Finder decay is returned in claims

kingofaces

Stalker
Joined
Jun 10, 2013
Posts
1,522
Location
US
Avatar Name
Tony KingofAces Hans
Background

So a few years back I did some testing between different decay finders (F-211 (L) and Ziplex Z15 (L)) with a 1.25 PEC difference per drop to see if finder decay was returned in loot or not. This was started because there were claims that finder decay was returned in claims, but there was never any hard data to show it even though it's not that difficult to test. These finders were selected for having nearly the exact same stats for depth, radius, etc. where only decay varied. At that time, there was no significant difference in claim sizes.

However, the recent changes to enhancers even after 2.0 got me thinking, and since I'm a scientist, I like to formally test things in the game. We now get the TT value back of the enhancer, meaning if finder decay wasn't returned, UL finders with enhancers would have a significant advantage if nothing had changed. That was my assumption before enhancer changes, but I wanted to test this again since the new enhancer mechanics would help make decay-based differences more apparent using enhancers on a F-106 to control depth without additional decay.

New testing

So through Dec 2019-Jan 2020 I repeated this test with a larger decay difference: F106, 1.799 PEC decay with 7 depth enhancers and TerraMaster 6 (L), 4.372 PEC decay. If there was a difference, it should ballpark around 2.57 PEC difference in TT return / drop. Depth (885m) radius, etc. were also identical to control for any unknown confounding.

What I would do is alternate between F-106 and Terramaster 6 each time I hit an enmatter. I also dropped for ores and recorded those, but focused on enmatter since differences should be more apparent there with smaller claim size. Total TT of each claim was recorded, and repeated. This was unamped, so I also removed outliers (generally anything above size 7). That alternating was important since that's basically the randomization of treatments to a degree. That prevents things like low TT in a certain area affecting the results if I had happened to use just one of the finders for say 30 drops and then switched to the other one when TT would have returned to normal.

For those not familiar with the statistical tests done on this type of data check out this round of testing a bunch of us did on hit rates. Those threads also get into things to account for that could be unintended confounding for comparing averages. Basically, there are ways to determine if differences in averages are due to random chance or some actual underlying cause if performed correctly. Basically, these are all averages, standard error of the mean (SEM), and t-tests where if the t-tests p-value is < 0.05, we're confident those differences are not due to just random chance. No fancy graphs this time since it just boils down to a few numbers.

So like the testing I did a few years ago, I kept dropping and kept watching for differences to be detected. This time around, I actually did detect a difference for enmatter that was very definite at 160 enmatter claims:

Enmatter: Average F-106 claim size for enmatter was 1.3434 PED (SEM:0.0254) and average TerraMaster 6 claim size was 1.4415 PED (SEM:0.0262). The difference between those averages was statistically significant (p-value = 0.008).

Ore: Average F-106 claim size for enmatter was 2.7111 PED (SEM:0.6363) and average TerraMaster 6 claim size was 2.7865 PED (SEM:0.6232).The difference between those averages was not statistically significant (p-value = 0.311).

What was different this time is that the p-values for enmatter a few years ago had no trend towards being a significant difference as I increased sample size. This time, I could technically detect this difference at close to 100 claims, but I kept dropping to confirm. There was a slight similar trend in ore (absolutely no trend in the previous round), so I'm guessing there is a difference for ore too. However, since the decay effect is already minimal for enmatter, but ore has higher variation due to larger claim sizes, this could swamp out the decay effect until you have closer to say 1000 claims until the statistical test can detect it.

Why does this matter?

So it looks like something may have changed in mining mechanics in the last few years with finder decay, or I just wasn't able to detect it years ago with smaller finders. Either way, more finder decay does increase claim size.

Now this isn't going to matter for trying to increase your claim size much at all as a form of amping. Where it really matters though is there's not a penalty for using finders like the TerraMasters with higher decay from an "eco" perspective. Using enhancers on an UL finder to get to a certain depth also costs more than L finders in many cases with this in mind.

If I'm using that F-106 I would pay about 1.73 PEC per drop in MU (i.e., just the MU on the enhancers) based on my historical break rates and 190% MU. For the TerraMaster 6, it's about 0.67 MU PEC per drop. You'll have to pencil this out for each L finder vs enhancing on an UL, but generally L finders under 120% MU seem to be a better deal. Enhancers still have plenty of use if you want to push depth down temporarily for a little bit or if you don't have the skills for a deeper L finder.

