SIB Healing Tools Experiment

[DocKangey]

After some weird skilling behaviors I encountered with a Vivo, and the subsequent research about skilling here on the forums, I didn't find much data about skilling with healing tools (FAP, Hedoc, Vivo, etc.). I decided to start my own little experiment get some complete data on runs on different tools.

I will test several limited tools (with learning period and maxed ones) as well as several different unlimited tools. I will try to do at least 2 runs (completely use up a full tool) with each to be able to provide a larger sample of data.

Experiment Settings
With each tool I will do one run at Swamp Camp, preferrably at the same time of day. Skill gains obviously have very complex mechanics, so I will try to mimin a certain kind of behavior with each run.

Before and after every run I will scan my skills, and while I am on a run I will not use any other tools or weapons that have impact on any skills. About 50% of the time I will run around and heal people "in the field", a small part of it will be standing in the middle of a bunch of mobs and getting attacked without moving. The other 50% I will heal at or in the container, which means a lot of heals in a short time and hardly any movement.

Objective of the Experiment
Mainly this experiment was concieved to find out if a SIB healing tool does give more skills during the learning period than an unlimited or maxed healing tool. To measure the skill increase for each run, I will calculate the increase in TT value for Dexterity, Anatomy, Diagnosis and First Aid skills (defference between TT value before and after) Using the TT value makes the skill values actually compareable - thanks jdegre for the great scanner!

as a comparison factor I will calculate how much PEC TT-Value gain per heal I got from one tool per skill (PEC/heal) as well as how much PEC TT-Value gain per 1 PED decay (PEC/decay).

Columns

• Skill Amount of Skill Points gained
• TT TT Value of the Skill gained (PED)
• PEC/Heal Amount of Skill value gain in PEC per heal. Example: 0.5 PED TT gain; 1000 heals = 0.05 PEC/heal (0.05 PEC TT-value gain per heal)
• PEC/decay Amount of Skill value gain in PEC per 1 PED decay burned. Example: 0.5 PED TT gain; 50 PEDs tool decay = 1.0 PEC/decay (1.0 PEC TT-value gain per 1 PED decay)
• Lines Skill message lines (green messages)
• PEC/line Amount of Skill value gain in PEC per Skill message line

I chose to take the TT value of the tools to find out if decay and skill increase have any connection.

Results

#1: Vivo T20
(SIB Tool; Learning Period: Yes; Current Paramedic Level: 18)

VIVO T20    (1136 Heals, 5622 PEC decay)

            Skill   TT     PEC/heal   PEC/decay Lines PEC/line
--------------------------------------------------------------
Dexterity    1,87   0,04   0,003521   0,0711491     6 0,666667
Anatomy      8,32   0,18   0,014845   0,3201708    26 0,692308
Diagnosis   12,46   0,11   0,009683   0,1956599    28 0,392857
First Aid   15,93   0,50   0,044014   0,8893632    64 0,781250

#2: FAP-80
(unL Tool; No SIB; Current Paramedic Level: 18)

FAP-80      (1459 Heals, 8928 PEC decay)

            Skill     TT   PEC/heal   PEC/decay Lines PEC/line
--------------------------------------------------------------
Dexterity    0,93   0,02   0,001344   0,0224014     5 0,400000
Anatomy      6,49   0,14   0,009409   0,1568100    30 0,466666
Diagnosis    9,37   0,09   0,006048   0,1008065    31 0,290322
First Aid   12,62   0,36   0,024194   0,4032258    77 0,467532

#3: FAP-50
(unL Tool; No SIB; Current Paramedic Level: 18)

FAP-50      (1213 Heals, 4852 PEC decay)

            Skill     TT   PEC/heal   PEC/decay Lines PEC/line
--------------------------------------------------------------
Dexterity    0,64   0,01   0,000824   0,0206101     ? 0,000000
Anatomy      4,88   0,10   0,008244   0,2061006     ? 0,000000
Diagnosis    4,70   0,04   0,003298   0,0824402     ? 0,000000
First Aid    8,22   0,24   0,019786   0,4946414     ? 0,000000

#3: EMT Ek-2550
(unL Tool; No SIB; Current Paramedic Level: 18.16)

Ek-2550     (670 Heals, 6021 PEC decay)

            Skill     TT   PEC/heal   PEC/decay Lines PEC/line
--------------------------------------------------------------
Dexterity    0,15   0,01   0,001493   0,0166085     ? 0,000000
Anatomy      1,84   0,04   0,005970   0,0664341     ? 0,000000
Diagnosis    3,20   0,04   0,005970   0,0664341     ? 0,000000
First Aid    6,73   0,19   0,028358   0,3155622     ? 0,000000

What does that mean?

