experience tweak

[mist]

thinking about xp, and all its evilness , i was wondering...

roids produce resources for the whole round, so if you steal a roid right at the start it'll be doing your score a favour right untill the end (assuming noone nicks it, but if you're going to win you're not going to be having that happen too much, right?)

xp on the other hand, you get and then it sits there doing nothing...

this would seem to suggest that the way forward is to bash the hell out of people early on, when the roids will give more return than xp would, and then find worry about xp later on to boost you to the top. following logically, xp therefore seems to encourage newbie bashing at the start of the round, when it's allegedly at its worst anyway, and then works to discourage it as it naturally gets to be less of a problem.

personally, this rather seems like a bad thing.

so, how about either giving roids a limited lifespan - that produce for so many ticks and then they're useless (deals with the exponential growth problem nicely), or making it so that XP 'breeds' (makes it well worth getting early on).

-mist

[Methedrine]

What about having XP influencing combat? Basically the more experienced your fleet is, the more efficient it will fight...

[scotty1234]

but this would make it very difficult to gain extra xp,because if you attack someone with a better score than you chances are very high that you will lose and this would encourage people to attack people their own size and also encourag nOOb bashing

[Blacknova]

Originally Posted by Methedrine

What about having XP influencing combat? Basically the more experienced your fleet is, the more efficient it will fight...

The people who attack successfully first then get a massive advantage right the way throughout the round and results in player stratification.

That is only of any use if they reinstate XP for ships killed as opposed to roids capped to give people who defend their alliance a chance to grow on an equal footing.

Nova

[Tactitus]

XP is just a flawed concept, as it attempts to subjectively award points for "good" behavior (without ever rigorously defining/explaining what's good about it).

Irrespective of XP, it's always been the case that roids are most valuable early in the round (when they offer the most potential income). Roids become progressively less valuable as the round progresses, as their potential income decreases--finally to the point where you can't even expect to recoup the normal losses accrued in obtaining them. Hence, in the endgame, roiding becomes virtually pointless and the primary focus of attacking shifts to destroying opponents' fleets (and, of course, preserving your own).

Giving roids a finite lifetime might limit growth, but it won't change this aspect of game dynamics. Roids would still be most valuable early and least valuable in the endgame.

[mist]

giving roids a finite lifetime would alter the curve tho, so that instead of starting high and then dropping from day one it would be level untill the end of the round was going to happen before the rocks expired

however, this would mean that when you stole roids you wouldn't know what they'd be worth to you, so it would make attacking much more risky. for this reason it doesn't seem particularly ideal.

i agree tho, xp does suck, but spinner likes it, so it seems worthwhile looking at ways to make it suck less

-mist

[Tactitus]

Originally Posted by mist

giving roids a finite lifetime would alter the curve tho, so that instead of starting high and then dropping from day one it would be level untill the end of the round was going to happen before the rocks expired

I don't think it would be that level. First, what kind of lifetime were you thinking about? I was imagining something on the order of, say, 30 days (anything much shorter and attacks will have a hard time paying off at all). So, suppose a 30 day roid life. Everyone initiates roids the first several days of the round. Then they fight over those roids for a couple of weeks (some roids get initiated during this period, but not a lot compared to the total number in the universe). After that, everyone knows that most roids are about to expire so attacks taper off. Then roids start disappearing en masse. Then another couple days of initiation, etc. It seems like there'd be less attacking overall (either because you're at the beginning of a cycle and it's cheaper to initiate roids than to attack or because you're near the end of a cycle and roids aren't worth much because they're about to evaporate). Eventually (after a couple of cycles) everyone's roid life is spread out and the universe isn't so cyclical, but by then the round is over. :/

however, this would mean that when you stole roids you wouldn't know what they'd be worth to you, so it would make attacking much more risky. for this reason it doesn't seem particularly ideal.

One idea would be to reset the roid clock when it's stolen. That is, treat a stolen roid as a new roid. Not very "realistic" though and it would make roid-swapping insidious.

[ComradeRob]

Originally Posted by mist

this would seem to suggest that the way forward is to bash the hell out of people early on, when the roids will give more return than xp would, and then find worry about xp later on to boost you to the top. following logically, xp therefore seems to encourage newbie bashing at the start of the round, when it's allegedly at its worst anyway, and then works to discourage it as it naturally gets to be less of a problem.t

That's why we have the 40% limit. I couldn't bash a newbie even if I wanted to.

The idea of roids disappearing is also fatally flawed, for one simple reason. Many roids will be initiated in the first 3 days, but far fewer are initiated in the following days and weeks. This would lead to a situation where, 30 days (or whatever) after protection ends, a bunch of roids all start disappearing at once. The end result is a bunch of planets with a month's worth of fleets and no roids, making attacking impossible as nobody is worth hitting. The moment anyone tries to initiate their disappeared roids back, they will be swarmed by everyone else.

