Atwar Analysis

This is ezzatam speaking, known for map making and formula breaking This time I bring some interesting facts and my opinion of the AW system.

Data between 22/12/2012-09/01/2013

1) Number of Players in AW by Rank



This is meant to be a graph with positive skew (imagine it flipped) but there is something wrong with the curve. There are more Rank 4s than Rank 3s and there is a big drop down to the number of Rank 5s. I call this the Game Threshold or the Rank 4-5 Gap. Also, there is a gap between Rank 6 and Rank 7, I suspect this is due to premium medals being needed to sufficiently rank up from 6 into Rank 7. I call this the Premium Threshold or the Rank 6-7 Gap.

2) Percentage of Premium players by Rank



High ranks players are mostly premium payers... NO surprises there. The percentage of Rank 7 premium payers compared to Rank 6 is evidence of my theory about Graph 1 working, and the precentage of Rank 8 premium payers are obviously people who bought premium at Rank 7 and progressed. You should also note how the premium bar of Rank 5 hasn't jumped compared to Rank 4 despite what I said about them in Graph 1. Clearly, even with ~30 hours of gametime, those Rank 4 players don't deem it as sufficient to invest their money on premium.

3) Number of AW players with Premium Status by Rank



This is how AtWar game community is right now. The free players mostly hover over the Rank 1 to Rank 6 mark and the Pay-to-Win factor kicks in at about Rank 7 mark as said before. Rank 7 is about 200 hours of gametime for references.

4) Premium v Free % of Players



The strongest proof that AW isn't pay-to-win for the 80%+ majority of the players. This links in to Graph 3 and not much needs to be said.

5) Number of Premium Players by Rank



However, this is the most interesting part. Remember that I mentioned about Rank 4 not ranking up and leaving? Believe it or not they are the greatest source of income in the WHOLE game. The contribute through subscription packs and yet they are treated very badly. The Rank 8 peak as explained in Graph 2 already, yet this Premium Threshold is only the 4th largest group, behind the all important trio of Rank 4,5,6 respectively.

6) Cumulative number of Players by Rank



You might ask, by now, what's wrong with Atwar system? At first sight this above Graph seems to be showing no problems. 1/5 of players are Rank 6+ and the admins are always catering for those top 20%. But, that 20% don't deliver 80% of the income as Pareto principle states. See Graph 5? Those ~300 premium players (Rank 4 and 5) are not given the run of their money and that's another reason they are quitting prematurely.

7) SP Increments by Rank (Gradient)



The last Graph brings us to the root of the problem in my opinion. The SP jumps. As clearly shown above, the linear proportionality disappears after Rank 5. In other words, they aren't ranking up enough for them to consider continuing to play this game (See Graph 1 and 5). We need to prolong this linear line in order to enlarge the number of players online at any one time. And I think I have the solution.

Suggestion:HERE

COMMENTS PLEASE!

How are rank 4's treated badly?

¡Excelent work! Nice statistics...
... but I have some questions:
(1) the gaps between ranks 4 and 5 (graph 1), ¿can't they be explained with the SP promotion weeks that boosted some players? or ¿the end-of-the-year-vacation-player-boom?
(2) ¿where do you get the information about the premium members and when they acquired their premium membership? I mean, the graphs do not tell us the moment players acquired their premium membership; could have been when tehy were tank 3 or 4 or 5 or 6, etc. Probably most of them acquired premium membership between rank 4 and 5. ¿Isn't it?
(3) Premium members are more persistent and, conclusively, they reach higher ranks. Not that premium membership is required to get to these superior ranks... ¿what do you think about this?

Again, excelent work... nat only this post but all the maps and stuff...
Best regards,
Columna Durruti

Skrivet av Columna Durruti, 10.01.2013 at 14:30

(1) the gaps between ranks 4 and 5 (graph 1), ¿can't they be explained with the SP promotion weeks that boosted some players? or ¿the end-of-the-year-vacation-player-boom?

SP promotion week would mean that more Rank 4s would rank up to Rank 5 and closing the gap. That would imply that the gap was even wider before. Also, I assumed that the proportions of players in each rank are the same for the last 30 days+ so player boom shouldn't offset the numbers that much.

Skrivet av Columna Durruti, 10.01.2013 at 14:30

(2) ¿where do you get the information about the premium members and when they acquired their premium membership? I mean, the graphs do not tell us the moment players acquired their premium membership; could have been when tehy were tank 3 or 4 or 5 or 6, etc. Probably most of them acquired premium membership between rank 4 and 5. ¿Isn't it?

I counted the number of premium (Gold) rank icons for each rank but date of acquiring premium membership is something too difficult to trace. The moment players acquired premium is more likely to be discount-dependent than rank to be honest, I just assumed that their proportions are the same at all times. By premium, I think you mean subscription package and if most acquired subscription between Rank 4 and 5 then they won't have enough game time to get those SP boosts by medals. That is an admin problem, since many low-ranks miss out on the benefits.

Skrivet av Columna Durruti, 10.01.2013 at 14:30

(3) Premium members are more persistent and, conclusively, they reach higher ranks. Not that premium membership is required to get to these superior ranks... ¿what do you think about this?

It is the other way around, persistent players are more likely to be premium members and reach higher ranks. Therefore, we have to adapt and influence members to be more persistent. I think there is a lot of potential in the Premium membership, 85%:15% is too low. I think many of us will be happy to accept 70%:30% balance

Thanks for your comment!

Skrivet av ezzatam, 10.01.2013 at 15:06

Skrivet av Columna Durruti, 10.01.2013 at 14:30

(1) the gaps between ranks 4 and 5 (graph 1), ¿can't they be explained with the SP promotion weeks that boosted some players? or ¿the end-of-the-year-vacation-player-boom?