From the first quick glance it seems that the SIB tool in learning period DOES actually give more skills per PED decay.

If we compare the PEC/decay values, we can see for example, that for each PED decay on the Vivo T20 I got an average of 0.8893632 PEC value gain on my First Aid skills, compared to only 0.4032258 PEC value gain on the FAP-80.

It seems I would get about twice as much skills with the SIB tool compared to the unL tool. For the PEC/decay value I used the TT value of the Vivo because the "game mechanics" (if there is any relationshop between decay and skill increase) would take only TT value into consideration.

But even with the current markup of 170% on the Vivo T20 there is still a slight benefit for using the SIB tool for skilling.

I havent yet found a conclusive explanation for the amount of skill lines and their relationship to the actualy skill gain. Data so far suggests that there is indeed only a gain in skill points if you see a skill message, and if no message show up I haven't seen any gain at all. Also it seems that there is not a significant difference in the amount of skill lines, but SIB tools in learning period give more skill points per message - more data needs to be collected for proof.

The datapoints gathered from my FAP-50 run suggests that the main influence on skill increased is the decay spent - I am trying to confirm that with more data.

[DocKangey]

Run #1

My first run started with Paramedic Level 17.95
I chose the Vivo T20 because I had one available Right now I am within the learning period of that SIB tool, the speed of course is slow.

Stats:
1136 heals
53477 HP total, that makes an average of 47.075 HP/heal
6 Skill Lines Dexterity
26 Skill Lines Anatomy
28 Skill Lines Diagnosis
64 Skill Lines First Aid

            Points Before   Points After
----------------------------------------
Dexterity         3195.62        3197.49
Anatomy           3155.57        3163.89
Diagnosis         1488.00        1500.46
Fist Aid          3035.33        3051.26

Run #2

Second run was done with a FAP-80, because it has a decent heal amount and rate. I do an almost consistent 32.0 heals per use, some very rarely it drops below that.

Stats:
1459 heals
46276 HP total, that makes an average of 31.7177 HP/heal
5 Skill Lines Dexterity
30 Skill Lines Anatomy
31 Skill Lines Diagnosis
77 Skill Lines First Aid

            Points Before   Points After
----------------------------------------
Dexterity         3197.49        3198.42
Anatomy           3163.89        3170.38
Diagnosis         1500.46        1509.83
Fist Aid          3051.26        3063.88

[MiguelNykvist]

Nice going Doc K
Hope you find something real useful, so i know where to go when i need to skill my heal skills hehehe and ty for this, and all the heals at Swamp

[MG Mighty]

Nice study
Provides information and helps the newcommers.

[kygon]

I didnt do data but I'm 100% sure the gains on SIB vs no SIB is quite high!

[DocKangey]

I just finished getting the 2 initial data sets together (documented each dingle heal, HPs given, area, skill messages) and have done the first comparisons (see the main post)

From other SIB/noSIB threads I have the feeling that "skill increase bonus" is not that obvious and people interpret SIB very differently. For healing tools it seems that SIB during learning period gives a significant higher skill increase than using noSIB items.

My next tests will include maxed (L) items and a rerun of the Vivo T20 and FAP-80, to make sure my data is consistent.

[Doer]

Thanks a lot for doing that test and sharing. I look forward to seeing your next results. I think it would be particularly interesting to see results for "Not Anymore" after you max that SIB tool, and for a FAP with a substantially different heal/pec.

I've gotten the impression that SIB means a lot more for BPs and healing tools than it does for weapons. Your experiment here certainly supports the idea that it's very beneficial for healing tools, anyway.

I checked to see that your skill gains follow the proportion of their contribution to the profession. Not too surprisingly, they do, and actually strikingly well for the SIB test. It's good to see that.

