Bot Alliances!

[Max]

Just an idea to help take the heat off of alliances like Funkaphilia this round, how about we introduce a series of BOT ALLIANCES???

You could inject say another 120 bots, forming 20 alliances. Then you could split them into 3 tiers of difficulty, strongly correlated with how many are "online" to defend incoming they receive.

For example, a low tier bot alliance would have perhaps 2-3 bots (perhaps randomly selected) to defend any incoming the bot alliance receives.

A mid-tier bot alliance may have 5-6 sending at all times.

A top-tier bot alliance would be able to have all 20 bots on at all times to defend incoming 24/7

And clearly the alliances names would be selected from a pool of past alliance names just like the individual IDs are selected, which could prove entertaining to see

Q1) Would it actually solve the problem of top alliances ganging up on less active/contactable teams around ranks 4-6?

Q2) Could you improve the idea to make things more competitive, that ideally would only require a simple code or algorithm?

 

[Hamilton]

a top-tier bot alliance would be able to have all 20 bots on at all times to defend incoming 24/7 :d

resistance against the bots!!!!

 

[Dimitar]

It would be far too hard to code the bots to defend properly. If they just defend any incs you could draw *all* the defences pretty easily.

 

[Max]

It would be far too hard to code the bots to defend properly. If they just defend any incs you could draw *all* the defences pretty easily.

Very good point - unlike human alliances (who can send fakes) - you could guarantee pulling away all the defenders.

Could use a two mob system perhaps where lethals could be sent to defend at range and non-lethal close firing troops only at close, but yes you are right it would be pretty easy to ensure the defence was pulled elsewhere.

 

[LuckySports]

The amount of coding required to have semi-intelligent defense is insane.. They would have to be able to do hacks and such as well.

But, I think this would be an interesting idea.

 

[CFalcon]

Amount of coding would not be "insane". It would take more than half an hour I'll grant you, but simulating intelligence is by no means impossible, or even difficult in a rock/paper/scissors game like this.

if incoming == route x:
send route y to defend

if incoming == route x + route y:
send route a + route b to defend

if incoming is slightly overwhelming:
send last tick blockers

if incoming is massively overwhelming:
send out to ID 1

Then throw in a random factor governing whether they send fake or real.

That pretty much sums up how non-ftw defence goes. Obviously, what I've put above won't cover all situations, but then I just jotted that down in 30 seconds. Give it proper scrutiny for 30 minutes instead and it'd be fine.

As to your original questions, I don't think it would do much to change pressure at the top. It might push the top alliances out of range faster, but they'll always be willing to gang up to hit the fattest targets, which probably won't be the bots.

 

[Dimitar]

It'll still be pretty easy to draw away defence

 

[Atlantis]

i like this idea alot!

Although i feel sorry for the bots as it is, imagine how hard it would be for them in an alliance though with no gov help!

 

[CFalcon]

It'll still be pretty easy to draw away defence

Only as easy as with any other alliance.

 

[timtadams]

Amount of coding would not be "insane". It would take more than half an hour I'll grant you, but simulating intelligence is by no means impossible, or even difficult in a rock/paper/scissors game like this.

if incoming == route x:
send route y to defend

if incoming == route x + route y:
send route a + route b to defend

if incoming is slightly overwhelming:
send last tick blockers

if incoming is massively overwhelming:
send out to ID 1

Then throw in a random factor governing whether they send fake or real.

That pretty much sums up how non-ftw defence goes. Obviously, what I've put above won't cover all situations, but then I just jotted that down in 30 seconds. Give it proper scrutiny for 30 minutes instead and it'd be fine.

As to your original questions, I don't think it would do much to change pressure at the top. It might push the top alliances out of range faster, but they'll always be willing to gang up to hit the fattest targets, which probably won't be the bots.

I was gonna say something like this. And you could easily have set activity times for the bots depending on tiers, with a small random factor. And then another random factor on going offline for 1-8 hours or something.

It'll still be pretty easy to draw away defence

why is this a problem. I think this is good. It makes bot alliances easier targets than regular alliance, which is supposed to be the point (imo).

 

[willymchilybily]

The amount of coding required to have semi-intelligent defense is insane.. They would have to be able to do hacks and such as well.