Skrivet av ezzatam, 10.01.2013 at 15:06

SP promotion week would mean that more Rank 4s would rank up to Rank 5 and closing the gap. That would imply that the gap was even wider before. Also, I assumed that the proportions of players in each rank are the same for the last 30 days+ so player boom shouldn't offset the numbers that much.

And rank 5 would boost upwards to rank 6... leaving a certain gap... and so on. To know more about this 4-5 gap we have to check other evidence like: when have they started playing (seasonal variation), for how long, how many games, how many turns, etc. To know more about the 4-5 gap. But you are right about this phenomenon.

Skrivet av Columna Durruti, 10.01.2013 at 14:30

(2) ¿where do you get the information about the premium members and when they acquired their premium membership? I mean, the graphs do not tell us the moment players acquired their premium membership; could have been when tehy were tank 3 or 4 or 5 or 6, etc. Probably most of them acquired premium membership between rank 4 and 5. ¿Isn't it?

Skrivet av ezzatam, 10.01.2013 at 15:06

I counted the number of premium (Gold) rank icons for each rank but date of acquiring premium membership is something too difficult to trace. The moment players acquired premium is more likely to be discount-dependent than rank to be honest, I just assumed that their proportions are the same at all times. By premium, I think you mean subscription package and if most acquired subscription between Rank 4 and 5 then they won't have enough game time to get those SP boosts by medals. That is an admin problem, since many low-ranks miss out on the benefits.

Again, depends on the moment of acquisition...

Skrivet av Columna Durruti, 10.01.2013 at 14:30

(3) Premium members are more persistent and, conclusively, they reach higher ranks. Not that premium membership is required to get to these superior ranks... ¿what do you think about this?

Skrivet av ezzatam, 10.01.2013 at 15:06

It is the other way around, persistent players are more likely to be premium members and reach higher ranks. Therefore, we have to adapt and influence members to be more persistent. I think there is a lot of potential in the Premium membership, 85%:15% is too low. I think many of us will be happy to accept 70%:30% balance

Thats what I said: persistent player becomes premium player... right about 70:30!!!

Skrivet av ezzatam, 10.01.2013 at 15:06

Thanks for your comment!

Thnx,
Columna Durruti

This is awesome! Great to see these stats!
Can i ask where you find this information?

I attribute the Rank 4 players leaving to the fact that they're forced out of the Beginners room and therefore completely raped by high level players, leading to frustration. I've been through that and I believe a lot of players when going from Rank 3 to 4 went through that as well.

I hardly, HARDLY, believe that AW is Pay-to-Win.

Skrivet av notserral, 11.01.2013 at 11:52

I attribute the Rank 4 players leaving to the fact that they're forced out of the Beginners room and therefore completely raped by high level players, leading to frustration. I've been through that and I believe a lot of players when going from Rank 3 to 4 went through that as well.

I hardly, HARDLY, believe that AW is Pay-to-Win.

I agree 100% with this statement. AW is extremely far from a Pay-to-Win game.

Also most of high ranks are Beta troppers.

Skrivet av notserral, 11.01.2013 at 11:52

I attribute the Rank 4 players leaving to the fact that they're forced out of the Beginners room and therefore completely raped by high level players, leading to frustration. I've been through that and I believe a lot of players when going from Rank 3 to 4 went through that as well.

It's a good interpretation of the stats in my opinion.

Skrivet av notserral, 11.01.2013 at 11:52

I attribute the Rank 4 players leaving to the fact that they're forced out of the Beginners room and therefore completely raped by high level players, leading to frustration. I've been through that and I believe a lot of players when going from Rank 3 to 4 went through that as well.

The problem with raising the Beginners Room Rank will then be Rank 4s raping Rank 1s... Which is why we need some sort of tier system, or in my opinion, more ranks.

Skrivet av Hugosch, 11.01.2013 at 02:20

This is awesome! Great to see these stats!
Can i ask where you find this information?

I compiled it myself using the player list open to all.
You can find all my data:HERE

How did you count premium members? You can't see all players in player list.

WoW ezza you are awesome.

Nice stats

Superbe!!

It seems like this has been thoroughly explained on all your points. I do agree with your other thread's suggestion as well.


I know you don't have access to everything, but it seems to be that without the time periods the information is incomplete. It'd be hard to complete without when and at what rank people got their premium.

Also there are a number of "beta" accounts that were made and have sat there for a long time, they all automatically have gotten premium and probably take up half of the premium low ranks.

Skrivet av Desu, 11.01.2013 at 15:35

Also there are a number of "beta" accounts that were made and have sat there for a long time, they all automatically have gotten premium and probably take up half of the premium low ranks.

Data from the last 30 days is used... beta accounts would have either ranked up or quitted by now. lol

Skrivet av ezzatam, 11.01.2013 at 16:05

Skrivet av Desu, 11.01.2013 at 15:35

Also there are a number of "beta" accounts that were made and have sat there for a long time, they all automatically have gotten premium and probably take up half of the premium low ranks.

Data from the last 30 days is used... beta accounts would have either ranked up or quitted by now. lol

Solved then.

At least we have a time period for the statistics--put that in your original post.

Updated Time Range for people who needed to know:
22/12/2012-09/01/2013

Great job... thnx for the XLSX...
Cheers,
CD

I appreciate your graph
CHEERS, I'M KALIRAA

Nice work.

I think the gap between rank 4 and 5 is made by the difference between beginners and casual lobby, while dominating in beginners room, rank 4s die in the casual room. getting kicked out of games and being killed all the way makes rank 5s leave the game...

Great graph I liked it

I closed this because this thread was necro'd from last year. I did not make a post saying I did so, sorry about that.

Closed again unless ezzatam is back and wants to discuss NEW info. Not last years.