Concentration of power

DeletedUser

Guest
A very interesting topic MK. Very thought-provoking and painful, both at the same time. I think I may have it a little easier since I've yet to take any economics/statistics classes so I have no predetermined ideas that I feel may be concrete. We hit the jackpot with you in our tribe.

Dunbop did bring up two good points. One being his Monty Python reference, which I got a good laugh out of. If it was ever put to vote, there are still those stubborn players who would refuse to vote to end the world, whether it be out of spite or just plain ignorance. His second point on the limit of tribe members affecting the C4 ratio would also be correct, seeing as W1 has a much higher tribe limit than ours, so I'd save that it would be safe to add the "junior tribes" for statistical comparison.

Savik, I didn't think it was possible but we seem to have another MK on our hands. I only understood about half of what you said, but that half was very interesting.
 

DeletedUser

Guest
Although my knowledge of economics is very limited, I must say that this is very interesting indeed. I hope to see a lot more of your statistics to give (in my opinion) an objective look on this world.
 

MichielK

Guest
W16 update

Points-based:
C4: 59.73% (+ 0.92%)
HHI: 10.76% (+ 0.24%)

Village-based:
C4: 57.41% (+ 0.99%)
HHI: 10.05% (+ 0.26%)

Discussion
This is the first weekly update I've done, so there's not much to say except that the world became a little bit more concentrated. As we collect more data like this, we should be able to see the effect of certain events on the concentration of power.

For example, what happens to (the change in) our C4 and HHI when:
- a major war breaks out?
- a major merger occurs?
- a target (Ad Inf, PnX, Dust) disbands?
- summer vacation ends?

We'll find out more about this as time goes by, but in the meantime I did two projects (one big, one small) that might be interesting:

Project 1: where does W16 rank among other worlds?

Last week, I compared W16 to W1 and came up with scenarios that would put our concentration of power in line with theirs. That served as a decent introduction to the numbers and how they work.

This week, I'm going to broaden our scope a little bit and look at the first 20 worlds. Here are the top 12 worlds ranked by their C4 (all non-listed worlds are below 50%):

1. W8: 83%
2. W7: 81%
3. W10: 81%
4. W1: 79%
5. W6: 75%
6. W12: 72%
7. W4: 71%
8. W2: 64%
9. W5: 60%
10. W9: 60%
11. W16: 60%
12. W11: 52%
All other tribes are below 50%.

The first thing that strikes me is that the order is definitely not based on when a world started, since the highest-ranking worlds are by no means the oldest ones.

The most striking example of this is W3, which doesn't even pass 50%. There's a logical explanation for this though: family tribes. W3 has three large families calling the shots, and they have "lower branches" high in the rankings (4th, 5th and 7th). If you add up all the branches as a single tribe for all three families, the C4 goes to 69%...more in line with other worlds that started at the same time.

A little digging shows a factor that is important: maximum tribe size. As it turns out, every single world in the top 12 has a tribal limit of 80 or higher, and the entire top 6 has a tribal limit of 100 or above (in some cases far above).

(Kudos to Dunbop for realising this immediately!)

These high limits have a number of effects, all of which contribute to a higher concentration of power:
- family tribes don't exist, since everyone can be squished into one tribe
- mass-recruitment is more effective
- there is less reason to cut inactive/incompetent tribemates
- mergers can occur far more easily

For next week, I'd like to dig a little further into this effect. However, I'll need your help! I don't play other worlds, but I know many of you do and you could really help me out here.

What I need are the names of worlds that started at roughly the same time but have very different tribal limits. E.g. if world 27 and 28 started at the same time but W27 has a tribal limit of 40 and W28 of 200, that allows me to compare them directly. With a few pairs like that, I can confirm and possibly quantify the effect of max tribe size on concentration...leading to a better analysis :)

Project 2: what's required for the end of the world?

As we all know, TW is essentially an open-ended game. In principle, there's no reason for any world to ever close as long as the owner make money on it.

However, in reality there are two worlds that have already been closed: DE1 and DE2. These were (to my knowledge) not closed due to lack of income, but rather due to lack of competition: one tribe dominated the world to such an extent that there was no point in going on.

Assuming TW is consistent in their logic, the stats of these worlds should provide a clue to what's required for further worlds to close. It is to be expected that if e.g. a tribe in W16 ever shows the same domination, W16 will be closed and we'll all find ourselves with a ton of spare time :icon_wink:

In this project, I'm using HHI instead of C4. The reason for this is that C4 focuses on the concentration of power among the top 4 tribes; it doesn't matter if these 4 tribes are all the same size or if one is huge and the other 3 are small.

Obviously, for this analysis it is important if the power is concentrated among one tribe or four, and HHI can deal with that quite nicely. Here's how they look:

Points-based HHI:
DE1: 63.2%
DE2: 52.2%

Village-based HHI:
DE1: 62.8%
DE2: 51.2%

Unless evidence to the contrary pops up, I'm going to assume that 50% is the cutoff. If a world goes above 50% in HHI, a very valid case can be made that the world is over and it should be closed.

