/10
== 16/20 BER10, 8/20 BER9, 2/20 BER8
For the BER10's I usually had 2 points left for storage hoppers, but on at least one occasion I had to put all the experiment into BER in order to achieve BER10.
So Oprolan had to make twice as many LOMU's as Pawlin did to get the same amount of usable quality parts. As a master, Pawlin had perfect results when making LOMU's.
Oprolan had to make 17 harvesters to get 7 BER10's or 2.4 harvesters per 1 max BER and Pawlin only had to make 20 harvesters to get 16 max BER or 1.25 harvesters for 1 max BER.
Conclusions:
A 0-4-x-x architect has to make TWICE as many parts and harvesters as a master does in order to get maximum BER.
plus..
A 9 point architect CAN make maximum BER medium harvesters.
Next I'll test making heavy harvesters and also check how well Oprolan can make stuff with 2-4-x-x skill for 90 structure experimentation and 50 structure assembly.
CoretDenvin
Tue Feb 22, 2005 10:02 am
#2
Excellent job Pawlin. Now if we only had a reason to use experimentation points for anything other then harvesters.
I have a feeling your not going to get BER13 heavy mineral harvesters on the next trial. I think you'll hit BER12. When I was in the process of mastering architect I made a couple heavies and was using very HQ stuff. I had to dump all Exp points into the quality just to get the BER12.
Though I don't remember off hand the quality and parts used for the OMU's. So that obviously would have been the deal breaker. Will be interesting to see.
I have a feeling your not going to get BER13 heavy mineral harvesters on the next trial. I think you'll hit BER12. When I was in the process of mastering architect I made a couple heavies and was using very HQ stuff. I had to dump all Exp points into the quality just to get the BER12.
Though I don't remember off hand the quality and parts used for the OMU's. So that obviously would have been the deal breaker. Will be interesting to see.
Crimsonsplat
Tue Feb 22, 2005 10:14 am
#4
How did you control for quality of the resources across both servers, when making the LOMU's?
CoretDenvin
Tue Feb 22, 2005 10:31 am
#5
LonelyGhost wrote:I thought harvs were the third tree? As in x-x-4-x?
They are - but you can make Medium Mineral Harvesters at the second tier. Which is what Pawlin's second character had.
Pawlin
Tue Feb 22, 2005 12:03 pm
#6
Crimsonsplat wrote:
How did you control for quality of the resources across both servers, when making the LOMU's?
I couldn't get the exact same resource quality obviously.So I made due with what I had and at least used steel in the same 'ballpark' of 900-950 range for stats.
I am working with the assumption that quality of materials does not impact the success level for experimentation results as far as the frequency of bag, good or great successes.
I can retest on Oprolan as a master with the same materials just to make sure that the material quality doesn't impact experiment success results.
Pawlin
Tue Feb 22, 2005 12:21 pm
#7
Hang on a minute! It just occurred to me that I think the city Pawlin lives in is a research city. I'll have to check when I get home but I think thats what Crimson Flats is now. I am so used to living there that its somethign I totally forgot about. I do believe that the research city would impact the quality of the experiment results a noticable amount. If it is a research city then that would be too much of an apples to oranges comparison and would throw the experiment out of whack.
I'll have to run another LOMU test with Oprolan as a master architect and that will be a better comparison to the tests I already ran on Oprolan. And/or I could run another LOMU test on Pawlin somewhere else to see how much difference the research city makes.
In general this isn't meant to be extremely scientific.The sample sizes I usedare fairly small. And there are various little differences between the 2 characters that I'm ignoring. Like I haven't even mentioned the quality of the crafting tools I used, and I'm just assuming it doesn't matter.
If anyone thinks of other major factors that I'm not mentioning then please let me know. I might be assuming they don't matter but I might have forgotten something.
Crimsonsplat
Tue Feb 22, 2005 6:26 pm
#8
Definately a few other things I'd like to make sure were consistant here.
Like flushing the RNG with a macro to do 100 combines quickly after two bad experiments in a row.
Always clicking the final combine with your left hand.
Discarding any components with a "13" anywhere in the serial number.
Making sure all factory made parts were from runs of an even number of units.
Standing up, turning around three times, and bowing in the direction of MIT (east coast) or CalTech (west coast) before you do the first combine. (Non-Americans substitute technical university of your choice).
You can't expect us to comment on the validity of your test results if we don't know whether you followed proper procedure or not!
Like flushing the RNG with a macro to do 100 combines quickly after two bad experiments in a row.
Always clicking the final combine with your left hand.
Discarding any components with a "13" anywhere in the serial number.
Making sure all factory made parts were from runs of an even number of units.
Standing up, turning around three times, and bowing in the direction of MIT (east coast) or CalTech (west coast) before you do the first combine. (Non-Americans substitute technical university of your choice).
You can't expect us to comment on the validity of your test results if we don't know whether you followed proper procedure or not!