I also can't pin down exactly how much of that decay is returned in claims from this, but it looks like most of it is, while some goes to skills etc. That means eco for finders isn't avoiding TT decay like it used to be and using enhancers to reach a specific depth, but simply comparing L finder MU or enhancer MU.

Update 6/2020:
Since the last VU, enhancer break rates have changed to break quicker based on how much added effect they give. That means my calculations above for whether L or UL + enhancers is more eco are out of date. Not sure what direction that changes things though. Decay was not affected based on the VU notes though, so the core findings still remain.
 
Last edited:
I posted about this in like 2008, or earlier.
 
I posted about this in like 2008, or earlier.

Part of the history on this is that it had been claimed by quite a few people, but repeated testing since around 2015-16 could never actually replicate that finder decay was returned. I haven't seen anyone provide any hard data or anything that was testable with formal statistics until I did the previous round of testing that showed no difference (and should have had some marginal significance if finder decay mechanics were the same back then).

For that time, the data completely contradicted the idea that finder decay was included, so it's went from decay being mostly a sort of MU to account for to now mostly not being an added cost.
 
Last edited:
I checked some of my screenshots and it was from early 2009. I was in a mining competition and you had to find claims of various sizes of ores/enmatters, one of the sizes being "tiny" I think. Anyway, this size claim was very hard or impossible to find until you used a finder with very low decay.
 
Regardless of who did it first it needed re-doing after 2.0 to confirm again - thanks.
 
I checked some of my screenshots and it was from early 2009. I was in a mining competition and you had to find claims of various sizes of ores/enmatters, one of the sizes being "tiny" I think. Anyway, this size claim was very hard or impossible to find until you used a finder with very low decay.

That actually had come up in past discussions that ended up spurring this since no one had tested it yet. Usually, switching to other finders means you might be dealing with different stats that might indirectly affect claim size. Just the timing of when you were doing drops could mean you had a higher or lower TT when you happen to use a certain finder as well. That could have easily been the cause for what you saw in your screenshots without any good way to differentiate that. That's basically why this was started.

That's one set of confounding factors that's controlled for above, but a big problem in this game is people thinking they see a trend, and it's just randomness. It's really easy for people to make claims that are just allegory, so that's why I was always skeptical when the evidence for this one just came down to people saying they only noticed tiny claims when they happened to be using a low decay finder. It's a broken clock can still be right twice a day kind of idea that we needed good data to get figured out.
 
Last edited:
Makes sense to me and I agree that decay is returned back. I hope that MA should announce such things or changes. I dont get it why it is a "secret" to let some know and others no.. So many things that hsould be known for public is not known. No wonder newbies find it hard to stay.

Thanks for confirming this.
 
Thanks for statistically confirming. As you know, I have always thought this but I had no concrete proof. As Xen said, i was in that competition and trying to find that damn tiny, and i have to switch to a lower decay finder to get one. But like you said that wasn't proof, and it was a long time ago, MA like to change things.

But either way, its nice to see the maths for it now.

Thanks for your efforts.

Rgds

Ace
 
This is good to know. If it is undeniably true. It will incentivise using terramaster finders, with them being higher decay and all ?
 
The question is now how much skills affect findings?
 
The question is now how much skills affect findings?

Only as much as is needed to use finders to reach their maximum depth. Finder radius isn’t affected by learning period, etc., so only resource composition is really affected.

There’s no data out there to suggest that skills alone have any bearing on finds.
 
Did we get Mining 2.0?
I think I’ve mentioned it before, but hunting 2.0 basically made hunting mechanics more like mining in terms of calculating base TT return. There wasn’t any need to change mining in that “2.0” sense.

Some of the “wave” mechanic has become more prominent lately in certain areas, but I don’t know if I’d call that 2.0 since that’s just been a slowly changing thing over the years.
 
However, since the decay effect is already minimal for enmatter, but ore has higher variation due to larger claim sizes, this could swamp out the decay effect until you have closer to say 1000 claims until the statistical test can detect it.

Did you ever continue the ore test to see if there is statistically significant difference in size of claims?
 
Did you ever continue the ore test to see if there is statistically significant difference in size of claims?

