A creator controlled bot?

[Morden]

now people might laugh at the topic but it would help the game a lot.

right basic outline of what i think would help the game a lot for this round and bring more players into the game:

[1]50'000 new accounts run by a bot controlled by the creators to increase possible targets for everyone.
[2]for those bots to have limited attack capabilities but so that they will attack people but not so much that they are overwelming for anyone.
[3] to defend in galaxy unless a full galaxy attack is launched on them, in which case cluster bot galaxies would defend.

^ that would help this round no end. there would be less noob bashing and a lot less people leave because theres no targets or that they get bashed every day.

otherwise we'll lose another 1000 people this round.

it aint hard to do fs so many peoiple have made bots over the round, you could always borrow a copy of the wac bot if you cant make your own.

[Radium-^]

Good idea

[Morden]

its been said before, but i think they need to take it seriously for once.

[Cochese]

I think it's a feasable idea, really.

My only concern would be the Creators using their bot-fleets to interfere in alliance affairs when they thought it was prudent:eek:

[Morden]

Originally posted by Cochese
I think it's a feasable idea, really.

My only concern would be the Creators using their bot-fleets to interfere in alliance affairs when they thought it was prudent:eek:

well with that sort of influx of planets alliance affairs wouldnt involve so many of the new people.

so it shouldnt be nesacary as they'd be fighting the bots and then each other, while the others fight at there different levels.

it wouldnt be too bad if they opened signups for free in like 3 weeks

with limits ofc: no scans past unit, de class is the biggest you can get, and eta -2 is the most you can get till you upgrade with a 15mil planet cap.

^ that coupled with a computer controlled bot would bring players into the game for once. as well as stop alliances having to bash people so much which is in a way a sad thing that we must do for our own survival because pother alliances havent got any other targets apart from us, and it is the same for them.

if anything it would get more players willing to pay for r10, because quite frankly saying that it will be different is one thing prooving it is a completely different ball game,

so why not show us you want to change pa for the better and at leats give this a thought if not implementation.

[whoop]

I like the idea.

-whoop

[Scouse]

I don't remember what the outcome was last time it was suggested but it's certainly a good idea.

And I don't think they'll open free sign ups this round at all. It would just drive people away who might have signed up for rd 10 instead and loved it. Whereas playing this rd, whilst getting bashed and trying to learn in a stagnated universe, would mean they give the game a few days before binning it and never trying it again.

[Petru]

One phrase.

"It requires coding"


Now ask yourself if it's likely to be done

[at0mic.c0w]

Originally posted by Petru
One phrase.

"It requires coding"

they could license killmark's bot. takes care of signing up etc as well...

[Ruler]

pretty impossible to make if you ask me!
If you attack such a bot the whole cluster defends it or you have to code it so complex that would take years I think! (and that's just 1 problem I mention)
though I like the idea if it is possible, GOOD LUCK to the creators if they try to make it!

[Kill]

If I was asked I could put PA Team in contact with a group who could program a bot to handle that, however it would require them buying a pretty good server.

Tempest was reasonably good bot and had goodish AI however would use up most the system resources on the server running it (and that was handling a mere 1600 accounts at most).


Killmark

[ParraCida]

Actually its rather easy, the only problem would be the sending of appropiate fleets and AI's are easely outsmarted, so easy roids for anyone.

[Morden]

Originally posted by Ruler
pretty impossible to make if you ask me!
If you attack such a bot the whole cluster defends it or you have to code it so complex that would take years I think! (and that's just 1 problem I mention)
though I like the idea if it is possible, GOOD LUCK to the creators if they try to make it!

you didnt read my second post properly.

i said that it would only get half decent defence if it came under attack by an organised attack and i suppose could be done so big people have a harder time than smaller.

as for the code nah some of the bots that've been used in the past were very advanced.

[whoop]

Originally posted by Petru
One phrase.

"It requires coding"


Now ask yourself if it's likely to be done

-whoop

[Starbucks]

Originally posted by Kill
If I was asked I could put PA Team in contact with a group who could program a bot to handle that, however it would require them buying a pretty good server.

Tempest was reasonably good bot and had goodish AI however would use up most the system resources on the server running it (and that was handling a mere 1600 accounts at most).


Killmark

what language is it written in?

[G.K Zhukov]

Wont solve anything.

NAHRSWEETFZ would just NAP VVOMM+the rest of the human universe to roid flat those bot planets.

