PC Lucky Looter

Boidster said:
Boidster: Doctor, it hurts when I do this <solves polynomial equation>

Doctor: Well, don't do that.

Gazz (I think) has said previously not to use CM to test out the RNG system. If I remember correctly, a few months back I used 'giveself' and it also showed unexpected results especially if quantity>1. The only way I got results consistent with the expected ranges was to load up on mats and run 20 workstations crafting weapons.

Here's part of the XML (items.xml) which governs SMG attributes:

<passive_effect name="EntityDamage" operation="perc_add" value="-.15,.15"/> <!-- random EntityDmg -->
<passive_effect name="EntityDamage" operation="perc_add" value=".1,.5" tier="2,6"/> <!-- tier bonus -->


If I'm reading it correctly, a T3 SMG has a damage modifier value calculated like this: [-0.15 to +0.15] + [0.1 to 0.5] + [0.1 to 0.5]

And a T4 would be: [-0.15 to +0.15] + [0.1 to 0.5] + [0.1 to 0.5] + [0.1 to 0.5]

So the range of modifiers is:

T3: from (-0.15 + .1 + .1) = +0.05 to (0.15 + .5 + .5) = +0.85

T4: from (-0.15 + .1 + .1 + .1 = +0.15 to (0.15 + .5 + .5 + .5) = +1.35

Putting that together with 9mm Ball ammo's base entity damage of 32 gives you the actual damage range you could see:

T3: from (32 x 1.05) = 34 to (32 x 1.85) = 59

T4: from (32 x 1.15) = 37 to (32 x 2.35) = 75

Quite a range and quite an overlap. Too much overlap, IMO.

Click to expand...
Very sure this is wrong.

The fist line adds a random value between -15% and +0.15% (because a range is given without qualifer).

The second line adds a fixed value dependant on tier qualifier. So tier3 adds 20% fixed and tier4 adds 30% fixed

I already tested this interpretation empirically a few weeks ago and it showed exactly the +-15% variation in each tier

 
Last edited by a moderator:
I've seen these kinds of entries in the xmls, but never knew for sure what they meant. So thank you both for the useful info!

If I'm following this then these would be results of those xml lines?

T3: from (-0.15 + .2) = +0.05 to (0.15 + .2) = +0.35

T4: from (-0.15 + .3) = +0.15 to (0.15 + .3) = +0.45

T5: from (-0.15 + .4) = +0.25 to (0.15 + .4) = +0.55

T6: from (-0.15 + .5) = +0.35 to (0.15 + .5) = +0.65

T3: from (32 x 1.05) = 34 to (32 x 1.35) = 43

T4: from (32 x 1.15) = 37 to (32 x 1.45) = 46

T5: from (32 x 1.25) = 40 to (32 x 1.55) = 50

T6: from (32 x 1.35) = 43 to (32 x 1.65) = 53

Off topic but understanding this a bit more will let me tweak something I've been wanting to for quite a while; up SMG damage compared to Pistol by a bit. Longer barrel adds around ~100 fps, assuming 'basic' ammo, so ~8%+. (if interested: http://www.ballisticsbytheinch.com/9luger.html)

 
meganoth said:
Very sure this is wrong.

The fist line adds a random value between -15% and +0.15% (because a range is given without qualifer).

The second line adds a fixed value dependant on tier qualifier. So tier3 adds 20% fixed and tier4 adds 30% fixed
Click to expand...
Hmm. We may have to agree to disagree, then, as I also tested this a few weeks back. I will find that thread and re-post the table of my results. Then we can discuss where I went wrong.

Regarding the first line: I am not sure we disagree here. It adds a random modifier from -0.15 to +0.15, as I said. Are you saying that each tier is applying this modification? So conceivably a T6 could have -0.15*6 = -0.90 as its base modifier? I did not see any evidence of this in my testing, but I am interested in your results.

Regarding the second line: I do not think this is correct. Every "value" parameter in the XMLs is a range, when given as "value=x,y". That is, the value is RNG between x and y. The "tier" parameter indicates which tiers (i.e. only tiers 2-6; T1 doesn't get this additional perc_add) that RNG modifier would apply to. I do not think that the game reads the "tier" parameter and divides the "value" range equally from .1 to .5 and assigns increasing values to each tier. At least it did not seem to do that in my previous tests.