(Note that I have no evidence of worlds with a high HHI that are still open. Most worlds tend to hover under 20%, so it's perfectly possible that the cutoff is much lower than 50%...we just don't have any examples yet!)

For those of you wondering how we can get W16 to 50%...one tribe would have to eat every village in C², Plight, HRV, BANG, TKR and Ni. That's over 400 million points and nearly 45,000 villages, and the next-biggest tribe would be LS with 26 million and 2760 villages. It's that difficult.

Looks like we'll be here a bit longer! :icon_wink:
 
Last edited:

DeletedUser

Guest
well all i can say is wow..... i need another scotch just reading that..... made my head spin, thats some genius work there... so does that mean that 1 tribe needs to own a certain number of villages or a certain percentage of them on the world? (i know im daft lol)
 

DeletedUser

Guest
Yay, finally a topic worth reading, it's been awhile.

Very intrigued by this post, please tell more! :icon_eek:
 

DeletedUser

Guest
In this spoiler is a map of our Worlds End as said by MK :)

[SPOIL]
worldsend.jpg
[/SPOIL]
 

DeletedUser

Guest
Just going by the tribes MK stated at least. I'm pretty sure we wouldn't see those gaps where PnX and Ad Inf currently reside, and just have an arc like Bang! has in the north.

Of course, very interesting as always MK. I have to admit I got lost a few times and had to look back at the original post to see what C4 and HHI meant again :icon_redface:
 

DeletedUser

Guest
Ok mk as you stated w3 is family. w2 is also family. I am going to break it down would you be willing to go back and calculate both equations for the groups?

Thanks :)

Familys:
TW
TW1
TW3
TW4
TW5


AtenCE
AtenRX
AtenRa
Aten-X

Ok all others are now solo :)
If you look through you will see two hot tribes, but they are not family anymore. :p

As for the w16 part of the stats I find it very interesting but until multiple people are quitting, that are over a million points, then we can start saying that the world is closer to the end. lol

The player count down I find slightly more accurate since the people in the monopolizing tribe do get bored or tired of stalemate fighting and those people quitting help equal out the tribes as you near the end.
 

MichielK

Guest
Ok mk as you stated w3 is family. w2 is also family. I am going to break it down would you be willing to go back and calculate both equations for the groups?

If it was just you, sure. However, it's you plus all the in-game mails plus all the forum mails, so...no.

Yes, I realise that some worlds are extremely 'family-oriented'. Yes, I realise that this strongly affects the concentration variables. However, even with the families broken down it would cost me a ton of time to go through all the requests I've received. This stuff is already pretty time-consuming, and redoing more than half the worlds would overshoot the purpose...which is finding and posting information about W16.

I intend to take requests for data occasionally, but I want to limit that to W16 or to truly special situations. I understand that for a player of W2 that world is unique, but for me it simply isn't :icon_wink:

There's another reason why I won't do this, and I'll use W18 to illustrate why since it's such a ridiculous example. Here are the top 20 tribes for W18:

1. Apoc-M - 179 million points
2. Apoc-D - 178 M
3. Apoc-W - 124 M
4. ~S~ - 116 M
5. Apoc-I - 114 M
6. Apoc-C - 108 M
7. {BA} - 104 M
8. Apoc-F - 103 M
9. [BA] - 84 M
10. Apoc-N - 78 M
11. FEAR - 72 M
12. ApocM2 - 36 M
13. Apoc-R - 34 M
14. INF-AM - 27 M
15. ~S1~ - 25 M
16. Apoc-H - 25 M
17. ANAR - 19 M
18. outz - 13 M
19. WarGod - 7 M
20. BANZAI - 7 M

Now, it looks like this world essentially consists of:

  • Apoc family - 1st, 2nd, 3rd, 5th, 6th, 8th, 10th, 12th, 13th, 16th - 2405 million points
  • {BA} family - 7th, 9th - 245 million points
  • ~S~ family - 4th, 15th - 214 million points
Here is what the concentration of power looks like:

C4 (points) - 95.8%
HHI (points) - 84.9%
C4 (villages) - 92.6%
HHI (villages) - 75.1%

Compare this to the two German worlds posted above, and this world is far more concentrated. A good case could be made that the world is over and should be closed. However, I disagree.

(In the example below, I'm using a fictional duke of Apoc-M as an example. I do not know who the duke of Apoc-M is, I do not know if he is as bad as I make him out to be, and for all I know he could be the best thing that ever happened to TW. I simply needed an example, so please remember that this is fictional.)

Imagine that you're the leader of Apoc-M. You're the leader of the number 1 ranked tribe, and you're a duke in the family that effectively won this world. You must feel like you've earned some respect and credit, right?

However, we see a tribalhugging and mass-recruiting player who leads a tribe with more than 30 members who show no growth at all in the past week (and well over 50 with inactivity problems), who formed a family with more 7 of the top 10 tribes, and who still didn't feel safe enough and allied himself with 2 more. What's there to respect or give credit for?

Given the fact that he gets laughed at anywhere outside of his own world, his own players are quitting and going inactive due to boredom, and frankly the whole TW experience has been very unfulfilling, who here doesn't expect him to turn on his brother and sister tribes? Seriously?