[Petru]

Originally posted by G.K Zhukov
Wont solve anything.

NAHRSWEETFZ would just NAP VVOMM+the rest of the human universe to roid flat those bot planets.

Fighting for humanities freedom against the machine evermind.

Too fking right!

[G.K Zhukov]

Originally posted by Petru
Fighting for humanities freedom against the machine evermind.

Too fking right!

With the amounts of lamers and hypocrits playing this game, I would prefer that the machines won! Die Arnold, DIE!!!!

[InTGuRu]

If you want to play with AI, get PAOL. PA has, is and should always be with and against real people. And when it is listed as a Massive Multiplayer online stategy game, I don't think they mean massive amounts of AI controlled planets.

[Morden]

Originally posted by Petru
Fighting for humanities freedom against the machine evermind.

Too fking right!

sounds like you've been reading the new dune novels eh? The Butlerian Jihad etc ;-)

[furssie]

Oi Morden

The idea is feasible, in fact an offline PA (PAOL) has done it, but to perfect the AI of those bots is extremely hard and will barely take 'til kingdom come of testings/beta in order to perfect those in a marketable p2p quality.

But then again, the 'community' concept of PA pride will be lost. Probably that's why the proposal was put to thrash in the past.

[Tactitus]

Good grief; isn't there anyone left who remembers this game in its heyday?

A bunch of creator-bot-controlled planets would be a tacit admission that not only can't they figure out how to make the game interesting enough to attract sufficient numbers of real players but also that they'd essentially given up trying.

If the creators have any creative juices left they should spend them on improving the game to attract new and more players, and not rearranging the deck chairs on a sinking ship. If they don't, well, stick a fork in it.

Good idea? Certainly

Does the creators wanna hear all day by some whiners that they favorise a certain powerblock and they set their target on a specific powerblock on purpose because they are uberleet and PA creators hates them? Certainly not.

[InTGuRu]

Originally posted by [7]Gunn3r
Good idea? Certainly

I disagree. tbh, there is no way I can stay up, miss sleep/school/work to play a war game v. AI. I can do that by myself, on my own time.

[cypher]

i've said this 2 rounds ago i think but doubt it'll ever be implemented:P

[hydridia]

I love this idea!!!!!!!!1
plus i have to say i also love killmark for once hehe sounds like a great idea and im sure it would make the universe a little more interesting

[hinch]

to those of you that thinks it needs coding you`d be wrong

one was written in r4 for exactly this purpose.

it was ran for several weeks in a high speed pa version on crumble.pa.com ( vish`s machine )

main use was to test the new middleware and the server optimisations when handling 180k players on limited resources.

now depending on if vish took a copy of the code (knowing him he did) depends on how quickly it could be implemented.

It would obviously need adapting to races etc for account creation but appart from that if they can find where it was put then it wouldnt be that hard to get it working and padd out the universe

yes and no

I like the idea in priciple, but not so many bot planets that its hard to move for bots tyvm

Perhaps 1 gal per Cluster sort of thing 2 at a max, and ofc the bots would need to go to sleep for 6-8 hours each day ofc

[SpaceLegend]

Originally posted by hinch
to those of you that thinks it needs coding you`d be wrong

one was written in r4 for exactly this purpose.

this thingie surely had no "intelligence".
that bot hadn't to make decisions like
"who do i attack"
"who do i defend"
"do i pull or not"
"which ships do i build now"

the time and efford for creating such a bot is better spend on improving the game and such a bot never solves the fundamental flaws of the game.

[hinch]

"who do i attack" - yes within a certain set of score/size parameters

"who do i defend" - planets randomly groupped them selfs into 100 planet alliances defence went to galaxy and alliance.

"do i pull or not" - didnt do this once launch was done it always landed

"which ships do i build now" - built based on the best ships to buy ie: told to before hand by us so concentrated on the relevent ship at each step of the tech tree

[hyfe]

Originally posted by hinch

"which ships do i build now" - built based on the best ships to buy ie: told to before hand by us so concentrated on the relevent ship at each step of the tech tree

Could be based on averages from the players too though..They have access to the database; and the strain would be low; as it needn't be done too often.

I still think its the wrong way to go though..

[Telhilion]

Originally posted by Tactitus
I think new players really stand a chance of having a good enough round to want to stay -in effect-

I'd hate to come into the game now. Bot padding would detract some of the flak and perhaps keep them round long enough to tally up some nice contacts and allies.

