Smuggler Archive

---

Syzygy-Gorath wrote:

The problem is, technically speaking, the advocate's right. Since we can safely assume that Wayne didn't use the same quality tools in all 4000+ slices, and since we don't know which tools were used for which slices, it's possible that lower quality tools produced lower quality slices while higher quality tools produced higher quality slices and that both the slices and the quality of tools used followed a normalized distribution. Now, I don't believe this to be the case because of experiments I conducted while I was still on Live (I wasn't a smuggler, but I provided UKs to a guildmate and we did a fair bit of testing) but it's possible. As far as it goes, the only way to proove whether or not tool quality actually matters is to make 200 top-quality clamps, 40 top-quality PLKs, 100 top-quality AUKs, 100 top-quality WUKs, and an equal number of crap tools, then 200 identical weapons and pieces of armor, slice all 400 items, and run two 2-tailed statistical analysis of the results. Even then you'll only be addressing tool quality, and slicing could be effected by, say, slicing while sitting on your left hand.

Now, like I said, I don't believe that tool quality affects slicing because of tests I did with them on Live. I saw no difference between 99.8% tools and 5% tools in terms of results. Both provided a normalized distribution of slices. But since I no longer have the results of those tests to point at, and since technically the number of tools using the godly copper wasn't high enough to be statistically significant, I'd have to do the work again. The sad thing is, if I had nice enough non-ferrous on TC I'd do it in a heartbeat. But since I don't have that kind of non-ferrous, I'll just shrug and say "sure, it may not be up to the scientific method, but it's enough data that the chances that it was predicated on tool quality is slim enough for me to accept as non-existant."

Update: I just bought 30000 units of 977CD copper. I'll be crafting AUKs, WUKs, clamps, and knives from it as well as a matching set from the crappiest non-ferrous I can find. I'll report back with all the data as soon as I can to finally put the tool debate to rest for even the most skeptical of forum-goers.

Message Edited by Syzygy-Gorath on 08-26-2004 07:51 AM

---

Also, this isn't something that the devs are promising, it's done, it's live, it's in the game. While I will have my doubts aboutthe validityof statements made concerninga release date of an update, I have no reason to doubt a statement concerning something already in existence.

---

jbezorg wrote:

LOL.... the devs must get a kick out of this. Here we are talking about 2-tailed statistical analysis and the result of slicing sitting on a hand when they're looking at the section of the code that calculates the result of a slice. This isn't statistical analysis of mass marketing trends on a population filled with unknown variables. We are dealing with an absolute.Someone just has to point to lines X through Y insection Z of the source code.

Also, this isn't something that the devs are promising, it's done, it's live, it's in the game. While I will have my doubts aboutthe validityof statements made concerninga release date of an update, I have no reason to doubt a statement concerning something already in existence.

---

---

Izhozz wrote:

Why do you waste time on experimenting with a random system? I dont think the Devs are lying about this,why would they? Just accept the fact that random means random.

---

---

Syzygy-Gorath wrote:

---

Izhozz wrote:

Why do you waste time on experimenting with a random system? I dont think the Devs are lying about this,why would they? Just accept the fact that random means random.

---

/sigh I know it's random. I expect the tests to bear that out. I believe GreenMarine. However, I would like to have cold, hard, indisputable data to point at because of the seemingly large number of people that insist that GreenMarine is mistaken or outright lying. That's all. I'm not trying to refute anything, just make the case that much stronger for what we all know to be true.

---

---

jbezorg wrote:

---

Syzygy-Gorath wrote:

---

Izhozz wrote:

Why do you waste time on experimenting with a random system? I dont think the Devs are lying about this,why would they? Just accept the fact that random means random.

---

/sigh I know it's random. I expect the tests to bear that out. I believe GreenMarine. However, I would like to have cold, hard, indisputable data to point at because of the seemingly large number of people that insist that GreenMarine is mistaken or outright lying. That's all. I'm not trying to refute anything, just make the case that much stronger for what we all know to be true.

---

Good luck my friend. You're a better man than I am Syzygy-Gorath.

---

---

stupid_people_happen . .
Pappi Inc Tailoring (home of the black tax) - Odi's meds and chef tissues - closed
- I support literacy, common sense, and apostrophes

Well, work's been killer lately, but I've managed to get the first half of the tests done. The results can be found here. In order for the tests to indicate that tool quality matters, P must be less than 0.05. As you can see from the right-most column, P is not less than 0.05. That means that the sample of slice results taken using 15% tool and that taken using 97.7% tools represent the same overal population, and that differences are the result of pure chance. In other words, for armor, tool quality does not make any difference. I'll get the statistics for weapons up as soon as I recover from pulling 40.5 hours of overtime over the past two weeks and, y'know, actually have a weekend to unwind during.

Message Edited by Syzygy-Gorath on 09-02-2004 07:45 AM

---

Scoooter wrote:

I still say that after taking LSD, while drinking brandy, eating ahrissa,smoking a new crack pipe while watching a trandosian dancer wearing hawtpants, listening to a white wookie musician while meditating (in hawtpants) you will get better slices.

---

No, scooter.... they just SEEM better! LMAO.