Goblin Squad Member
|
...and I'm 53. Welcome to the club, you'll find patience and friendliness here, no worries; you chose a good game to get started with, and you'll be helping in its development.
Look over the Guild list in Land Rush, to see with whom you might be sympatico. Consulting (or joining) with others of like mind is a quick way to get started.
|
Goblin Squad Member
|
I'm talking about pulling the data down and putting it in a spreadsheet or a database. I know one of our resident numbers geeks is doing so with the intent to chart the daily growth of the different guilds, and I imagine Nightdrifter is doing something similar
Not quite. My coding skills are more analysis related. Anything I wrote to grab info from a webpage would be laughably crude at best (probably some combo of 'wget' plus 'grep' in a python loop).
What I've been working on is a simplistic empirical model to predict the minimum number of votes to get a settlement (bear in mind this completely ignores whatever modifiers Lee will be applying!):
*Take the total number of votes. Subtract off the number of votes for TEO, Pax Aeternum, and Seventh Veil as they aren't vying for settlements. Call this number V_T.
*Look at the top contenders who are well above everyone else and count how many there are (don't include the above 3). For example, Magistry and up gives 5. Call this number N.
*For those N settlements, total up their votes and subtract this total from V_T. Call this new number V.
(In essence V is just the total of all votes of all competing settlements who don't have a large lead on the rest of the pack. N is the total number of such settlements way out front, not including the 3 who already have one.)
*G = number of guilds - N - 3
*m = minimum number of votes to get a settlement
Then:
m ~ 1+[(V-G)/G]*log[G/(30-N)]
(The log is the natural logarithm, ie. base "e", NOT base 10.)
Right now this predicts m~2.59 and the 30th place competing settlement has 2. Obviously the prediction will change as time passes.
Where it comes from:
The first landrush had a vote distribution which sorta looked like an exponential (ie. a decay function). Dunno why. It just did. The lead settlements mess with the distribution, which is why they are chopped off in the algorithm.
Then it's just a matter of 1st year calculus doing some basic integrals to know relationships between V, G, the normalization of the exponential, and the decay constant of the exponential. Once that's done it's another simple integral to get a prediction for m.
Note that the above equation doesn't quite apply to the 1st land rush as that one had guilds with 0 votes whereas here they all have a minimum of 1. So the equations change since the integral's lower limit is 0 there and 1 for this 2nd land rush. The formula for a 0 vote minimum is:
m ~ [V/G]*log[G/(30-N)]
m was actually 6 there and this postdicts 5.5.
|
Goblin Squad Member
|
| 1 person marked this as a favorite. |
Then it's just a matter of... Once that's done it's another simple...
There you go making me feel bad at math again :) (not really bad, since I do actually understand integrals enough to do it again with some reference material, but still...)
What I'm most curious about is if that formula is a standard formula, or something you devised to fit the data you saw. If it's the latter, then that's where you're leaps and bounds ahead of me.
|
Goblin Squad Member
|
Might I suggest you consider joining a group that's very committed to being Kind and Helpful, and doesn't make demands on how you spend your time?
The Seventh Veil
Recruitment Thread
Land Rush Guild LinkI'm also happy to vouch for the members of the Roseblood Accord.
And although I don't really have any relationship with them other than my immense respect for the people involved, I hope they won't mind if I also recommend Ozem's Vigil and The Golden Flask.
Thanks!
I'm hoping to tag along with a group of my regular table-top gamers, but if they don't have a preconceived notion, I will look into those.
|
Goblin Squad Member
|
Then it's just a matter of... Once that's done it's another simple...There you go making me feel bad at math again :) (not really bad, since I do actually understand integrals enough to do it again with some reference material, but still...)
What I'm most curious about is if that formula is a standard formula, or something you devised to fit the data you saw. If it's the latter, then that's where you're leaps and bounds ahead of me.
Land Rush 1 vote distribution (excludes TEO, Pax, T7V). You can fit the data there to an exponential. It's not perfect (which is always the case when dealing with real data), but it works.
Edit: As for why it's an exponential decay which describes the number of guilds vs. number of votes...no idea. It just happens to work and is a simple function to work with.
|
Goblin Squad Member
|
Looking at some of the guild names, I started to think about if there is going to be some kind of naming policy for settlements, companies etc...
I know the guilds are temporary, but still.
I would find it a bit annoying to see a settlement called Ogrimmar in the game (just an example).
This is being covered here: Naming Settlements and Companies
|
Goblin Squad Member
|
My coding skills are more analysis related.
In my world (banking), we call that the difference between being a programmer, an analyst, or a programmer/analyst; it all depends on whom one can rely for adequate backup to cover one's weak skills. I've been the latter two in my career, but I'm also primarily an analyst who thrives on data from others.
It's incredibly important, though, to make sure to have a hand in getting that data, though, or one'll have no idea what the hell one's being handed.
|
Goblin Squad Member
|
| 1 person marked this as a favorite. |
It's incredibly important, though, to make sure to have a hand in getting that data, though, or one'll have no idea what the hell one's being handed.
So true!
Been bitten by that one more than a few times in my research.
"Here's a plot of Y vs. X which shows ..."
"What's Y?"
"Ummm...lemme get back to you on that."