Evaluating Recruitment

[[violet]]

I guess my surprise was more in the fact that there weren't nearly as many recruitment TGs targeting revived nations

The majority target new nations. I'm not sure if this is by choice or whether NS++ makes that the default (or the only?) choice or what. But to give you an idea of the scale, of the last 10,000 telegrams the system processed, 9,914 were recruitment telegrams that bounced back into the queue due to being directed at a new nation.

[Shadow Afforess]

I guess my surprise was more in the fact that there weren't nearly as many recruitment TGs targeting revived nations

The majority target new nations. I'm not sure if this is by choice or whether NS++ makes that the default (or the only?) choice or what. But to give you an idea of the scale, of the last 10,000 telegrams the system processed, 9,914 were recruitment telegrams that bounced back into the queue due to being directed at a new nation.

NS++ let you target new, refounded, or rejected (TRR new arrivals ), and customize the split between them. I am planning on adding a lot more variety next update: viewtopic.php?f=15&t=269464&start=475#p18508612

[[violet]]

I am planning on adding a lot more variety next update

I just want to remind you about our previous discussions, where I've mentioned the importance of the Telegrams API having a barrier to entry. Because I see you exploring options like distributing the client key throughout all of a region's residents, so they will all automatically send TGs whenever they're logged in. This is clever, but will bring us right back to the same situation, where you've turned script-based recruiting into "check this box and forget about it," requiring no thought or effort, which means everyone will use it, which is unworkable, which will force me to restrict the API.

[Shadow Afforess]

I am planning on adding a lot more variety next update

I just want to remind you about our previous discussions, where I've mentioned the importance of the Telegrams API having a barrier to entry. Because I see you exploring options like distributing the client key throughout all of a region's residents, so they will all automatically send TGs whenever they're logged in. This is clever, but will bring us right back to the same situation, where you've turned script-based recruiting into "check this box and forget about it," requiring no thought or effort, which means everyone will use it, which is unworkable, which will force me to restrict the API.

Not all. Recruitment officers will be able to see the client key, so regions need to appoint trusted members. I will probably put an upper limit on officers too, probably no more than 10 nations (excluding delegate / founder).

I think this is a reasonable barrier to entry. I think recruiting for new/refounded nations is always going to end up a zero-sum game with few winners, no matter the system. That is why I will also be bringing out many new options for regions to try out, so they can opt-out of the new nation game altogether.

If you have other suggestions on possible recruitment variations I have not already listed, or other suggestions regarding recruitment officers, do share.

[Mousebumples]

Question for the techies, as I'm evaluating the best recruiting options, going forward. If a given region uses all 3 recruitment methods (API, stamps, and manual) - or even 2 of the 3, how does the "Recruiting for RegionA" tag work with the theoretical new system?

It seems likely that one of two things will happen:

1) A region can *queue* 3 TGs for their region to the same nation (one through stamps, one through manual, one through script) and as soon as one of them gets *delivered* (determined by the queueing-round robin system), the other 2 get blocked.

2) A region can queue only 1 TG for the same region to the same nation, regardless of method.

Care to provide any insight, so we can make appropriate plans for whenever the future system is coded and released? Thanks!

[[violet]]

You can queue as many as you like, but once the first one makes it through, the rest will fail with "Previous Recruitment Too Recent." So that's your Scenario 1. It would increase your chances of delivery, though.

Code is written for the round-robin; here's how it will work. It won't be a true round-robin, since that would require quite a lot of tracking of data that isn't able to be easily manipulated; instead, it will be more random, which will achieve the same result on average. How the queue works currently is the "postie" service just iterates over and over the queue, trying to deliver recruitment TGs in random order, with most of the ones targeted at new nations bouncing back to the queue because the recipient isn't ready for a new one yet.

What will happen soon is postie will iterate through four stages: Normal, Manual Recruitment, Mass TG Recruitment, and API Recruitment. For each of the three recruitment stages, it will randomly select one matching TG for each receipient. It will then attempt to deliver this, same as now.

What this means in practice is each time a new nation is ready to receive a recruitment TG, there will be a 1/3 chance that it will come from the Manual, Mass TG, or API queues. If the queue it comes from is short--say, it's the Manual queue and only 2 Manual recruitment TGs are queued for this nation--then each TG will have a 50% chance of being the one chosen for delivery. But if the queue is long--e.g. it's the API queue and there are 20 script-sent recruitment TGs for this nation--each TG only has a 5% chance of being the one chosen.

Another thing I'm considering (but haven't coded) is phasing this in, to avoid manual recruitment suddenly becoming astonishingly more effective, a situation that may invite some of the more unscrupulous script behavior we used to see under the old TG system.

[Tyler Kazakov IV]

Why can't we just get rid of API and Stamps?

We don't need them, and its causing trouble, theres no need to keep them.

[Shizensky]

If something isn't working as expected, you fix it. You don't throw it away. These are just the growing pains of new systems.

[Tyler Kazakov IV]

We'll never be able to contain it.