Back later with my results from the earlier thread.

Previous experiment was for tool damage. I am going to redo the experiment using SMGs. Since the only way to reliably test this is to build SMGs at workstations, it will take me a little bit. Step 1: CM 50 workstations into existence...

 
Last edited by a moderator:
Boidster said:
Hmm. We may have to agree to disagree, then, as I also tested this a few weeks back. I will find that thread and re-post the table of my results. Then we can discuss where I went wrong.

Regarding the first line: I am not sure we disagree here. It adds a random modifier from -0.15 to +0.15, as I said. Are you saying that each tier is applying this modification? So conceivably a T6 could have -0.15*6 = -0.90 as its base modifier? I did not see any evidence of this in my testing, but I am interested in your results.

Regarding the second line: I do not thing this is correct. Every "value" parameter in the XMLs is a range, when given as "value=x,y". That is, the value is RNG between x and y. The "tier" parameter indicates which tiers (i.e. only tiers 2-6; T1 doesn't get this additional perc_add) that RNG modifier would apply to.

Back later with my results from the earlier thread.
Click to expand...


I cited the first line to show the complete "picture" and the difference to the second line. I'm only disagreeing about the second line.

 
Last edited by a moderator:
Boidster said:
Okay, got 100 T1 and 100 T2 SMGs cooking. Coming soon...statistics!
Click to expand...
um, unless you modded the Workbenches they only have 6 output slots? won't the other 4 just disappear or something?

Btw, definately see what you're saying about the xml lines. they really seem goofy. like there should be another modifier if the range is in set steps of .1, especially compared to the line above where it seems to be a 'full', random range _between_ -.15 & .15...

Looking forward to the results!

 
FileMachete said:
um, unless you modded the Workbenches they only have 6 output slots? won't the other 4 just disappear or something?

Btw, definately see what you're saying about the xml lines. they really seem goofy. like there should be another modifier if the range is in set steps of .1, especially compared to the line above where it seems to be a 'full', random range _between_ -.15 & .15...

Looking forward to the results!
Click to expand...
Nothing goofy about it, the step of .1 can be easily calculated (by the program) with stepsize= (max-min+1) / (# of steps)

 
meganoth said:
Nothing goofy about it, the step of .1 can be easily calculated (by the program) with stepsize= (max-min+1) / (# of steps)
Click to expand...
By goofy I meant the seeming lack of an 'operator' in the xml line telling the engine to "step".

I'm guessing that there are hidden options/modifiers that TFPs configure/write that get applied when the xml files are parsed.

Which are likely mainly for ease of xml writing/editting.. or maybe just so the syntax is more uniform?

 
Made a test right now: I removed the first line with the "random EntityDmg" from SMGs in items.xml and produced 12 SMGs of tier4. If I were right, all should have the same damage, as the random modifier is gone. If Boidster were right, there should still be randomness from the second line.

Result: All 12 SMGs had the same 41 damage.

 
Re: workstations - well my plan was to empty them out periodically, but instead I stopped at 25 units of each tier.

Conclusion: Meganoth is absolutely correct. And, in fact, reviewing my experiments with tools from a couple of months back I made the same mistake. Lernin' is hard.

Using Meganoth's correct interpretation, the expected min/max damage values are:


Tier


Min​

Modifier​


Max​

Modifier​


Min​

Dmg*​


Max​

Dmg*​


1


-0.15​


0.15


27


37


2


-0.05


0.25


30


40


3


0.05


0.35


34


43


4


0.15


0.45


37


46


5


0.25


0.55


40


50


6


0.35


0.65


43


53

* These are rounded

The results from 25 units of T1-T4 line up pretty well with those expected numbers.

image.png

FileMachete said:
I'm guessing that there are hidden options/modifiers that TFPs configure/write that get applied when the xml files are parsed.
Click to expand...
Agree. There is something behind the scenes looking for the "tier" attribute and, when found, it's used to calculate the # of divisions to apply to the "value" range, and then applies the value adjustments in equal increasing amounts for each tier.

 
Very cool info guys. It's really nice to have clarity on this.

Now we just need a 7 Days To Die Journal so this can be 'printed' as an "independently confirmed" "peer reviewed" article :biggrin1:

---

Go Team!

 
You must log in or register to reply here.
Back
Top