Families of this size won't last. They'll collapse under their own weight sooner or later, at which point the world (and the concentration numbers) turn back to normal.

The families on W2 might last a bit longer since they can at least fight eachother for supremacy, but after one family wins I suspect the same thing will happen.

My personal philosophy is that a family that "wins" a world will soon fall apart and fight eachother, and that's a good enough reason for me not to calculate the stats of highly 'family-oriented' worlds.
 

DeletedUser22370

Guest
Wow, if I were playing that world I would definately quit :p
 

DeletedUser22370

Guest
Well, if I were in Apoc it would be so boring that I couldn't cope, and if I were in any other tribe, there would be so many noobs to wade through that it wouldn't be worth bothering with.
 

DeletedUser

Guest
The W18 example brings back an issue that I was considering in the "Victory Conditions" thread. To what extent should a family of tribes be considered a single tribe? Obviously the scenario outlined by MK suggests that if we were to consider the APOC family as a single tribe, the world would be properly saturated. However, the qualitative discussion of picking off weaker elements of one's own family is one with which I personally agree.

In order to keep this on topic, to what extent should family tribes be considered a single entity as opposed to multiple individual entities for the purpose of measuring concentration of power? And as a tie in to the Victory Conditions thread, to what extent should a family of tribes be considered as a single tribe for declaring victory?
 

DeletedUser

Guest
this is why i quit w18 back at new years lol...... i was in the INF family over in K89 or there abouts.... they kept hugging the other tribes and i had nowhere to expand so i left lol
 

MichielK

Guest
The W18 example brings back an issue that I was considering in the "Victory Conditions" thread. To what extent should a family of tribes be considered a single tribe? Obviously the scenario outlined by MK suggests that if we were to consider the APOC family as a single tribe, the world would be properly saturated. However, the qualitative discussion of picking off weaker elements of one's own family is one with which I personally agree.

In order to keep this on topic, to what extent should family tribes be considered a single entity as opposed to multiple individual entities for the purpose of measuring concentration of power? And as a tie in to the Victory Conditions thread, to what extent should a family of tribes be considered as a single tribe for declaring victory?

That is really a discussion for the Victory Conditions thread, so I'll move the quote there if you don't mind (and if you do, sorry! lol)
 

MichielK

Guest
I just wanted to let you know that I haven't forgotten about this thread, and will start making regular posts and updates again soon. The reason for my lack of new info is that even statistical freaks need vacations (and on that topic, anyone with a love for history, culture, and/or beaches should visit Cyprus!).

Sorry for the bump :icon_redface:
 

DeletedUser

Guest
I just wanted to let you know that I haven't forgotten about this thread, and will start making regular posts and updates again soon. The reason for my lack of new info is that even statistical freaks need vacations (and on that topic, anyone with a love for history, culture, and/or beaches should visit Cyprus!).

Sorry for the bump :icon_redface:

Hmm you go away when your tribe goes to war? If so, thats very gutsy :icon_razz:

Or have you just got back?
 

MichielK

Guest
Hmm you go away when your tribe goes to war? If so, thats very gutsy :icon_razz:

Or have you just got back?

Just got back. I put on my saddest face and whined a lot :)icon_cry: <---see?) to make sure I'd be here when war broke out :icon_wink:

On a different note, does anyone happen to have the tribal ranking for W16 from the before the HRV-LS merger? What I need is A) all the tribes, B) the total points of each tribe (not top 40), and if possible C) the villages of each tribe.

I want to calculate the effect the HRV-LS merger had on the concentration of power in W16, but came back too late to get an accurate pre-merger ranking. I know some of you save this information regularly, so hopefully you can help me out.
 

Seagryfn

Guest
does anyone happen to have the tribal ranking for W16 from the before the HRV-LS merger? What I need is A) all the tribes, B) the total points of each tribe (not top 40), and if possible C) the villages of each tribe.
If you need to, you can get that stuff (for the last 3 weeks) on TWStats by clicking on the "view tribe history" link for each tribe. It gives total points (not top 40) and total villages.

Date Tribe Rank Members Points Villages OD ODA ODD
15/09/2009 LSHRV 3 +72 +97,145,339 +9,877 -406,108,891 -327,727,963 -78,380,928
14/09/2009 LSHRV 3 -65 -83,993,930 -8,499 +415,452,033 +334,442,036 +81,009,997
13/09/2009 LSHRV 3 =69 +85,251,845 +8,626 +413,274,482 +332,627,321 +80,647,161

And if you gather it for the top 20-30 tribes, I'm sure the remaining tribes won't be too statistically different today than they were a day or two ago.
 

MichielK

Guest
If you need to, you can get that stuff (for the last 3 weeks) on TWStats by clicking on the "view tribe history" link for each tribe. It gives total points (not top 40) and total villages.



And if you gather it for the top 20-30 tribes, I'm sure the remaining tribes won't be too statistically different today than they were a day or two ago.

True, and I'll do it that way if I have no other choice. I figured I'd give it a shot to get exact data first though to put the accuracy of the data beyond doubt...basically to avoid some rim tribe coming on here and complaining that their (for them) significant merger was or was not included.
 
Top