Doctor Archive

• BEC: 3 Power, 6 Charges


• CRDM: 17 Power


• LS: 129 Power

• Combined Power of Subcomponents (Z): 149


• Combined Charges of Subcomponents (Y): 6


• Stimpack Power: 257 (At 25% Effectiveness)


• Stimpack Charges: 23 (At 15% Effectiveness)

Stimpack 2

• BEC: 11 Power, 20 Charges


• CRDM: 64 Power


• LS: 129 Power

• Combined Power of Subcomponents (Z): 204


• Combined Charges of Subcomponents (Y): 20


• Stimpack Power: 312 (At 25% Effectiveness)


• Stimpack Charges: 37 (At 15% Effectiveness)

Now, if we substitute our experimentation results into our earlier forumla, we get this:

312-257 =204 - 149

That, of course, breaks down to this:

55 = 55

From that experiment, it would appear that our initial assumption that the values of subcomponents are added to the "base" value is true. You can do the same thing with charges to see that it works for that attribute, as well, but I won't include that here. You may have noticed that I noted the experimentation level at which I recorded the stimpack power and charges. The reason for this is that, after assembly but prior to experimentation, your Experimental Effectiveness is already greater than 0%. Therefore, when an item is initially assembled, it is as if some degree of experimentation has already been done to that item. However, as I used the same resources to craft each stimpack, the experimentation numbers were identical for each stimpack, which allows me to go forward with my comparison.

On the other hand, this discovery poses a new quesion which must be answered before going forward. Is experimentation of both stimpacks equal? In other words, will an additional 7% of experimental effectiveness (the equivalent of 1 experimentation point resulting in a great success) produce the same power increase in stimpack1 as it would in stimpack 2? My assumption was that the increase in power from experimentation is based somehow on this "base" value, not on the resources that are used to produce the item. Of course, to determine which is true, we need only to experiment on each stimpack and record the results. Here is what I found:

Effectiveness: Stimpack 1 Power Stimpack 2 Power

25% 257 312

32% 266 321

39% 275 330

At this point, the percentage values begin to differ (I achieved an amazing success on Stimpack 2 while achieving only a great success on Stimpack 1). From this small sample, however, it seems that every additional 7% of effectiveness equates to an additional 9 points of power for either stimpack. Therefore, the subcomponents that were used to craft each stimpack are irrelevant. Of course, this is a small sample of data. I continued this experimentation until I ran out of points and this is what I achieved for my highest value:

• Stimpack 1 Power: 337 @ 89% Effectiveness


• Stimpack 2 Power: 385 @ 83% Effectiveness

I can then generate an approximation to the value of a single percentage point of effectiveness to each stimpack by using the simple slope forumla: (y2 - y1)/(x2 - x1)

Stimpack 1: (337 - 257)/(89 - 25) = 80/64 = 1.25

Stimpack 2: (385 - 312)/(83 - 25) = 73/58 = 1.259

Given rounding that occurs along the way (it's impossible to have a Stimpack that heals for 350.5 damage), I believe this is suffuciently accurate to state that the experimentation is based on the item you're producing, not on the subcomponents being used to produce it. This formula also holds true for charges, but I won't bother to show that here.

PRE-EXPERIMENTATION CONCLUSIONS:

So what have we learned so far? Well, in my case, I hadn't learned anything new - I had only confirmed what I had suspected all along. We now know that the values of subcomponents are added directly to the item being made. That seems to prove that some "base" value exists. We have also proved that experimentation is constant over all items of the same type - regardless of the components that are used to craft a given item, say a Small Stimpack - B, the results of experimenting on that item will always be constant. In the case of a Small Stimpack - B, each point of experimental effectiveness is equal to roughly 1.25 points of power.

EXPERIMENTATION:

Finally, with all of that work out of the way, we can start really testing what we're after - the base values for any given item. Now, we just need to determine how we'll get that value. Based on the previous work we've done, we know that the base value of an item is equal to the power of the item (or charges, etc) minus the combined power of the subcomponents minus any effect of experimentation on that item at that point. Mathematically, it looks like this:

BASE_POWER = POWER - Z - (EFFECTIVENESS * 5E)

You can see the new term that I've added to the equaton: EFFECTIVENESS * 5E. That term represents the current experimental effectiveness of the item multipled by the effect of 1% of effectiveness (5E).

So, based upon that equation, I need to gather a handful of data for each item. I will need to craft each item and record the initial attributes of the item, the attributes of the subcomponents used in the crafting process, and the initial experimentation level of the item. Once I've got that information, I need to fully experiment on any one attribute (power, charges, medicine use, etc). This experimentation allows me to determine my 5E component.

