Estei’s Guide to experimentation
Inspired by a post I made on an experimentation question I have decided to sit down and make a little guide on experimentation and different ways of going about it.
Experimentation points and values of successes
As master you get ten experimentation points that can be upped to twelve points by procuring some experimentation “skilltapes”.
Skilltape is the common word for Clothing Attachments that you can put into socketed clothes. For every ten points of armor experimentation you put in clothes you will get one extra point to work with when experimenting (+10 = 11 points in all +20 = 12 points in all).
Experimentation is like many other bonuses you can get through the use of skilltapes, capped at +25 so the max amounts of experimentation points you can have to work with will be twelve.
Value of successes
Crafting in “swg” is based on percentages. This means that the different kinds of successes and failures you get are also based in the same domain.
When experimenting, you will depending on how lucky you are add or subtract to how good your product is in percent up to 100%. How far up you can experiment is capped by the quality of your resources.
This I will be talking about in the part about the math of crafting. Normally I only work with the parts where I get either a great or an amazing success, so I will concentrate on these here.
A great success will give you 7%, and an amazing will add 8% to your experimentation line.
Force Sensitive Experimentation
At the village of Aurillia you can through your knowledge of the force enhance your experimentation skills. Through the FS Experimentation line you can get up to +20 experimentation. These +20 do not give you more experimentation points, they will on the other hand greatly enhance your chances of getting amazing successes. For more experimentation points you will still have to procure the expensive experimentation skilltapes.
Math of crafting
Most of this is already covered in “The Complete Guide to Armor and Armorcrafting” stickied at the top of the armorsmith forums. But when that is said some of the math used in the said guide aren’t entirely accurate and there are better formulas out there that are more precise. The ones I use here I have “filked” from the helper tool “AS little helper” that I have linked to in the links section.
Why you want to know all this, you might ask. My answer will be. Fractions baby!
Fractions count all the way up through your craft so leftover fractions from your layers will influence you segments and fractions from your segments will influence your final product. This might not always show but trust me they are there and sometimes you can get a lot out of them.
This is the weighted average for the stats that are important for the experiment. For instance OQ 50% /SR 50% for resists.
The weighted average is used to determine how good a product you will be able to make. This hasn’t changed from the formulas used in “the complete guide” so I will take the liberty of “filking” it from there.
(( ( ((OQ1*n1) + (OQ2*n2) + (OQ3*n3)) / (n1+n2+n3) ) + ( ((SR1*n1) + (SR2*n2) + (SR3*n3)) )/ (n1+n2+n3)) )/2
OQ1: Material 1 Overall Quality
SR1: Material 1 Shock Resistance
n1: Number of Material 1 required in schematic
OQ2: Material 2 Overall Quality
Material 2 Shock Resistance
n2: Number of Material 2 required in schematic
OQ3: Material 3 Overall Quality
SR3: Material 3 Shock Resistance
n3: Number of Material 3 required in schematic
The assembly is the same whether you get a great or an amazing. Don’t ask me why but that’s the way it is *shakes head at the devs*
The formula for calculating what your initial assembly percentage isn’t the most straightforward and I won’t really say I understand why it looks as it looks but it works :-)
Y = X*(0,000015*X+0,015)
X = The weighted average calculated above
Y = The initial assembly percentage
The max experimentation is very interesting in the respect that it determines just how good your product can get. Again this will also reveal to us those lovely fractions that we can play with to really get the best out of our materials.
This one is simple as it is directly linked to your weighted average. It’s actually just the percentage of your weighted average out of the perfect 1000 that is max for resources.
P = (X/1000)*100
P = X/10
X = The weighted average calculated above P = The max experimental percentage possible.
Experimentation is just adding the experimentation results to your assembly percentage, keeping in mind that it’s capped at the max percentage for your resources. So the formula would look something like this
Z = Y + (N1*7) + (N2*8) (< P)
Z = The experimented percentage
N1 = Amount of great successes
N2 = Amount of amazing successes
Example of the formulas in action
We will here be making a kinetic layer using the formulas and three fictional resources.
8 * Polymer: Overall Quality = 900, Shock Resistance = 800
8 * Amorpheus Gemstone: Overall Quality = 900, Shock Resistance = 800
20 * Metal: Overall Quality = 900, Shock Resistance = 800
X = (((((900*8) + (900*8) + (900*20)) / (8+8+20)) + (((800*8) + (800*8) + (800*20)) / (8+8+20)))/2
X = 850
Y = 850*(0,000015*850+0,015)
Y = 23.5875%
Max Experimental percentage:
P = 850/10
P = 85%
We play that we are using 7 points in this line and we got 3 amazing and 4 great successes.
Z = 23.5875 + (4*7) + (3*8)
Z = 75.5875%
You might have been thinking about the fractions I’ve mentioned a couple of times, thinking what is the guy talking about. Well this section will try and explain that.
To really get to know our fractions we will need to know the different cutoffs. Cutoff is the percentage where our values will make a jum. For this example I will use one of those I know, the base on layers.
The cutoffs are every 25%
1-24.99%: 1-1.99% Base
25-49.99%: 2-2.99% Base
50-74.99%: 3-3.99% Base
75-100%: 4-5% Base
Now back to those fractions. If we take our example from the math of crafting where we with three amazings and four greats got to 75.5875%, this would have taken us to 4% base which is ok.
If this was triple layered comp we would have been able to add 12% to the base of our segments from these layers.