The "population" for the nations will do nothing either.

The API doesn't even help regions. I've learned the only way to actually do anything is with Stamps. I've only seen one region which actually used the API and it got them a lot of nations.

Look at Europeria, perfect example. It doesn't help, so:

1. You need to get rid of those stupid rate times for the nations that get the recruitment telegrams.
2. We need to increase the amount of inbox for people, from 20 to 30.
3. We need to make Manuel Recruitment the best option.
4. Get rid of the API, or increase the rate times.
5. Get deals with Stamps, so that more people buy them.

But then again, no one will listen to me.

[Andacantra]

The point is that old "manual" recruitment was either scripted (and just bloody hard to detect/keep a handle on) or so close to scripted that it just involved a person sitting there clicking a button. That's not an ideal situation either and the manual/API/Stamps system does have the potential to work, especially if there are also alternative avenues for recruitment and it doesn't just result in a crudton of spam in the early stages of a nation's life.

[The Black Hat Guy]

Our API did quite a good job of keeping the population at least stable in the VF, at least until Afforess' NS++ came along. With that reduced again and manual recruiting weighted, I don't see a problem at all with keeping stamps and scripts.

[[violet]]

Anyway, I think if we take my marketing ideas into actual thought, then we should make stamps the best thing. But thats just my $0.02

Like I mentioned, best way would be for Manual recruitment to have the upper hand. With API and stamps not.

For one, the game has made it extremely harder to manually recruit. With the time it takes to reach new members, along with with stamps are cheap, easy ways to get the nations, plus they've allowed us to recruit from UCR's. They made it a game that money buys you a region.

Why can't we just get rid of API and Stamps?

5. Get deals with Stamps, so that more people buy them.

I'm finding your position kind of perplexing, Tyler.

[Tyler Kazakov IV]

I'm bipolar on this issue.


OFFICIAL, NO matter what I say: I think API and Stamps should get the boot, while Manuel becomes easier.

[Astarial]

The point is that old "manual" recruitment was either scripted (and just bloody hard to detect/keep a handle on) or so close to scripted that it just involved a person sitting there clicking a button.

This is a claim that gets tossed around a lot, without any kind of evidentiary support. I'm quite sure that there were scripts of various kinds (both legal and illegal) being used for a long time, but not in all regions and not by all members of the regions who did have them. I can assert for myself, and I'm sure for several others in Equilism (though not necessarily all), until about 3/4 of the way through 2012 I was still lurking the five Pacifics' region pages for new nation creations, and customizing my telegrams myself before pasting them in and sending them. A fantastic day sometimes got me 1 nation every 20 TGs, while an average one was perhaps 1 for every 100.

It was only near the end of 2012 that I got access to a semi-automater - a script which pulled names off of the 50 most recently created nations page, made them link to their nations, and automatically filled out a telegram for them. I can't say this dramatically changed my results, but it certainly made it easier to recruit while also doing other things. That was barely more than a year ago, and I'm relatively confident that the vast majority of my region-mates had no access to recruitment assistance before that - and there's simply no way we're unique there.

I guess my surprise was more in the fact that there weren't nearly as many recruitment TGs targeting revived nations

The majority target new nations. I'm not sure if this is by choice or whether NS++ makes that the default (or the only?) choice or what. But to give you an idea of the scale, of the last 10,000 telegrams the system processed, 9,914 were recruitment telegrams that bounced back into the queue due to being directed at a new nation.

For me, it's always been a deliberate choice - I've never known refounded nations to give anywhere near the kind of return on investment to make recruiting them worth it. When we recruited manually, it was incredibly rare to get a nation from any of the sinkers, and during recent (admittedly relatively small-scale) tests with stamps, there were several new nations who responded but none from refoundeds. And that fits broadly with what I would expect, too - most refounded nations are either puppets (often R/D), or players wanting to return to what they were doing - if that's issue answering, they might stay put, but others already know which region they want to rejoin. There are very, very few who are returning to the game, with an old nation, and open to being recruited somewhere.

I would be curious to know overall how the response rate to recruitment TGs varies between the two groups - what percentage of nations that move out of their founding/refounding region do so in response to a recruitment TG?

[Bears Armed]

I was still lurking the five Pacifics' region pages for new nation creations, and customizing my telegrams myself before pasting them in and sending them. A fantastic day sometimes got me 1 nation every 20 TGs, while an average one was perhaps 1 for every 100.

Same here, back when I had enough time to do much recruiting.

[Andacantra]

Well, around that time's about when I'm referring to - especially in the interim after auto-scripts got banned. I don't have numbers because I can't have numbers, but from personal experience that's how things were and had been for a while.

[Unibot III]

And if you contact people more personablely and selectively in a UCR, you can get a 3/5 acceptance rate - but some people don't want you to know that.

