UESP Forums

I found this site nearly a decade ago, but I did not have the time nor the technology to put a lot of effort into recording data for the game. For the moment, I have some interest and my technology has improved drastically.

I started by measuring the 'actual' travel distances between all of the settlements in the game. The data is saved to an Excel workbook.

Having read most of the pages that relate to the gameplay, I noticed that there were tables for the enchanted items in the game. Also, the tables were incomplete. I started by writing down in an Excel workbook the names, the materials, the attributes, the spells, the weight, and the cost of every weapon and piece of armour better than leather. I was trying to fill the gaps. When I started this, I was using my Windows XP to play the game in DOS-Box, while I used my Acer laptop with Windows 7 to record the data. It was not a quick process and it was not particularly ergonomic either. I recorded information for 3-4 settlements before I put the project on hold, for one reason or another.

Recently, I noticed that I could download and install Arena for free, so I installed it onto my HP laptop with Windows 10. With three screens and better software, I should be able to accomplish more in less time. However, it was still going to take time to record everything that I wanted.

I decided to change my strategy, after I rediscovered the lists of stores for each settlement. I manually selected, copied, and pasted the lists into a new Excel workbook. I did not care where the stores were located, I just needed the names. With nearly 2900 store names, I analysed the data using a PivotTable and a PivotChart. There were certain names that stood out, because they showed up so often.

When I looked again at the sorted entries in the PivotTable, I could see a pattern in the names. The leader was different, but the same 'types' of store kept showing up. I returned to the original column that I created and I add a new column for "Collections". According to the data, there are 10 separate collections.

Accouterments
Armaments
Equipment Store
Gear Store
Merchandise
Provisions
Sundries
Supply Store
Tool Store
Weaponry Store

During my analysis, I found a few spelling mistakes, but I would like to verify them in-game, before I make any changes.

Returning to my previous project of recording what each store sells, based on my discovery above, I suspect that the list of items will be the same for all the stores that belong to a collection. My current sample size is too small, but I believe that after recording the data for 10, 20, or 30 stores, in a few of the collections, that it will be possible to extrapolate a better and more complete list of items.

I remember grinding out the game nearly 30 years ago to locate or to buy all the ebony items. I think that is as far as I got with it at the time because the old 486/66 DX2 just did not have the hardware necessary to play it well, or I switched to a different game.

So, if I decide to actually play the game now, I would like to have a roadmap to find everything that I want in order to equip my character.

And, that is why I am questioning the accuracy of the lists of stores in settlements in Arena. If they are complete, then I have reduced my work effort immensely. Also, if my hypothesis is correct, it might be possible to locate items in more stores then the respective tables indicate.

Anyway, if someone happens to know the answer to my question about the lists of stores, I would appreciate it before I get too far along.

Thank you.

Most stores have fixed (unchanging) names, but stores that have an owner's name in their names-- such as "Tom's Store"-- can have random owners' names, although the rest of the stores' names will be fixed. For instance, a given store might be called "Tom's Used Goods" one day, but be called "John's Used Good's" another day.

In the UESPWiki pages, those stores are indicated using underlines or "blanks" for the owner's name, such as "___'s Used Goods." However, as far as I know the list of equipment stores that you get when you're talking to people on the street will always be the same, so I think it would make the most sense to use the names as given in those lists, with a note about the ones that can have random owners' names.

I am not so much worried about the names of the random stores. I figured that out by looking through lists and following how the naming convention was used.

I am looking at the items that are available between two stores that belong to the same collection. I just skimmed through the offerings of 'The Practical Equipment Store' and 'Elite Equipment Store' in Eagle Brook, High Rock. My original hope was that stores in the same collection would offer the exact same items. As it turns out, TPES has 22 items, while EES has 52 items. So, my first hypothesis seems to be busted.

However, because there are 22 'named' stores within the Equipment Store collection, perhaps the 21 'Bargain Equipment Store' entities have the same offering. Now, that would be a large group to verify, so it would be better to tackle something like 'The city-state Equipment Store' with 2 entities and 'The town Equipment Store' with 4 entities. If the hypothesis holds for the small values, then I can verify that it holds true with largest one, '<random>'s Equipment Store', which has 30 entities. The same hypothesis testing would need to be accomplished with the other collections.

If I end up disproving all of my hypotheses, then it becomes a true grind to extract the data from the game and sharing it here with the community.

With the amount of deterministic results that exist with this game, there cannot (or there should not) be nearly 3000 distinct lists that the different stores use to create their individual item offerings whenever a player visits them.

Does this help to clarify the part about accuracy that I was asking about earlier? Can the lists that were created be trusted?

I see what you mean, but I suspect the data is probably stored within numerous data files which are loaded as needed rather than kept within a single giant database that's in memory all the time, so there could well be a unique list for each store. If this is the case, it might be possible to extract the lists from the data files.

I had a similar sort of hope related to the houses in the cities, because I wanted to map out which house layouts are used for each of the houses. The cities are split up into what one might call city blocks, and there are a specific number of layouts for the various city blocks, including layouts which are mirror images or rotations of other layouts. I was hoping that a specific house within a specific layout might be the same regardless of which city it's in, so I traveled around looking for cities with a specific city block layout and checked a specific house within that city block, but the house had a different interior in each of the cities I checked, which blew my theory out of the water.