		Norm.	SIB	No SIB	SIB	No SIB
	% Cont.	% Cont.	Volume	Volume	%	%
dex	3	4.9	0.04	0.02	4.8	3.3
anat	13	21.3	0.18	0.14	21.7	23.0
diag	8	13.1	0.11	0.09	13.3	14.8
fa	37	60.7	0.5	0.36	60.2	59.0
	61	100	0.83	0.61

[Doer]

I pulled out log entries comprising 606 PEDs decay on my Imp. FAP to see how it compares. I already know that skill gains in Paramedic profession are heavily weighted to PED decay, making an Imp. FAP an awful means to skilling. My hypothesis is that skill gains are a function of uses and decay (s=f(#,$)) and this is a good chance to investigate that.

I just looked at First Aid and Diagnosis gains, because most of my Anatomy and Dexterity come from combat. First of all, i checked to verify that they were gained in the expected proportions for these skills. I found a ratio of 0.216 Diagnosis to First Aid; 8/37 (their contributions to Paramedic) = 0.216, so the sanity check is passed.

Now, comparing the gains (in PEC volume)/PED decay with your two results:
Healing tool: Imp. T20 FAP-80
Diagnosis: 0.144 0.196 0.101
_First Aid: 0.665 0.889 0.403
-------------------------------
Combined: 0.809 1.085 0.504

The Imp. FAP falls between the SIB T20 and the FAP-80 here in skills/PED. More uses are possible per PED decay, so it does seem that there is some component of #uses (~hp healed) involved. With another data point or two for a traditional FAP we could maybe find the relative contribution of $ and # to skill gains, and then from there make better comparisons for SIB and no SIB.

[Immortal]

Originally Posted by Doer

I just looked at First Aid and Diagnosis gains, because most of my Anatomy and Dexterity come from combat. First of all, i checked to verify that they were gained in the expected proportions for these skills. I found a ratio of 0.216 Diagnosis to First Aid; 8/37 (their contributions to Paramedic) = 0.216, so the sanity check is passed.

Yes convenitently skillgains are %'d by their contributions in teh long run. Confirmed this using Star's data in his foe vs laser setup. (deconvoluting complex multivariate data sets is fun)

The Imp. FAP falls between the SIB T20 and the FAP-80 here in skills/PED. More uses are possible per PED decay, so it does seem that there is some component of #uses (~hp healed) involved. With another data point or two for a traditional FAP we could maybe find the relative contribution of $ and # to skill gains, and then from there make better comparisons for SIB and no SIB.

I think a point of note to make here is that the data collected in this thread was for healing only. I am assuming your data would have included hunting too?

PS you've managed to burn 606ped on an imp fap
PPS insert note here on 'skill speed of imp fap is 43.2% that of the t20' and 57% that of fap80.

I think there have been similar threads showing where skill gain is sacrificed slightly ..but the numbers are too small (i think) to be conclusive

PPS uhm.... insert note on behaviours of skill levelling such that skill_gain=derivative_x{exp(b.sin(ax)*C+f(x))+g(x) }*random_EU_component_function (don't know the approximation formula of my head :p)

[DocKangey]

My hypothesis is that skill gains are a function of uses and decay (s=f(#,$))
...
More uses are possible per PED decay, so it does seem that there is some component of #uses (~hp healed) involved.

I am not so sure about the involvement of #heals in that formula. The decay seems to be directly connected to the #heals ( d=#h*C ), so #heals would not be an independent input parameter to the function. Unless of course decay depends on the actual HPs given with each heal - but I have no evidence for that. Maybe another thing to investigate more closely.

From what I see in the data so far, there are 3 potential factors that could determin skillgain: #heals, $decay, HPs given. Since $decay presumably connected to #heals I think $decay, HPs or both are involved in skillgain as well as a tool specific factor. (sidenote: maybe there is a connection to a weapon skillgain model, $decay and damage determines skillgain)

I will get us some more data from a FAP-50 and maybe another FAP. Right now I have a gut feeling that the EMT's might have their own skill gain function, so I will stick with FAPs until we have a more solid hypothesis and then compare with L-tools.