It's important to note that, even though the effect of experimentation appears to be linear in nature, in practice, it is not truly linear, but stepped. A stepped equation is one which holds a given value for some period and then "steps" to the next value, then holds at that value for some time until "stepping" to the next value. For example, A Small Stimpack - A might have 11 Charges at 15% and 11 Charges at 22% before finally jumping to 12 Charges at 29%. The reason for this "stepping" is rounding - a Small Stimpack - A can't have 11.3 charges, so it remains at 11 until the experimental charges are high enough to cause it to round up to 12. Because of this, it is important to use high quality resources to do this testing. High quality resources will allow me to experiment to higher levels and, thereby, acquire two points on that line that are farther apart. The more distant those two points are, the more accurate my approximation becomes.

So, now that we've got the methodology laid out, let's see what some of my results were:

Small Stimpack - A

*For a Small Stimpack - A, no subcomponents are required, so our "Z" and "Y" valuesare 0.

• Charges: 11 @ 15%, 16 @ 62%, 5E = 0.10638


• Power: 81 @ 25%, 95 @ 81%, 5E = 0.25

Base Charges = 11 -Y - (15 * 0.10638) = 11 - 1.5957 = 9.4 = 9

Base Power = 81 - Z - (25 * 0.25) = 81 - 6.25 = 74.75 = 75

• Base Charges = 9, 5E = 0.10638


• Base Power = 75, 5E = 0.25

Small Stimpack - B

*For this experiment, my "Z" value was 204 and my "Y" value was 20

• Charges: 37 @ 15%, 44 @ 62%, 5E = 0.1489


• Power: 312 @ 25%, 392 @ 89%, 5E = 1.25

Base Charges = 37 - 20 - (15 * 0.1489) = 17 - 2.2335 = 14.7665 = 15

Base Power = 312 - 204 - (25 * 1.25) = 108 - 31.25 = 76.75 = 77

• Base Charges = 15, 5E = 0.1489


• Base Power = 77, 5E = 1.25

Small Stimpack - C

*For this experiment, my "Z" value was 204 and my "Y" value was 20

• Charges: 44 @ 20%, 56 @ 77%, 5E = 0.2105


• Power: 381 @ 25%, 573 @ 89%, 5E = 3


• Med Use: 31 @ 25%, 21 @ 89%, 5E = -0.15625

Base Charges = 44 - 20 - (20 * 0.2105) = 24 - 4.21 = 19.79 = 20

Base Power = 381 - 204 - (25 * 3) = 177 75 = 102

Base Med Use = 31 - (25 * -0.15625) = 31 + 3.906 = 34.906 = 35

• Base Charges = 20, 5E = 0.2105


• Base Power = 102, 5E = 3


• Base Med Use= 35, 5E = -0.15625

CONCLUSION:

With this information, one should be able to predict just how powerful of an item could be made prior to construction and, if a given power is desired, the quality of the subcomponents required could be determined ahead of time. More importantly to me, however, the base values for each item helps me to see how each item breaks down in comparison with the others and how experimentation effects each one.

Another thing that becomes apparent with these tests (and is something that I was also hoping to show by doing this) is that the quality of the subcomponents used becomes less important as you craft higher level meds while the quality of the resources used (which determine your maximum experimentation percentage) becomes more important.

For example, the Small Stimpack - B that I made received 204 points of power from subcomponents (and I used a rather poor LS at 129 - this number could have been much higher - probably closer to 250) while it received only 80 points from experimentation (and I experimented up to 89% effectiveness). Of course, both aspects are important if you wish to make a truly effective item, but with the Small Stimpack - B, the subcomponents provide over two times as much power and the experimentation.

Now, compare that to the Small Stimpack - C that I crafted. That stimpack also gained 204 points of power from subcomponents but it gained 192 points of power from experimentation. At that point, the subcomponents used are only marginally more important than the resources used. Going on to a Small Stimpack - D (not shown here as I'm not yet complete with my calculations), I found that, while I gained 204 points from subcomponents, I gained 347 points from experimentation. As you can see, the trend continues - while the power gained from subcomponents is static, the power gained from experimentation varies from item to item.

REQUEST FOR COMMENTS:

Here is where I ask all of you for a little help. I'd like to validate my theories here and I can only do so much myself. What I would request is that, next time you craft one of the above items, insert the values you get into my formulas and see if they hold true.

For example, if you create a Small Stimpack - B with power 405, insert that value, as well as the power of the subcomponents used and your experimental effectiveness into this formula:

405 = 77 + Z + (Eff * 1.25)

I've done the work of getting you the base value and the5E value. Now I just ask that you use them a couple times and let me know how the results turn out. Does the formula hold true?

Any other questions or comments about my assumptions and/or methodologies are greatly appreciated.