I didn't mostly for the reasons I outlined in the OP. It would be very odd for MA to have an entirely different mechanic for ore than for enmatter, and I was getting some marginal effects with ore over time.
 
I didn't mostly for the reasons I outlined in the OP. It would be very odd for MA to have an entirely different mechanic for ore than for enmatter, and I was getting some marginal effects with ore over time.

It makes sense. I also think it is logical that the finder decay should be mostly returned, otherwise someone mining unamped enmatter with higher-end TerraMaster would be at a 8-10% TT disadvantage (4-5 PEC finder decay) over an ore miner using F-10x with big amps.
 
I wouldn't bet my money on this.
 
I might've completely missed that but does the same conclusion holds for extractors?
 
It makes sense. I also think it is logical that the finder decay should be mostly returned, otherwise someone mining unamped enmatter with higher-end TerraMaster would be at a 8-10% TT disadvantage (4-5 PEC finder decay) over an ore miner using F-10x with big amps.

Is about 6% and is exactly like that. Terra can hit dunkle, F-10x with big amps can hit sh*t. If finders' decay would be returned in a meaningful measure it would mean depth and range for free, doesn't exist such thing with the two most important mining modifiers. All respect king, but your hypothesis and method are biased, you can't possibly control the environment and assuming this or that only works as an intellectual exercise, not for a material conclusion. Preamped finders have any consistency only if decay of finders is a cost.


I might've completely missed that but does the same conclusion holds for extractors?

No. Neither refiners. That's why people selling lyst at 101% are sorrow imbeciles.

Hell, not even armour is payed back totally, why does everybody arround is running this hypothesis that MA is suddenly the f'ing tooth fairy? (This is just an attempt at humour, apologies if you find the tone displeasing :) ). I see your account is rather new. Safest bet, until you get your own data/practices etc is to use tt excavator and, respectively, imperium b1 refiner and be selective what you refine. Some things under certain markup rather belong in tt (or if you have the patience and auction slots, sell small stacks overpriced).
 
Last edited:
@Kerham did you even read the first post of this thread? I only see claims from you that the OP is wrong but no data to back it up.
 
@Kerham did you even read the first post of this thread? I only see claims from you that the OP is wrong but no data to back it up.

Indeed, their comments really did just amount to handwaving not a addressing what was actually done, and I laid that all out pretty clearly in the OP.

I mentioned it in some of the other testing threads I linked, but the whole point of using basic experimental design and statistics like we do in real scientific research is to be a step above all the pet theories out there. Some people mean well, but some "data" in there falls into the garbage in, garbage out mantra when it comes to potential for using it to interpret something. Others have trouble when there's solid data that contradicts how they think the game works, but that's exactly where data like this plays a defining role. That's how us scientists deal with enthusiastic people in controversial real-world topics afterall.

For this set, this is exactly the kind of data these tests are meant for. You control what you can (in this case, the finder stats), and treat everything else in the background as random. Those tests actually need that random effect and specifically test whether any difference found are due to sheer randomness (p > 0.05 in a nutshell), or correlated with the treatment effect (p < 0.05). That the p-value was much lower than that was also a good indication we aren't dealing with a marginal case. This is honestly the kind of thing I would teach to undergraduate students in an intro to statistics course early on, and isn't anything particularly high-level at all when it comes to scientific research and basic experimental design. This is not a hard one to setup a proper experimental design for.

As I mentioned earlier, I was one of those that was convinced of the opposite before I saw any data, and a good chunk of my living is tearing apart datasets like this to see if there's a reason not to believe what the data is showing. I let the data do the talking once I had something solid to work with though. The first round of testing was too low of a decay difference to detect anything if it did exist (similar to how ore worked in this later round), but even stress testing this larger decay dataset (checking for stratification, normality, etc. that would require a few extra steps) still yielded the same significant difference.

In the end, if someone doesn't "believe" the data, they are free to repeat it exactly as I did and run the same tests. Until then, I don't see anyone with a better designed dataset adressing all the things I did in the OP able to show otherwise.
 
Last edited:
I might've completely missed that but does the same conclusion holds for extractors?

The short answer is no, because everything up to the point I described is before a claim's TT value is determined. Extractors (and refiners) are extra costs after you have already been awarded the TT value, so it's like those two are essentially MU instead. It's kind like how if someone spends TT on fuel for a vehicle. Extractors at least are somewhat equivalent (though not exactly the same comparison) to the looter professions in hunting. Both determine how much of the net TT from a loot event you get. Those similarities diverge a bit as you get into each system (or what is known about hunting looters at least).
 