But if we instead had capped the experimentation and taken it to 85% we would see a little but very significant difference.
Let’s calculate the exact base of this one instead. We know that it will be 4% because we are over 75%, but we want to know the fractions also, so we take the 85% we experimented and subtract the 75% needed to get to 4% leaving us with 10%.
To find out how much base this gives us we will need to find out how big a fraction 10% is of the 25% needed to go to the next cutoff.
10/25 = 0.4
Now we add the two numbers 4 and 0.4 giving us a base in all of 4.4%.
Now we not only add 12 but 13.2% to the segment because of the fractions :-)
These fractions carry all the way through so these leftover 0.2 that we probably won’t see make a difference in the segment can still play a role. Maybe the segment calculation will yield a fraction of 0.4 also and the final product might also have an 0.4 fraction. This will result in 1 extra percent base in the final product. (.2+.4+.4 = 1)
There are some fractions that you can see when making your segments these are 0.33 and 0.66. 0.33 will give you one extra percent on the segment base while 0.66 will give you 2%, making it possible to add 14%+ from three layers even if the layers only say 4% base.
For ease of use I have made this little table for layer base showing at which percentage you will get the good fractions that we can see on segments.
Risk is a modifier that is added to your experiment when experimenting. This increases your chance of getting critical failures and lowers your chance of amazings.
From my experience risk is only added to AS experimentation when you cross lines. Like say Durability and Resists at the same time. I know this differs from craft to craft, for instance artisans seem to be punished already at three or four points in the same line.
For each time you push the experiment button your raise the complexity of your product. Each point of complexity means 8 seconds extra factory time per piece. This can result in a significant difference. If you make a run of 1000 layers 8 seconds will result in an increase in factory time of 8000 seconds, or 2 hours 14 minutes per point of complexity.
I normally use one of three methods for experimentation. Which one I use is heavily dependant on what I’m making and what kind of components I’m using. They all have their pros and cons so I have tried to put in the ones I could come up with.
"The analytical way"
The analytical person would sit down and check the numbers thinking "ok I need to free up a point how many amazing would I need to do this?"
A little example might be in order here :-) After you assembly your line starts at 27% and you want to max out the line at 93%. Getting only great successes you would need 10 points to max out this line 10 * 7 = 70 + 27 = 97%
Well we know from the little calculation above that with 10 greats we get 97%. One less point would mean 90% So to free up a point we would need 3% extra from amazing successes, ergo we could free up a point by doing 3 points at a time resulting in 6 greats and three amazings. 27 + (6 * 7) + (3 * 8) = 93%
Pros and Cons
Three chances of getting the amazings you want
Cons: Three chances of getting “crit fails”!
8 Seconds extra per experiment
I find this way is good for final combines when not working for factories, because it gives those three chances of freeing up that elusive extra point. Great for special orders using looted subcomponents. If you know your resources well and know where they cap because you would want to make your client the best possible armor :-). (bespin port is also your friend in this situation :-)
"The easy/risky way"
The easy way of doing is the way I do it on most things because its.... well.... easy. 7 points that are amazing will free up one point seeing as 8*7 = 56 and 7*8 does the same ;-) (rocketscience)
That will leave you an extra point to work with.
Pros and cons
Easy and you don’t have to do any calculations
Only use two experiments to fill out the line, resulting in less complexity again resulting in less factory time
Only one chance to get the amazing you are looking for!
I use this for making regular non enhanced layers and segments because I don’t want to do all the math and the complexity is most likely lower than doing the analytical. You might burn through some resources but sitting in a research center town you should be able to get that amazing pretty quick. (my record in my current shop is 12 tries on segments and maybe 15 on layers. That is the worst Ive tried by the way. Sometimes you get lucky and get it on first try :-)
"The double amazing/Even more risky way"
Seeing as fractions count for something all the way up through the craft you sometimes want to get as much out of your points as possible. Especially when working with Special resist maxed layers, like stun and kinetic layers for ubese and padded.
What we do when using this method is to put as many points as possible into the line you want to max.
Ill use my stun as an example. When doing stun layers I can put 7 points into resists on the first experiment. If I do this and get an amazing it takes one more great success to max out the line. So knowing I need to save one point I go all out and place 7 points in resists and 4 points in base.
I continue to try until I get an amazing on these eleven points then top off the resist line with the last point.
Pros and Cons
Lets you get your absolute max out of your resources!
Even less complexity (only 16 seconds extra per unit)
Risk! When doing this the little risk meter at your right will show that you are on more shaky ground than usual :-(
Can burn up great amounts of resources before you get that amazing.
I only use this one on layers, because they don’t take enhancers or subcomponents so the waste is minimal. If you can say minimal waste when going through that many resources. But what can I say the results speak for themselves. It got me from 78%/51% padded to 80%/52% recently.
10 pts. vs. 12 pts.
To get the max out of your resources takes some knowledge of the working of crafting and you resources. Many people complain that because they are "only 10 pointers" they can’t make as good armor as 12 pointers. Well there are some niche products that might be a problem. I see heavy stun armor as one of them as the 10 pointer will have a hard time getting as high base as the 12 pointer. The other is low encumberance comp where the extra points in layers and segments really make a big difference. But using the methods above the "10 pointer" can really get a lot out of his resources.
AS little helper
: A great tool when you want to experiment but don’t want to use resources.
: No intro needed. Great place to track resources.
: Uses the swgcraft.com DB and order them by profession
Message Edited by zaphb on 01-25-2005 01:35 PM