Message Edited by mcglonec on 02-11-2004 09:06 AM

---

Ledao wrote:

One additional thing you may wish to include (though only of peripheral relevance to your query) -- the stats on subcomponents are taken to at least 1 decimal place. I haven't checked the rounding to see if it functions as one would expect it to, but I'm quite sure that a 24 charge aBEC at 95% experimentation will make slightly different meds than a similar 24 charge aBEC with 98% experimentation (netting one more charge at combine on a stim E, with all other resources and comps being the same). It does, at any rate, seem to function predictably -- the actual rating of the subcomponent is given by multiplying the theoretical max for that comp by the experimentation percentage attained on the schematic.

---

Very nice post.

One additional thing you may wish to include (though only of peripheral relevance to your query) -- the stats on subcomponents are taken to at least 1 decimal place. I haven't checked the rounding to see if it functions as one would expect it to, but I'm quite sure that a 24 charge aBEC at 95% experimentation will make slightly different meds than a similar 24 charge aBEC with 98% experimentation (netting one more charge at combine on a stim E, with all other resources and comps being the same). It does, at any rate, seem to function predictably -- the actual rating of the subcomponent is given by multiplying the theoretical max for that comp by the experimentation percentage attained on the schematic.

Any other questions and/or comments? Or should I take the silence to mean that you all agree and that my logic is correct?

Sorry ... its just that you are blowing my mind!

NICE WORK...

gotta print it off and re-read it.

VERY nice job - hope to get you some feedback

---

Blue-Quick wrote:

Using many experimentation pointmust give some different results in some cases though. I put 6 exp points into effectivness of a Health Enhance D. Got a great result. THen 2 more. Then 2 more. All great results. I ended up with power 880.

Using the same components, I got 3 great results. I ended up with power 925. I have had wonderful results like this with great expermintation rolls like 4 times out of 50. Has this happened to anyone else?

There's got to be something random in the experimentation part of crafting--especially if you use 5 or 7 points at the same time.

Blue-Quick

---

I believe the reason for this isn't to do so much with the crafting system in general, but enhancement packs, in particular.

You see, with enhancement packs, when you experiment on effectiveness, you're really experimenting on two things at once: power and duration. Instead of "rolling the dice" just once for both attributes, the crafting system really rolls the dice twice, once for each attribute. It then averages the two results and tells you the "average" success. Therefore, it might be possible to get a great success on duration and a good success on power and the system will tell you that you gota great success. However, while your duration will go up enough to be consistent with a great success, your power will only increase by an amount consistent with a good success.

You can see even more dramatic results when you fail on one attribute and get a great or amazing success with the other. You'll be informed that you had a good success or success and one attribute will go up while the other will go down. Or perhaps you'll be told that you had a failure and notice that one attribute still increases.

Enhancement packs are just weird that way.

Message Edited by Happymob on 02-23-2004 08:41 AM

Yes there are base values.

Whenever Devs or CRS "summon" items (for replacement of llost items etc.) they will be created with base stats.

Base stats include base stats for subcomponents (non-advanced), which is usually 1. They also are considered the worst combine roll without actually failing the combine, and do not have any experimentation or + crafting skill modifiers applied.

I haven't seen any for a while, so I can't list stats. But considering we almost never use sub 100 materials (in all stats) to make meds, and we get al lsorts of crafting bonuses (even at Novice Medic).. Just how truely scare-bad the stats on these things are is crazy.

You can't actually create a base item if you have any + crafting skills (though you can get close at very low level medic). But, sicne you need + crafting pts to get new schematics (up medic and doc craft trees), you are scewing your results on higher meds because of your + skills (your assumptions about base stats will be too high, but will accurately reflect what a real doc would actually get, because all docs have crafting skills too.)

-T

I believe experimentation is changing next patch though. Let's hope it's for the best.

So, I had: 36 @ 27% and 334 @ 28%. Then at the end I had 40 @ 48% and 413 @ 91%. My values for delta were: charges 5e = 0.190, the power was 1.254. Comparing this to your tests, I noticed that my power was the same as what you stated but charges was completely off. I decided to investigate this further, this time concentrating on charges instead of power. Using the same materials and starting at 36 @ 27% and 334@ 28%, I finished with 46 @ 90% and 361 @ 49%. Charges 5e: 0.159 and power: 1.29. Both values completely messed up.

I decided to go one step further and change the Biological Effect Controller to 19 power, 18 charge. I had 37 @ 27%, 335 @ 28% (I just realised that this is an error too..surely I'd get 334 with this, seeing as the power on my last BEC was the same as this one). So, at the end I had 40 @ 48% and 414 @ 91 %. Values were for 5e, charges = 0.143 and power = 1.254. These were the same values as my first test, and seem to go with yours.

My theory on this is that the values of the BECs are shown as integers whereas on the server they are stored as their accurate value.

Once again mcglonec, thanks so much for the info. Its actually really helped me understand the medicines more, and with it I'm going to work on a calculator much like the Architects had