This certainly makes sense to me and why I have never been willing to pay the price premium for Omegaton Detectonator finders.
 
This certainly makes sense to me and why I have never been willing to pay the price premium for Omegaton Detectonator finders.

That's exactly why I sold off my F-106 after I did this testing. UL can work in narrow instances, but enhancers have somewhat limited utility at current break rates unless you want depth and can't use an appropriate L finder (low MU ones at least).
 
@Kerham did you even read the first post of this thread? I only see claims from you that the OP is wrong but no data to back it up.

Yeah, did you?

The fundamental problems which I see;

1. mining claims are not made entirely of shrapnel (or, at least oil in our case), such as to have a significant granularity of data. Is enough to have something like growth molecules or solis in those finds to throw off everything at 160 claims.
2. the concluded difference is larger than the actual decay difference (3.3 pec vs 2.6 pec), which would mean that you're actually making money tt wise using Terra6 on enmatter, which allow me to say is a very daring conclusion. Not impossible, I would love to see somehow skills being rewarded even if in a very modest form, but I would like to see ALOT more proof before throwing my wallet on this.
3. there is no reason for which the decay difference to be more apparent in enmatter than ores., the decay spent is exactly the same.
4. I honestly haven't understood wether it was double dropping or single-dropping, that would play a major role.
5. as I said above, the medium is impossible to be controlled. We just don't know how claims are generated, how exactly is this circuit of payback mining drop-claim designed, wether is always instant or exist some form of compensation one claim to another (as in one claim of 6 peds of valurite being impossible to be taxed, for example) etc etc Simply by not knowing these kind of factors, the amount of biases/skewing elements can be in any quantity.

Then the actual hypothesis is a bias:

1. that UL finder enhanced shouldn't (for some obscure reason) be more eco than L. That's exactly what's happening in hunting, and to a very obvious degree. I don't see why that would be some sort of heresy in mining. (and no, I don't own a f-106 tiered or not, just a poor old md40)
2. the whole premise itself is a made-up bollocks alltogether. Why not testing (and I am speaking very seriously) wether wearing yellow shirt instead of green is somehow compensated in finds? Is a plain example of wishful thinking, just pushed because reasons. Why would the finder decay be repaid, to begin with, what's the logic behind this?

Then, as per basic debate rules, the burden of proof is with the affirmative part. Which is not (and cannot) be the case here. There is a fundamental difference, in terms of feasability, between measuring the probability of an event (hitrate and refresh time, good work there) and the quality of an event, where the whole underlying mechanism of said quality is godamn unknown. Whereas the mechanism of probability is, well, obvious, the claim is there or godam not. Cautioning this attempt with previous testing done on a different species is poor practice.

Finally, is safer for me to go with the hypothesis that it's not repaid. From a cost perspective is a very serious problem, Terra6 unamped on ores means a 4% (at 0 MU) cost, when the damn mu is so hard to find, so it must be used with caution.

Clear enough my answer, Mr 6k mining finds on tracker? Just do it yourself, do it like it would matter, because hey this is about real money we're talking about, and maybe you'd understand why such a daring and hazarded hypothesis, with serious economic implications, need to be godam seriously backed before being accepted. Speaking of data, this time some hundreds of thousands of tt circulated mining when I bothered with tracking, my conclusion is that even 1% matters alot.
 
Last edited:
That's exactly why I sold off my F-106 after I did this testing. UL can work in narrow instances, but enhancers have somewhat limited utility at current break rates unless you want depth and can't use an appropriate L finder (low MU ones at least).
The good thing about mining skilling is you get to a good depth pretty fast. Terras may have high TT but with that MU, they are definitely your friend.
 
Yeah, did you?