But, I think this would be an interesting idea.

bots dont need to hack. they automatically know your mob. there is no level of inteligence that is how they can run away from lethal incomings.
------------------------------------------------------------------------------
wanted this for ages. the game is better with more players. an injection of AI to never let the game fall below critical mass would be awsome. as all the problems now are target related particularly bashing. al;so i dont think bots should be public ids...no need.

also bots should be split. 20 bots per alliance, 4 bots per GMT. gmts: +0, +6, +12, -6. (+ or - a random value between 0-3) for all bots first of all. to simulate a potentially well rounded alliance timezone structure.

then activities. low alliance(6 hours a day) medium alliance (10 hours a day) top alliance (16 hours a day.

so you say well this is getting complex.. before you even talk about defence.... nope. the bots already have this... bots already have offline times. and activity times. azzer already gave a few that. (maybe all of them)..... so all you do is organise the bots into alliances based on activity.
----------------------------------------------------------------------

defence.

easy enough. the bots know what incoming there is. what you have sent, and hopw many out of your total staff have been sent. based on that they can defend. you have 100m units, and sent 50m. and have a troop score of X amount. then the bots send to match or beat 50/100 * x troop score defending.. based on who is online. and based on rank. largest rank defends first then next largest and so on. until they have sent a mob that potentially could be of equal troop score. the troop score is based on the bots troops score not on the actual troops that have been sent. they will aim to first tick where possile every time. and match eta. they always can only send out 1 attack. and 4 defence. they will always try to keep enough gardeners to mimic 3 times thier non innocent troops before buying more useful units. whether they send 1/3rd of thier gardeners + 1 lethal of appropriate eta. or send 100% lethals is random.(0-1) but the formulae only applies to the first bot who is sending.

the odds of the first bot sending real is 1 in 4....but the odds of the second bot sending real is affected by the bot before him. if the bot before him sent real he will send real, if the bot before sent fake.. he will fake.

this way there is not a separate program running the bot alliances and potential to conflict. each bot has its own code. and its actions are based on what has gone before and who is the largest defender... they wil always defend the largest member with incoming first if there are multiple incomings on the same tick. thus it doesnt matter what is sent how many sent or who sent.

and one final thing to stop random easy draws... the defence only sends if the incoming value of troops on that tick is = or greater than 10% of the sender for one mob sent, +3.5% for additional mob, aka up to 24% troops out for 5 mobs sent. that way even attacking at 30% you still need to be large enough to do the job in order to draw defence. or sending out all 5 mobs you need to send more than 24% of your total troops out to get defence. (24/30 = incoming is 80% of bot's score, fair battle needs no real defence)

 

[tobapopalos]

Do we really want more bots in this game than players? :/

Spamming bot IDs is not a replacement for actual alliances.

 

[Elevnos]

This is a good idea imo, even though it could result in a bot-dominant game, it would shake things up a bit too!

Another thing I was thinking of would be to have a 'sister' game, where players can script bots to play in it...

 

[timtadams]

also bots should be split. 20 bots per alliance, 4 bots per GMT. gmts: +0, +6, +12, -6. (+ or - a random value between 0-3) for all bots first of all. to simulate a potentially well rounded alliance timezone structure.

This is the only thing i would disagree with, only because that would not mimic most alliances. Rather, skewer the timezones to be weighted to 0 GMT. I reckon at least 50% of an alliance is typically, or close to, 0 GMT.

Although, you could determine the timezone distribution based on what tier the alliance is. So the higher tier would have a fairly even distribution, while a lower tier would tend to (not necessarily always) have minimal nightcover.

 

[timtadams]

This is a good idea imo, even though it could result in a bot-dominant game, it would shake things up a bit too!

Another thing I was thinking of would be to have a 'sister' game, where players can script bots to play in it...

that would be fun, I would love to do that.

 

[Polo]

This game is supposed to be PvP not PvE. There's already way too many bots as it is.

 

[timtadams]

This game is supposed to be PvP not PvE. There's already way too many bots as it is.

And this game is supposed to have at least 1k semi-active players.

 

[Max]

Do we really want more bots in this game than players? :/

Spamming bot IDs is not a replacement for actual alliances.

I do agree toby, I really do, the idea was intended to relieve the real alliances of some incoming / bashing due to the declining playerbase, that was what I had in mind.

Since it doesn't look like we're going to get an increase of players through advertising etc. this is one alternative that guarantees extra targets for the playerbase, even though they are sub-standard and not as good as the real thing.

On the subject of a sister programmable bot world...

I love it It would be like robot wars - where everyone would program their own automated ID to attack / defend with a particular route. Would be slightly more complicated for everyone to program their own alliances, but you could try a pure solo bot world and see what happens ;D

 

[Elevnos]

On the subject of a sister programmable bot world...

I love it It would be like robot wars - where everyone would program their own automated ID to attack / defend with a particular route. Would be slightly more complicated for everyone to program their own alliances, but you could try a pure solo bot world and see what happens ;D

Of course you could just have it so pre-round start, the players who make the bots can assign them to different alliances, and can manage what alliance their bot(s) are in throughout the game!

 

[Max]

Of course you could just have it so pre-round start, the players who make the bots can assign them to different alliances, and can manage what alliance their bot(s) are in throughout the game!

Hehehe, like an agent working for his player Alternatively - an additional program could be written:

-> If (number of times left undefended in alliance > 3)
-> Leave alliance and apply to another better one
-> Else
-> Remain stalwart