Posting on PD, must be ill again.

[Ultimate Newbie]

I quite like the idea, and i think i mentioned it myself a couple of rounds ago.

Granted, as Tac said, its re-arranging the deck chairs on a sinking ship, but before it sinks we as a community would have alot more fun. We (generally) play PA for fun as well as fighting with (and/or against) old mates.

The introduction of AI planets would go a long way to keep the game interesting for people who are not at the pinnicle of the players (such as myself, and i suspect a couple of thousand of other people).

Or, we could go one step further and exapnd on the idea that somone above said about humans vs. the AI - perhaps make it part of the storyline (heh, what storyline ), and make the human players must resist a powerful and determined Bot Army. if you are outnumbered 4 to 1 by reasonably intelligent Bots then it should rpove interesting at least.

THus the AI may not solely be farms.

anyway, i'll go see if i can dig up some of the old posts, but i dont think i have them anymore.

[TeK]

Morden, there would be plenty of targets if the alliance blocks were done right.

[-o|DarkJedi|o-]

Originally posted by Morden
now people might laugh at the topic but it would help the game a lot.

right basic outline of what i think would help the game a lot for this round and bring more players into the game:

[1]50'000 new accounts run by a bot controlled by the creators to increase possible targets for everyone.
[2]for those bots to have limited attack capabilities but so that they will attack people but not so much that they are overwelming for anyone.
[3] to defend in galaxy unless a full galaxy attack is launched on them, in which case cluster bot galaxies would defend.

^ that would help this round no end. there would be less noob bashing and a lot less people leave because theres no targets or that they get bashed every day.

otherwise we'll lose another 1000 people this round.

it aint hard to do fs so many peoiple have made bots over the round, you could always borrow a copy of the wac bot if you cant make your own.

spinner wanted to do this back in round 5

it should at least be tried....

[Ultimate Newbie]

hah! i found it.

It was the second column i had ever written - in R5 i think.


AI planets to solve PA problems?

Currently, there are a number of problems in Planetarion. One of the major problems is the shrinking universe, which began with R5 becoming Pay to Play. Fewer and fewer people are enjoying their PA experience at both ends of the scale – high ranking planets napped with everyone whom they can attack, leading to boredom, and the masses of planets that have waves of incoming 24/7, leading to frustration because they cant get anywhere. Many suggestions have been raised to solve this problem, however most of them include making PA free once more, which is clearly impossible without killing the game off faster than it is dying now. Few people will pay to be worked over constantly for four / five months again, which leads me to my suggestion.

My suggestion is to involve AI planets at all levels of Planetarion. They effectively take the role of the Newbie at the very base of the Roid Chain (ie food chain). AI planets would not need as much bandwidth or server load as a conventionally player, simply because they do not need to load all the silly little pictures (x4 now there are multiple races) that a human player needs to operate the game. However the AI’s wont just be tiny little planets with roids, there will be larger AI planets that can help sustain the higher levels of human players, and ultimately, if the round stagnates then a group of rather large AI’s will try and kill off the top 50 galaxies, for example. This would leave some galaxies open to attack from every else, like a Revolt type thing again – this would address some of the major issues above. It would eliminate the boredom (AI’s do not fjear your 1337 alliance, and they do not respond to threats ), AND it would give hope to the smaller people to 0wn a large planet before the round end (and they die :/), giving their meagre existence some value .

AI planet should not be their own little island. Depending on their scores, I think that AI’s should form alliances with other AI’s. for example, a 500k AI might be allied to another 500k AI for mutual defence, though a 5 mil AI might be allied to another 5 mil AI and 3 more 2.5 mil AI’s for mutual defence. This would make the AI’s running for the top spot more difficult to beat, creating a longer war with more fun . The idea behind this is so the AI planets don’t become free farms. Players who continuously attack a poor little farm over and over again (like they do now) requires no skill. At least with the AI being in a mini alliance, they have some sort of protection (however small). Clearly, AI’s shouldn’t be able to react right away at an event. For example, the AI might not know its under attack for a tick (while the server updates all the AI planets, just like we tick now), and help may not be received from an allied AI for another hour (for example), making it possible to attack these farms, but not for long, and not without risk (the AI might detect it right away, and get immediate defence, in which case you are screwed ). Possibly a limit should be imposed on how many times you can attack an AI planet (per week?) to minimise farming.

