Great. I'm very interested in Punch UV rates. So I would like to see your actual data. Then we could integrate it with other people's data, to get a much bigger data set. You need a big data set, to tease out the rates with certainty. Your post doesn't give us confidence intervals or even the number of data. Please consider posting the data here or there. Thanks.

"I like roleplaying and my adopted personality for that blog is "The Spy"."

Gentlemen...

Again with the shifty data lol.

I understand that you are a "spy", but I really would like to know how you got your data etc. before I consider it to be legitimate... as would I for your lockbox odds. I'd love a mail from you regarding your source(s) just to verify, and can promise confidentiality if necessary.

Spy's overarching numbers fit the bill in regards to Punch.

While you must take it with a grain of salt without the statistics in front of you, we also must recognize that quite a few people here, Spy included, recognize that in order to obtain applicable data, there needs to be a ton of statistical data to use.

Truly, Spy's latest post (dice) is pretty much what we'd expect from OOO in regards to how things are obtained. In the sense that, for the most part, Punch is just a RNG you throw your weapon to and he cranks out a number series. There's a good deal of weight in regards to what strength your item gains, but, again, the numbers fit the bill.

Well yeah, but there are quite a few numbers that would fit the bill in this case and I'd much rather be using the correct set of numbers than whatever numbers "fit the bill" because I'm perfectly capable of making up ballpark numbers on my own. Or at least I am when I have the random.org site with me and some idea of the bounds.

I will tackle rarities once you start trusting my numbers :P.

Methods are just as important as the un-reduced data. Both are required to accept conclusions. Display your methods and data then I will trust (or disprove) your numbers.

I was told that shields were the LAST thing to punch. Cause a shields status doesn't effect your gear status, meaning it only works with your shield up.

If this is wrong please tell me I don't punch uvs that much anyways I was just told that when I was about to punch my shields.

Onewingedhorror: Yes, shield protections (damage and status) only have an effect when you're actively using your shield. To learn more, read the wiki page "Shield". Furthermore, you can't get shield status UVs. You can get only damage UVs. And research (mainly by Glacies) suggests that damage UVs have very little effect. This is why people tell you not to bother Punching your shields.

Klipik-Forum: You have a point, but my sense is that this game is supposed to be casual. We're not supposed to worry about the details of the rules. For example, when you hit a monster, the default is to show you vague ticks instead of concrete damage numbers.

@Spiral-Spy
Just to clarify the point: that's not how data works. If your data points to a different conclusion from mine then that just means that I need more data. Increasing the sample size will only increase accuracy. Even if (especially if) that changes the result I have currently.
The only exception being if there is a differing methodology which significantly impacts the results. Since we have no way of knowing if there is a modifier in the formula based on star level, heat level, the color of your knight, the number of snipes in Haven, etc, we have to assume there is not and everything is the same.
That educational note aside, I will respect your non-disclosure.
Instead of disclosing your data could you provide more accurate measurements to compare with my data? Perhaps instead of "~2" you could give one decimal point, "~2.x"? I would like to see if your measurements falls within my own margin of error.

@Spiral-Spy & Sir-Pandabear
If you are looking for a very specific UV (and got money to burn), you can increase your chances of getting it by blocking slots.
I wouldn't even bother doing this... You don't need a calculator to figure out that...
But if one did have a calculator, one would see this!
Goal: ASI VH on Sword, using SS's ratios and plan
Punch 1: 20k, Gremlin or something.
Punch 2: 75k, Undead or something.
Punch 3: 225k, 1 / (( 1 / 6 ) * 0.02 ) = 1:299 chance
20k + 75k + 299 * 225k = 67370kCr
Punch 1: 1 / (( 1 / 8 ) * 0.02 ) = 1:399 chance
20k * 399 = 7980kCr
The key here is that SS is right: it would take fewer Punches but cost significantly more.

--Cheshireccat

If the test server had a button that let you generate any amounts of crowns in one click, this would be true.

As it stands, it takes ten times more clicks to generate ten times as much crowns. You're generating money at a rate of 30k per item you buy in the supply depot, which takes n*k clicks, where n is the amount of items you're selling, and k is the amount of clicks it takes to buy the item from the supply depot, skip through the dialogue telling you that you bought an item at the supply depot, and then sell the item to a vendor.

As for shields, you have a 50% chance of getting a status UV for your armour (I think it was?), which has considerable effect no matter what the status should turn out to be.

On shields, you have a 100% chance of getting a defence UV that barely does anything, whether you have existing defence or not. I refer, again, to Glacies research linked earlier.

A 3 UV ticket costs over 11 times more crowns than single UV ticket does. This means 11 times more clicks for buying and selling.

This increases your chances to 1.5 times, if we are being generous.

So you are clicking 1100% more often in order to click 33% less. Assuming punching and selling stuff takes the same amount of clicks (and I'm pretty sure selling supply depot stuff is actually more cumbersome, so I'm being generous again here), taking SpiralSpy's advice involves 733% more work.

Edit:

Specifically using Cheshireccat's numbers, in order to save 100 clicks on punch (out of 300), you'd need to sell 1980 extra pieces of 5* equipment.

That's 266 sales to roll 300 times, and 2246 sales to roll 200 times.

and besides, with crafting you also have a chance to get more than 1UV, which is another bonus crafting gets. 1-2K is still quite generous, you could easily sell your crafts for a higher price... Crafting is more economical for 2* items.

I compared the same thing you did: The cost of obtaining a new single UV'd item.

You forgot to mention specifics in your results (correctly noted below in paranthesis):
Cost of punching a 1UV: 20k (plus cost of item that you apply the UV to)
Cost of crafting a UV: 35k (minus whatever you sell the additional 9 items that dont get a UV)
Your conclusions miss a significant aspect of the problem that do modify the problem itself.

Your revisit of the topic again contains flaws (not as bad as the first).

750crns for a 2* item is bare minimum - as that is the vendor price. The 1-2k was an already effective sell-price. Normally if you play the market right, the price can range between 3-4.5k per 2* item from the AH. Furthermore, anyone that shops around for a massive quantity of materials (Thoranhippo), you might notice that I'm currently selling them for roughly half the prices you estimate and you can possibly find them even cheaper.

If your argument for this topic leaves quite a bit of items unaccounted for, I question the validity of other items that you do not show data or methods. Luckily for me, I have been questioning the validity of your conclusions since the beginning. I am a skeptic raven and pecking all these keys sure makes my beak hurt.

As I wrote before, there is a lot of room to wiggle since you are dealing with many variables. There is no general answer to sir-pandabear's question. It all depends on what margins you assume/what risk you are willing to take.

That wasn't the response you gave until I gave my input. You clearly stated prior that there was a clear 75% increase in cost when crafting compared to punching.

I was able to give input and you handled the input well clarifying that there are a whole ton of factors going into the answer (which is my opinion as well). I am not able to give input on your methods/data on any of your other topics because you don't share them. I suspect there may also be incorrect conclusions as well, which is why you notice my first post in all of your threads often are clarifications for where my information differs or links to other similar research objectives.

Plus, I must say it's a nice looking site.

I'm mostly just asking because if it is anecdotal, it could easily do more harm than good as Spy has moderate reputability due to her model viewer and website.

If it is ripped from the game in some way, that's fine with myself and I don't blame her for keeping that bit to herself publicly... privately is a different matter however. Since I did not receive a mail, I can assume this is not the case.