Interesting theory.

During my records creation process, I was also going to grab screen shots of each settlement in the game. I would need to grab many images for each one, because the entire settlement does not fit within the area available within the map view. Now that I have SnagIt, it is much easier to crop images in batches. The more difficult part is manually positioning each 'tile'. I made sure to have enough material to overlap the images, so that there are no seams. I have made a few of them, but I am not ready to continue with the rest of my idea.

Ultimately, I would like to have a set images for each settlement. I will probably assign some 'bright' colours to the inns, the stores, the temples, the mage guilds, and the palace. It will be assumed that everything else would be a house. Perhaps with enough effort, it might be possible to set up something like a radio button against the current lists. Depending on the user's selection, that version of the image would be displayed, so that you would know roughly which heading to take. It might be beyond the capabilities of the wiki, but it would be a 'nice-to-have'. I believe that something could be assembled into a dynamic PDF file, but that is beyond my current skill level. I might need to delve deeper into Adobe LiveCycle Designer to try out a few things.

Here is a different thought. If recording all the inventory lists for the different stores in the game is too big of a problem to solve, has anyone tried to tackle the Mage Guilds? Assuming that there is one and only one in each settlement, then the problem is 1/10 as difficult as the stores. There would be 256 Mage Guilds, instead of nearly 2900 stores. Also, they sell fewer items on average, so creating the lists should proceed much faster. It should be possible to create a complete inventory of magical items throughout Tamriel. I will probably start by confirming with my new game that the Mage Guild at Eagle Brook is still selling the same wares as a previous game that I started a few years ago.

I might try to tackle a few of them tonight.

It has been a productive few hours.

I confirmed that Eagle Brook always sells the same magic items, no matter the character that I play. Also, it seems that I was paying half the current amount that I am with my new character. I might need to retrieve the old save file and see what my character's attributes were. It could have been that I was visiting a Mages Guild on one of the special holidays. So, that was a happy coincidence that the values were so different.

After that part was taken care of, I decided to scrape the magic items web page and create a PivotTable to analyse the data. I wanted to know whether or not each settlement was listed. They are not. There are entire provinces that are missing from the list. Elsweyr, Morrowind, and Valenwood are completely missing. Only half of Summurset Isle was surveyed, while Black Marsh had only seven settlements listed. (There is a spelling mistake. 'Black Rose' should be 'Blackrose', like the other instances.) Thankfully, Hammerfell, High Rock, and Skyrim 'appear' to be complete, settlement-wise. I cannot attest to the accuracy of the inventory lists.

Tonight, I was able to complete Summurset Isle, by writing down every item that is sold by the missing 16 settlements. Depending on when I have more free time, I might try to tackle Black Marsh next. The other three provinces will take a bit longer.

It will take many passes through the map to gather all the data that I am setting out to record. By doing one thing (or collection) at a time, I do not run the risk of missing something during a visit to a settlement. Also, I try not to bite off too much during any single session.

Have you looked into the OpenTESArena project? I'm not very familiar with it, myself, but apparently they've done a good bit of digging into the data files to decipher them, so there might be some useful tidbits of information there.

I do not recall reading about the project before. If they have managed to read most or all of the game data, then why has not anyone brought it back here for the wiki to use?

I just finished with Elsweyr. I believe that I am able to complete a province every four hours, if I am only dealing with the Mages Guilds. Obviously, I could spend that amount of time just recording a handful of settlements in any province, when I decide to deal with weapons and the armour.

It could take me about a week to finish with the Mages Guilds and figure out how to parse my Excel table into raw text that the wiki can understand. That will be one task done. After that, I might tackle grabbing the screen shots of the settlements and stitching them together. At least I know where each Mages Guild is located. I am recording coordinates as I go. I can create a second copy with the Mages Guild filled with a different colour for locating them easier. Case in point, I got killed twice in Chasemoor trying to find the two inns when I arrived at night. Maybe if I had a colour-coded map, I might have been able to find them and save myself from the enemies that chased me down.

It is too bad that I did not have digitized notes of all this. I probably have lists of settlements that sell Ebony items, but not the rest. I know that I have done this before when I was kid.

I will be interested to see how the numbers break down for the magical items. There might be better versions based on the number of charges when the type and the spell are the same. The differences are still linear, but if you can get more charges from item X versus item Y, then would you travel throughout Tamriel to make such a purchase?

I'm not very familiar with the OpenTESArena project, myself, and honestly don't know what they have or haven't managed to decipher in the data files.

Well, I left a comment on the latest video that was uploaded to the YouTube channel that is showcasing the progress so far. According to the author, they "haven't gotten to store items yet." I might be able to complete my survey of all the magical items before they are ready. If my clear-text version of the data is helpful to the deciphering process, then it might make the process easier for the equipment stores.

When it comes to sharing information and knowledge, it is better to provide it to others, so that they can avoid reinventing the wheel or encountering the same frustrations that you did. I try to do the same thing when it comes to using Microsoft Office. A few tips and tricks, and you are actually using the applications like they were designed to be used, instead of struggling to force something that could be impossible to do.

I will keep chipping away at the data tables and performing a verification and validation of the existing data for the settlements.