AI planets should not have ghey fleets, however they should vary from AI to AI. Eg some AI planets might prefer Harpies over Phoenixes, making them more difficult to roid by Cathaar planets while another might prefer Phoenixes because of superior killing power. Each planet would not be an individual, of course. The type of load and management that would be required from the servers to achieve that, would screw PA HQ rather fast. The fleet preferences might be in groups of 100 (or 1000) depending on how many AI’s will be introduced.

I would like to see a large amount of AI’s present in the universe. Not just a couple of hundred like the how the free accounts are being added to the universe, but rather at least half of the total people playing. This would boost total universal roids substantially and keep the action throughout the round. This is what PA needs imo. Keep away the boredom at the top, as well as fending away the loneliness and hopelessness at the bottom.

[Morden]

Originally posted by Ultimate Newbie

ace idea

now that would be very interresting and at least it would help smaller people stand a chance of having a planet without getting hit every time they get of x amount of roids.

[Fungi]

Hoo hah, another moot point...

[Dave]

It shouldnt be that hard, take PAOL for example.

That runs anywhere between 1k-10k bots at great spead.

And thats on my crappy old machine (p3)

[Starbucks]

its a definate idea for havok

imagine a whole cluster of super strong AI controlled players, awake 24/7

that will make havok fun

[InTGuRu]

Originally posted by Dave
It shouldnt be that hard, take PAOL for example.

That runs anywhere between 1k-10k bots at great spead.

And thats on my crappy old machine (p3)

And that's using how many db queries per second? 0?

[Ramihyn]

Originally posted by InTGuRu
And that's using how many db queries per second? 0?

optimise the db queries away

it shouldn't be a performance problem to control 20000 planets on a single machine - we are still speaking about a whole hour for a tick and since the bot controlled planets should imitate user behaviour, they aren't all online and active each tick. Also by far the most decisions are incredibly simple - there are only a few computationally complex decisions to be made in PA.

i even heard such bots existed
http://www.codersger.de/bot3.gif

but this suggestion has been done many times by now ...

and for maxing out at 1600 planets on a game with 1 hour ticks - tehehehehe

[The Globe]

SUPERB IDEA

[MrLobster]

Keeping it simple is the key, in PAOL I cant mimick human behaviour because the player him/her self is online all the time so the bots have to be too. I hope to create a more human like response bot in the next version of PAOL V.2003

More Soon....

I think bots could prove to hard to beat or too easy to predict...

[Dave]

Originally posted by InTGuRu
And that's using how many db queries per second? 0?

So do you select all the data for the bots, then you do all yoru calculation in memory, then you update the database with the new data.

You just need a box with sod loads of ram, which is doable .

[Ramihyn]

Originally posted by Dave
So do you select all the data for the bots, then you do all yoru calculation in memory, then you update the database with the new data.

You just need a box with sod loads of ram, which is doable .

Its not even that much ram - a 512 MB machine was able to easily hold all data (speaking of pa dump information) of every single tick of a PA round in RAM (including the 6 month rounds). With the current Ram Prices, i wouldn't care if it needed 1gig either

Speedpa clones are more demanding as some run for 45000 ticks or more :P

[[Smurf]Pies]

hmm i love this idea, as we all know pa was best with more planets, tho if these bots attack ppl, ppl would say creators are getting invloved in atatcks, but the idea of gals retaling if u attacked and maybe gal groups of computer bots who will help each other would be kwl. and the bots would randomise how they played/race/skill lvl, so u had diffrent types scores sizes.

i cant code so dunno how hard all this would be...

[Dave]

Originally posted by Ramihyn
Its not even that much ram - a 512 MB machine was able to easily hold all data (speaking of pa dump information) of every single tick of a PA round in RAM (including the 6 month rounds). With the current Ram Prices, i wouldn't care if it needed 1gig either

Speedpa clones are more demanding as some run for 45000 ticks or more :P

But it depends on how in depth you want the bots to be, and whether they will simply take actions at ticks, or throughout a tick etc.

[Woodeh]

Writing the bot wouldn't be hard, assuming it was plugged right into the database server. However, getting good AI that simulated a wide variety of player styles and mimicing patterns of people logging in etc. would take alot of work. You would then also have to work out alliances etc between bots again to mimic the way real people work.

IMHO

[beefeb]

sounds like a very good idea to me