I might be alone in this, but I think the current xp system actually works pretty well, and doesn't need any major modifications :\

[mist]

tacticus

i was working on the assumption that there would be at least some newbies in the game who'd be continuously initiating. otherwise, you'd have a constant number of roids in the game after the first week and afaik that's not particuarly realistic

comraderob
i dislike experience in general. rather than modifying the game so that it plays the way the pateam/spinner want, they've added experience in an attempt to bribe people in to playing how they want them to. i'm gonna take a guess that there's no rankings by value in the game, ever wondered why?

-mist

[ComradeRob]

The idea of the never-ending supply of newbie roids doesn't sound plausible to me. It also only works if those newbies are repeatedly attacked, supplying their roids to the rest of the universe, which seems at odds with the idea of trying to protect them.

The biggest problem for PA rounds is that there is always a point at which most viable targets have so much fleet and so few roids (relative to fleet size) that attacking them is pointless. Disappearing roids would simply exacerbate this problem.

If the roid self-destruct timer were somehow reset when the roid is captured (as Tactitus suggested), then that problem would be mitigated somewhat, but there are still probably unintended consequences to work out.

I really don't have a problem with xp scoring. It does have real, positive effects on how people play the game, without any negative effects. If a genuinely superior system can be found then it should be used, but right now xp is here and it works. That's a major advantage over any other idea.

My final point is that, given the 40% value limit on attacks, I'm not sure that there is a problem that needs to be solved. Your point that a roid captured early on is worth more than a roid capture a week before the end of the round is true, but I don't see much of a problem with that. Also, at the start of the round there are no big and small planets - everyone starts equally. There isn't really any 'bashing' since everyone has tiny fleets, and any losses can easily be rebuilt in a day. It's only after a week that the 'top players', 'newbies' and so on become really apparent, by which time the 40% limit starts to kick in and protect the newbies. XP scoring also gives alliances an incentive to target galaxies with large planets in.

As for the ranking point, various PA tools sites offer value rankings. What I would like to see is a proper value rank within PA, and in particular listings of total and average XP for alliances.

[mist]

i'd be willing to put money on pa refusing to impliment a ranking based on value.

i don't particularly like the idea of destructing roids, mostly for the reasons you just mentioned. however, the 40% limit does annoy me - i dislike hard limits

as for XP causing no troubles. didn't it result in people breaking the servers with scanning last round?

-mist

[ComradeRob]

I didn't play last round, so I don't know for sure, but I'd assume that was a bug which has now been corrected, not a fundamental flaw in the concept of XP.

[mist]

afaik the definate rewards you get in the form of experience have never been published - increasing the luck factor in the game. this is imo a fundamental flaw in the implimentation of xp, which is rather the important thing if you're playing the game.

furthermore, there's always the danger that people will find more things like the scan exploit

-mist

[ComradeRob]

Only if there are bugs in the game though. Bugs are symptomatic of bad coding or bad design, but not of any inherent flaw with the idea of XP.

Most people understand how XP works pretty well - attack bigger planets for more XP. The incentive of XP gains for hitting bigger planets has worked well this round, and removing it would be a step backwards imo.

[Tactitus]

Originally Posted by mist

i was working on the assumption that there would be at least some newbies in the game who'd be continuously initiating. otherwise, you'd have a constant number of roids in the game after the first week and afaik that's not particuarly realistic

I'm sure there would be some roids being initiated continuously; but would there be a lot compared to the total number of roids in the universe? That seems unlikely to me.

If you get rid of exponential growth then you're left with linear growth, and as the coefficient decreases, linear growth starts looking a lot like zero growth.

Originally Posted by ComradeRob

I didn't play last round, so I don't know for sure, but I'd assume that was a bug which has now been corrected, not a fundamental flaw in the concept of XP.

I disagree. The fundamental flaw of XP is that players will always find ways to exploit the formulas in unanticipated ways. This is unavoidable when giving points using objective formulas to reward subjective actions.

[Gerbie]

Originally Posted by Tactitus

I disagree. The fundamental flaw of XP is that players will always find ways to exploit the formulas in unanticipated ways. This is unavoidable when giving points using objective formulas to reward subjective actions.

So far I haven't seen any exploits. Aparantly they were able to avoid it. The way to prevent unanticipated results is to keep the formula's simple and look at what they do in stead of just implementing it and wait how it turns out.

[Blacknova]

Originally Posted by Gerbie

So far I haven't seen any exploits. Aparantly they were able to avoid it. The way to prevent unanticipated results is to keep the formula's simple and look at what they do in stead of just implementing it and wait how it turns out.