The fundamental problems which I see;

1. mining claims are not made entirely of shrapnel (or, at least oil in our case), such as to have a significant granularity of data. Is enough to have something like growth molecules or solis in those finds to throw off everything at 160 claims.
2. the concluded difference is larger than the actual decay difference (3.3 pec vs 2.6 pec), which would mean that you're actually making money tt wise using Terra6 on enmatter, which allow me to say is a very daring conclusion. Not impossible, I would love to see somehow skills being rewarded even if in a very modest form, but I would like to see ALOT more proof before throwing my wallet on this.
3. there is no reason for which the decay difference to be more apparent in enmatter than ores., the decay spent is exactly the same.
4. I honestly haven't understood wether it was double dropping or single-dropping, that would play a major role.
5. as I said above, the medium is impossible to be controlled. We just don't know how claims are generated, how exactly is this circuit of payback mining drop-claim designed, wether is always instant or exist some form of compensation one claim to another (as in one claim of 6 peds of valurite being impossible to be taxed, for example) etc etc Simply by not knowing these kind of factors, the amount of biases/skewing elements can be in any quantity.

So I'll preface by saying you're not really addressing flaws in the data (or things that weren't already accounted for), but throwing out random ideas of your own. That's not how this works. I'm instead focusing on what the data says as a scientist with a reminder that I design these kinds of things for a living. I don't gain anything from posting about this one (and the testing I really do gain from generally stays private).

1. That's already accounted for in the study design. That's why you have paired tests between the two finders so in general you are hitting the same area an equal amount with both. If you're in an area with a lot of high TT value items (I did not have many of these, and nothing like valurite), that's an equal error across finders. If I mined an area like that too much, it would actually make it harder to detect a difference, not easier this way. If there was an issue related to this, that would have also shown up in the preliminary testing of the underlying data that's always done before running the test (e.g., testing for normality, homogeneity of variance, etc.).

2. There's no criticism of the study design here, though I noticed you did make a serious mistake. You're directly comparing raw averages, which you cannot do like that. That's why the statistical tests are required or some sort of 95% confidence interval (related to the SEMs provided). They use a range of values of where that average will likely fall based on the data. If you repeat what I did 100 times now, the averages are going to be slightly different each time. What the tests that account for that show is that the averages are too far apart, making it extremely unlikely that the true underlying averages are instead equal between the two groups.

3. As others' mentioned, please read the OP. You may believe there is no reason, but the data very clearly says otherwise. Standard deviations, SEM's etc. are always higher for ore than enmatter (SEM: 0.0254 vs. SEM: 0.6363 in this case). This is a very basic fundamental concept again in stats that as you move to larger numbers, variance increases. That's why we use variance stabilizing methods if we need to compare treatments that have a wide range in values (e.g., trt 1 mean = 3, trt 2 = 100) because you get a fanning out effect you see here as you increase in values (TT size in this case due to increased drop cost).

4. " that would play a major role" is an assumption not based in any data. Double drops did occur, though the alternations between finders were based solely on enmatter. I have not seen any data, including my own testing, that shows that claim sizes somehow vary due to what else you're dropping for. Please provide the data for your claim with that in mind.

5. Again, this is an assumption you are asserting, and why I mentioned handwaving either because it completely ignores basic study design. You don't need to know any of this because that is already accounted for in the random portion of the data where any nuisance effects are lumped together and occur in both treatments. You're just left with what you actually controlled for, finder in this case. Otherwise, you could make this exact same claim about any scientific paper out there where we are taking a small part of an unknown blackbox system and controlling for what we can while letting the randomness just do it's thing with the rest.

"that UL finder enhanced shouldn't (for some obscure reason) be more eco than L " - assumption without data

"the whole premise itself is a made-up bollocks alltogether " - Really? This is a question that comes pretty often if decay is returned in claims because it is an actual known input. This is simply how you would test it no different than if I was testing to see if there was a difference in any other set of treatment groups that have some random variation around an average. Sure you could test for any other random thing you make up like clothing, but you would be controlling for clothing rather that finder instead. In this case, clothing did not change. You're making it clear these data contradict your perceptions about the game, but that's not a reason to be lashing out like you are currently.

Overall, it's looking like you have some misunderstandings about how scientific research is conducted and basic statistics, so if this is a topic that interests you, I would suggest taking any beginner/introduction course to statistics. That's especially apparent with the comment " There is a fundamental difference, in terms of feasibility, between measuring the probability of an event (hitrate and refresh time, good work there) and the quality of an event, where the whole underlying mechanism of said quality is godamn unknown."