Pawlin
Tue Feb 22, 2005 6:47 pm
#9
Crimsonsplat wrote:
...Like flushing the RNG with a macro to do 100 combines quickly after two bad experiments in a row. ...
I honestly did actually think of doing something to flush the RNG when I got the batch of 3 bad experiments when making MMI's with Pawlin.
/target self
/sigh
Pawlin
Tue Feb 22, 2005 9:58 pm
#10
Ok, my next test:
Q: Can a dabbler with 9 experiment points make BER13 heavies?
I ground Oprolan up to 0-4-4-1 so that I had the full line of harvesters available yet still only 9 experimentation points and 20 structure assembly skill.
I decided to try making heavy chemicals since those seem to be the ones people have the most problems with. First I had to make a Turbo Fluidic Drilling Pump. This wasn't an easy task. To get BER13 harvester I would have to get something in the 85-90% range as a minimum.
It took 12-13 tries to get a TFDP up to 90%. I didn't take exact records on the numbers and am not sure if it was 12 or 13. I had a lot of lousy experiment results and a few near misses in the 70-80% range.
When I did get a TFDP at 90% I made a manufacturing schematic and ran off 20 of them. I also made schematics and did factory runs of structure modules, generator turbines and small structure storage modules.
Once I had all the parts I started to try making the Deep crusts.I made the walls by hand to get a little more XP.
Deep crusts:
35% = BER11, 76% = BER12, 80% = BER12, ? % = BER12 (failed to write down %), and finally 90% = BER13
So it took me 5 tries but I was able to finally make a single BER13. So the answer is:
A: Yes, 9 point Architects CAN make BER13 heavy harvesters.
Pawlin
Tue Feb 22, 2005 10:00 pm
#11
For reference, here are the material stats that I've used with Oprolan on Sunrunner:
Steel: omnikawoksis duralloy steel
HR 952, SR 986, UT 930
Iron: Tifeian Kammris iron
HR 939 SR 993 UT 989
HR 939 SR 993 UT 989
Ore: zinsiam carbonate ore
HR 971 SR 614 UT 945
HR 971 SR 614 UT 945
I've used the Iron whenever possible since I've got more of it and its higher quality than the steel.
I use lub oil for the chemical and the quality of the rest of the materials isn't consequential.
Pawlin
Tue Feb 22, 2005 11:07 pm
#12
If my structure assembly points go up then will my experimental success rate go up proportionally?
Considering that I got 9/20 success or 71.6% with 20 structure assembly and 20/20 or 95% experiment average with 110 structure assembly then 70 structue assembly should give about a number 5/9ths between or roughly 15 / 20 success or 84% average.
So I grinded up a few boxes of the furniture tree to make Oprolan 3-4-4-1 and giving him 70 structure assembly points.
I then reran the LOMU test doing another 20 assemblies at my +8.8 crafting station.
Oprolan: 70 structure assembly, LOMU's at +8.8 crafting station
62, 62, 90, 90, 90, 90, 62, 90, 91, 90, 62, 90, 62, 90, 58, 0, 90, 91, 94, 90
62, 62, 90, 90, 90, 90, 62, 90, 91, 90, 62, 90, 62, 90, 58, 0, 90, 91, 94, 90
== 13/20 usable > 85%
Average experiment result was 77% out of max 94%
This isn't exactly the 15/20 and 84% predicted but its close enough considering the small sample size.
Ok, now on to the MMI's
With 20 structure assembly I got 7/17 MMI's to hit BER 10 and with 110 structure assembly my success rate was 16/20. So thats 41% and 80%. So 70 structure assembly ought to result in about 63%. I've got 13 usable LOMU's so I ought to be able to get about 8BER 10's of them with the 70 structue assembly.
Oprolan: 70 structure assembly, MMI's from 90%+ LOMU's (percent / BER rate)
90's: 90/10, 90/10, 90/10, 48/8, 0/6 (crit fail), 69/9, 90/10, 91/10, 90/10, 90/10
90's: 90/10, 90/10, 90/10, 48/8, 0/6 (crit fail), 69/9, 90/10, 91/10, 90/10, 90/10
91's: 90/10, 90/10
94: 62/9
==9/13 BER10, 2/13 BER9, 1/13 BER8, 1/13 BER6
The 9 out of 13 BER10's is pretty close to the 8/13 that was expected.
So the LOMU results were a little lower than expected and the MMI results were a little higher than expected. But both aren't too far off.
YES it does appear that as structure assembly points increase your experimental success results increase proportionally.
Crimsonsplat
Wed Feb 23, 2005 10:31 am
#13
Pawlin wrote:
Crimsonsplat wrote:
...Like flushing the RNG with a macro to do 100 combines quickly after two bad experiments in a row. ...
I honestly did actually think of doing something to flush the RNG when I got the batch of 3 bad experiments when making MMI's with Pawlin.
/target self
/sigh
That one goes back to my earliest days of crafting in EQ. I started with something almost reasonable just to fake you out. Guess I faked me!
Page 1 of 2