Jerome demonstrated this in another thread, that by sitting back and deliberately allowing your targets to get roidfat, and grow to the point you can hit them for max XP, you can catapult up the rankings, the exponential factor has been removed to stop the most blatant excesses of this system, but the game essentially fosters a possible situation whereby a player could send a pure podfleet, escorted by a few killfleets, cap the roids off of a much bigger planet, and harvest XP at a phenomenal rate.

Nova

[mist]

the fact that they had to introduce an artificial cap to the amouunt of XP you get for hitting targets screams that the forumla itself is quite silly. something based around a 1-x^y curve, with x being a control variable and y being a relationship between the size of the planets would, assuming i remember some maths, give a naturally lowering return as you hit someone bigger than you. as they've got a striaght line i've got to assume the relationship is linear, which is naturally hard to control.

ofc, the system is utterly flawless.

-mist

[Gerbie]

Originally Posted by mist

the fact that they had to introduce an artificial cap to the amouunt of XP you get for hitting targets screams that the forumla itself is quite silly. something based around a 1-x^y curve, with x being a control variable and y being a relationship between the size of the planets would, assuming i remember some maths, give a naturally lowering return as you hit someone bigger than you. as they've got a striaght line i've got to assume the relationship is linear, which is naturally hard to control.

ofc, the system is utterly flawless.

-mist

That formula is flawed. There is a formula with the desired effect, but I don't think there's anyone in PA team that has the mathematical skills to design one (and comprehend the effects). The more complex the formula the harder it is to predict it's behaviour. Lines are hard to control as they have no limit. It's even worse than that as it's Target_value/your_value: the relation between score and your value therefore is: score ~1/your_value...

[Melesse]

I have no problem with experience, but I'd like to see people get some for defense. Would certainly encourage more people to keep a fleet around for that. My biggest problem in my gal is people are constantly attacking in the mad scramble for xp.

Melesse

I probably shouldent say this since I'm not 100% sure. But You do get xp for defending. If I remember correctly you get xp for killing shiøps no matter what but also gain xp for capping roids which is where the part of attacking larger planets come in.
Someone please correct me

[Kal]

I know the XP formulae, I think on the whole they are good, with a few minor issues. I personally think they should be public, but thats something for round 12.

You get XP for:
capping roids - based on a value ratio which is capped.
scanning - results in an oberall score loss
covert ops - more agents, more XP - only resource hack should result in a overall score gain (I think)
initiating roids - overall score loss, but growth potential

you do not get XP for anything else

[The Real Arfy]

whoa, so you mean that people who scan get a drop in score?

thats crazy!

[mist]

i think he means that those who scan loose score from resources, but gain XP. the loss from resources has to be greater than that gained in xp, otherwise you'd have people trying to pull the same trick as yogi et al did last round.

tbh, i don't think there should be any experience from scanning at all. the reward for scanning is the info you get, why give score for it as well?

-mist

[Tactitus]

Originally Posted by mist

tbh, i don't think there should be any experience from scanning at all. the reward for scanning is the info you get, why give score for it as well?

Heh. I feel the same way about the whole game. The reward for attacking is the roids you get; the reward for covert op-ing is the resources you steal, etc.

Down with XP!

[mist]

got to agree, however, the XP for scanning just seems to make scan planets more viable, and imo scan planets are one of the problems with the game

-mist

[Kal]

the idea behind XP in combat is to reward hitting people bigger than u and to make it less worthwhile hitting smaller people, which i think is a good intention at the very least

[Tactitus]

Originally Posted by Kal

the idea behind XP in combat is to reward hitting people bigger than u and to make it less worthwhile hitting smaller people, which i think is a good intention at the very least

Of course. And most of PAs good intentions have ended badly. :/

That was the same idea behind the roid/cov-op formulas based on size. That is, you get fewer or no roids attacking a planet that's too much smaller than your own (and more attacking a planet that's larger).

Two comments/questions:

1) If the formulas were effective at punishing bad behavior and/or rewarding good behavior, then why do we need XP?

2) If the formulas weren't effective, then why will XP work any better?

[ComradeRob]

XP works because it's a carrot rather than a stick. Forcing people to attack larger planets was always unpopular, and was often impossible. Since this round's ship stats favour attacking, it's much easier to hit someone your own size, with the added bonus of gaining extra XP for it.

XP was the reason 1up stayed ahead of LCH in score, despite LCH leading in roids for over a week.

[Tactitus]

Originally Posted by ComradeRob

XP works because it's a carrot rather than a stick.

I don't see why the formulas can't offer carrots in addition to sticks (e.g., a larger roid cap when attacking planets equal to or larger than your own).

XP was the reason 1up stayed ahead of LCH in score, despite LCH leading in roids for over a week.

Well then it must be good.