[[violet]]

I really only started paying close attention in 2013, and my impression is that the landscape changed quite a lot in the year or two before that, as autoTG scripts proliferated. But certainly in early 2013, before the new system came in, the great majority of recruitment TGs were sent by scripts. Not all, but an awful lot. There was also a gray area around some recruitment tools, where it was hard to tell how manual some recruiting really was, and at least a couple of tools existed that allowed a player to "manually recruit" by clicking repeatedly without even looking at the screen.

[Tyler Kazakov IV]

I really only started paying close attention in 2013, and my impression is that the landscape changed quite a lot in the year or two before that, as autoTG scripts proliferated. But certainly in early 2013, before the new system came in, the great majority of recruitment TGs were sent by scripts. Not all, but an awful lot. There was also a gray area around some recruitment tools, where it was hard to tell how manual some recruiting really was, and at least a couple of tools existed that allowed a player to "manually recruit" by clicking repeatedly without even looking at the screen.

She's onto us.

[Americans]

The problem with the old system as far as I knew was that players could hide their script so that it looked manual. I still think there were easier solutions. The check box for recruitment would have been an awesome addition by itself. If we had that under the old system, and each sent recruitment TG was logged on some page in regional admin or on the page it would be a lot easier to spot a bot (time stamps that show even rates, lack of eating or sleeping time, etc.) But even at that, I prefer a system where some can cheat to get ahead, and some can work to get ahead. The current system is you can't work to get ahead, you can only watch a script and pray or ditch the script and pay.

The old system wasn't perfect, but it worked for years. It is completely unsurprising that a system encouraging scripting is not working.

And unless you've fixed things since I was last awake (a month or two) most revived nations have full inboxes from when they played before. If they don't, then they were a puppet. Either way that'd be a waste.

[The Dark Star Republic]

What this thread doesn't have much of is input from new nations about how they want to be recruited, and how they find the whole thing. It's mostly recruiters, and people whose 'new nations' are puppets. The impact of signing up to a website and immediately getting deluged with join requests has to be different than getting those requests to a sock puppet account for which you can already anticipate such requests coming.

[Evil Lord Sauron]

What this thread doesn't have much of is input from new nations about how they want to be recruited, and how they find the whole thing. It's mostly recruiters, and people whose 'new nations' are puppets. The impact of signing up to a website and immediately getting deluged with join requests has to be different than getting those requests to a sock puppet account for which you can already anticipate such requests coming.

It's highly unlikely new nations would find there way here to comment and or read this with any amount of understanding. Though it would be good for them to do so, if they are new they'll be exploring the game and acclimatising to the game as a whole.

[[violet]]

What this thread doesn't have much of is input from new nations about how they want to be recruited, and how they find the whole thing.

Not a new player, but I think recruiting is really great for new players today. They create a nation and quickly receive a welcome telegram from their Feeder, which helps explain where they are and what they can do. Then recruitment TGs from other regions start to arrive, which are also very well-written these days; some are funny, some are interesting, all of them show there's plenty of life in the gameworld and a lot to explore. These are spaced out so the player gets a few messages in the first few minutes and then fewer as time goes by, so they're not overwhelmed all at once.

Recruitment TGs targeting new nations today do a much better job of welcoming new players to the game than the official site welcome TG does.

[[violet]]

What will happen soon is postie will iterate through four stages: Normal, Manual Recruitment, Mass TG Recruitment, and API Recruitment. For each of the three recruitment stages, it will randomly select one matching TG for each receipient. It will then attempt to deliver this, same as now.

What this means in practice is each time a new nation is ready to receive a recruitment TG, there will be a 1/3 chance that it will come from the Manual, Mass TG, or API queues. If the queue it comes from is short--say, it's the Manual queue and only 2 Manual recruitment TGs are queued for this nation--then each TG will have a 50% chance of being the one chosen for delivery. But if the queue is long--e.g. it's the API queue and there are 20 script-sent recruitment TGs for this nation--each TG only has a 5% chance of being the one chosen.

This is now implemented. I'll post a new thread in Technical to announce it properly, but just giving you the heads-up.

When Afforess turned off NS++ recruiting about a week ago, the number of API-based recruitment TGs fell from ~75% of the total to ~40%. Most of the rest is stamp-based recruiting, with very little Manual recruiting. I understand that a new form of NS++ recruiting is on the way, so presumably this will change again, but that's how it is today.

The new system means that the less of each type of recruiting there is, the more chance its TGs will arrive sooner in new nations' inboxes. So right now, manually sent recruitment TGs have a much higher chance of being delivered sooner to a new nation than TGs sent via the API or with stamps.

Please note that in the past we've had issues with scripts breaking the rules and attempting to perform manual recruitment, which is a violation of Script Rules. All telegrams sent with a script or browser tool must use the Telegrams API. Regions that violate this will be barred from recruiting.

[Mousebumples]

When Afforess turned off NS++ recruiting about a week ago, the number of API-based recruitment TGs fell from ~75% of the total to ~40%. Most of the rest is stamp-based recruiting, with very little Manual recruiting. I understand that a new form of NS++ recruiting is on the way, so presumably this will change again, but that's how it is today.

Would it be possible to request semi-occasional updates about the breakdown?