What you call "quality" is what we actually call a quantitative variable, which should be covered in day 1 of class along with things like normal distributions in a population. It's how we measure everything from average height, weight, etc. of a population when the underlying mechanism of how those individuals got to those numbers is " godamn unknown" as you put it. We don't care about all the underlying mechanisms when testing the weights of two different mouse groups fed different diet. We simply get an appropriate sample size from each group controlling for food type and measuring weight. They're treated the same outside of that, and any mystery mechanisms that somehow have a bearing in some mice but not others that could act as confounding are taken care of by the randomization. The statistical tests are meant for exactly that kind of situation here.
 
Last edited:
3. there is no reason for which the decay difference to be more apparent in enmatter than ores., the decay spent is exactly the same.
Sometimes it just takes one fairly short point to sort out those who may understand stuff from those who don't. Here, while the decay is said to be the same, the rest of the drop is more expensive for ores. Fluctuations in tt returns are thus more likely to cloud ore results than enmatter, which changes how much testing you need to do to get a high enough sample size to counteract this to an acceptable degree.
 
That's not how this works. I'm instead focusing on what the data says as a scientist with a reminder that I design these kinds of things for a living.

I understood already that you're a scientist. I am not, my education was law school.

In terms of cost, my observation is that finder+excavator+refiner (in this order of impact) tend to separate themselves as the obvious cost of mining (towards (but under) 3% over 155k) as opposed to probes+amps which trended, on this sample of mine, toward (but under) some 1%. Meaning, out of a total tt loss of about 3.2% including everything, two thirds towards three quarters were occupied by finder+excavator+refiner. Also meaning that over the course of the sample the decay cost remained rather constant, while probe+amp cost/loss decreased constantly.

Surely this could be alltogether just a coincidence where, on one hand I'd have a rather static decay cost, and on the other hand, a certain return rate which maybe over next 150k could see significant corrections and dispersion might be impacting at larger values, over larger samples, making kind of all my observations moot.

I do wholeheartly agree´with you on one point, my tone might have been inappropiate and you see it is as lashing out and personal. I apologies for that. I have no quarrels with you personally nor do I have any personal gain in the stated hypothesis, I am not this bigass merchant of either of the finder classes favoured by one or the other conclusion.

I am still not convinced by your explanations and you are right, this might be entirely due to my lack of education in domain. In my view we already know what causes variations in height or weight, namely genetic or biology. We can probe mices for said differences because we know how feeding/digestion works, in general. And that's where my valurite example was aiming. We do know for sure that 1. tax is tax 2. a stone of valurite has precisely 6 tt and an "average" of it will have precisely same value on a 5% tax, 2% tax or untaxed land. Hence it appears to me as a necessity to exist a sort of compensation mechanism, even if eventually on very short term, which to transfer certain values from one claim to another for who knows what necessities of the system. You might already have already covered for this with your approach and I am not able to understand that because, yes, I am not a scientist.

Finally,

Be it a dsec 30, decay 3.96 pec 130% MU
lvl 5 amp decay 200, 3% MU
(bollocks, was thinking at lvl3, so then there's only some 1%) - question still stands ref the impact of the answer

Dsec30 per 100 drops unamped, ores 2ped per drop tt, gives a finder cost of 3,96 ped, with a MU of 1.19 ped
Lvl5 amp (say with an UL finder) gives a MU cost per 100 drops ores of 6 peds.

Hence if decay of finder is returned in claims, it means that every 100 drops with unamped DSEC30, one would save 4,81 peds, which is an overall cost difference of 2.4%

If decay of finders is not returned, then we'd have to see what finder is used with L5, can be a difference of 0.5% or the amped finder can be more cheap, depends on the choice.

What this impacts, potentially, in mass:

- obviously bankroll
- blau, lyst, oil, narc, ganga, lvl5 bps, attachement skills (-)
- DSEC 30 clicks, maladrite, zoldenite, quantium, belkar, garcen, prospector/surveyor skills, tool tech skills (+)
similar reasoning, to some extent, for Roctec series

I am not a scrientist, as said above. I can only probe your test to the extent you saw above, "lashing". I must count on your superior understanding of statistics and provided data.

So, tell me, are you confident enough on your data such as I'd invest potentially thousands of $ in this paradigm? Or, what should be my position, go with something which I obviously don't understand or take the safest approach, which in this case means denying it?
 
Last edited:
I think i found this info elsewhere few years ago which is why i preferred f105 over f106. tiered f105 is damn near the terramasters in decay for lil extra :)
 
Back
Top