[Sinn]

Originally Posted by mist

got to agree, however, the XP for scanning just seems to make scan planets more viable, and imo scan planets are one of the problems with the game

-mist

Couldnt agree more. Battle XP system should be something like :


valueRatio = defenderValue/attackerValue

experiencePointsFleet = (defenderFleetLostValue - attackerFleetLostValue) / (1000 * valueRatio)

experiencePointsRoids = ((defenderValue - attackerValue) ^ (defenderRoids / attackerRoids)) / (1000 / valueRatio)

defenderExperience = defenderExperience - experiencePointsFleet
attackerExperience = attackerExperience + experiencePointsFleet + experiencePointsRoids


Obviously not balanced, but the general ideas :
- If fleet is lost (doesn matter on which side) the fleet owner loses XP
- Roiding a small planet and lose a lot of fleet may actually LOSE the attacker XP (crap attack, go find a bigger target!), the roids will make up for that somewhat
- XP gains decrease dramatically when defender size drops

I still think roiding should use a 25% cap, instead of the max cap being related to a value ratio.

[mist]

what's the max cap you can get when hitting a planet 40% your size?

-mist

[Sinn]

Originally Posted by mist

what's the max cap you can get when hitting a planet 40% your size?

-mist

25% initially, what you could do is add a 'recently hit' factor, the more you get hit within period X, the lower the max cap. As follows :


maxCap% = 25% - (20% / hits_last_50_ticks)


So, max cap would range between 5% and 25% depending in you having been pwned recently I would not, however, relate max cap to the value ratio of the battlng parties.

[mist]

why not?

the recently hit thingy is a nice idea, but kindof impractical. while it makes sense from a small planet's prespective it's rather pants to find that your target won't give you any roids through no fault of your own. it'd also make it rather hard to take down large planets

-mist

[Tactitus]

Originally Posted by mist

the recently hit thingy is a nice idea, but kindof impractical. while it makes sense from a small planet's prespective it's rather pants to find that your target won't give you any roids through no fault of your own. it'd also make it rather hard to take down large planets

It also seems really, really easy to abuse (e.g., allies "attacking" you with zero or one-pod fleets to jack up your recently-hit factor ).

[mist]

you mean people would abuse the system?

*shock*

-mist

[Sinn]

Originally Posted by mist

why not?

the recently hit thingy is a nice idea, but kindof impractical. while it makes sense from a small planet's prespective it's rather pants to find that your target won't give you any roids through no fault of your own. it'd also make it rather hard to take down large planets

-mist

Whot not? Quite simply because it'd remove easy roids from the game. If all roids are "hard to get" the round'll become stale in a hurry because deffing your own roids becomes cheaper than getting them at high losses.

As for the rest of your comment. If anything, it's too easy to take down large planets. People who spend 16 hours a day on the game and are t50 can be removed from the game with a single night of waves. After that they're borked because they can only hit planets with 40% of their value (which, in most cases, is still quite high due to remaining fleet). No problem you'd think since they usually still have their fleet. Problem is that they dont have access to easy roids anymore, so getting back to where they were is nearly impossible.

Originally Posted by Tacticus

It also seems really, really easy to abuse (e.g., allies "attacking" you with zero or one-pod fleets to jack up your recently-hit factor ).

Obviously, the recently hit factor should be related to actual losses, not number of atk fleets that land. Probably best to relate it to % of roids lost recently.

[mist]

given that you're a mil officer in one of the few decent alliances, i'll have to take your word for it.

however, isn't it hte same for everyone at the top? i was under the impression that once you can only reach the dizzying heights with a decent alliance behind you, and if this is the case, i assume the other planets up there will also have next to no roids?

-mist

[Sinn]

Originally Posted by mist

given that you're a mil officer in one of the few decent alliances, i'll have to take your word for it.

however, isn't it hte same for everyone at the top? i was under the impression that once you can only reach the dizzying heights with a decent alliance behind you, and if this is the case, i assume the other planets up there will also have next to no roids?

-mist

Well, i'm not opposed to big players being attacked and all, and i dont mind a 25% cap for the first wave, but being able to reduce a roid count from 5k to 200-ish in one night is a bit unfair. The only way he can get roids back (assuming he didnt lose his fleet, so his value is still very high) is to attack planets within his bash, which are still quite big. He'll lose more fleet than those 200 roids can repay, so he'll have to be more careful with attacks etc. etc. It's a safe bet he'll need a few weeks at least to get where he was, all because of one night of waves. But unfair. But the cap drop wil help smaller players as much or more, since they cant be raped all the time.

[mist]

how do you define 'all the time'?

i'd have thought that it was unprofitable to attack a newbie every night, as they'll very shortly have more fleet:roids than is worth attacking, so therefore you attack them and then leave them for a few days, then attack them again. this will still ruin the game for the newbie, but probably won't be covered by your cap reduction. unless you're going to give people a week of invulnerability when they're attacked?

-mist