XP formula ;p

[XelNaga]

Well, all ideas here have a lot of disadvantages (abuse, more imbalanced, etc etc etc), but how about this:

Stop the XP cap at 10. You simply don't get more XP from hitting bigger planets, you only get less if you hit smaller planets. This would make the xp whoring completely useless, since it wouldn't help at all to keep your own value down.

OWNED?!?

[Cannon_Fodder]

bravery factor is fine, maybe increased buts thats just me.

only change i would make is change score = xp*60 + value to xp*50

[Nadar]

Originally Posted by XelNaga

Well, all ideas here have a lot of disadvantages (abuse, more imbalanced, etc etc etc), but how about this:

Stop the XP cap at 10. You simply don't get more XP from hitting bigger planets, you only get less if you hit smaller planets. This would make the xp whoring completely useless, since it wouldn't help at all to keep your own value down.

OWNED?!?

No.

[furball]

No. People should be rewarded for taking on gambles rather than going for easy roids.

[Synthetic_Sid]

Originally Posted by furball

No. People should be rewarded for taking on gambles rather than going for easy roids.

I think the issue this round is that certain attack fleets work against planets substantially larger than yourself without there being any "gamble". Obvious examples being zik fi vs terran, xan fr vs terran and terran bs vs cath. In the first 2 cases the initiatives are such that it's not feasible for the target to defend vs the atatcker AT ALL themselves without spending lots of res on ships that are basically useless. In the latter case the issue isn't quite so bad - as at least BWs are useful for attacking with and alliance defence.

The sum total of the relevant stats is that it's not possible for a terran to play an old-style value/roid-based planet - but they're an ideal race to xp-whore with. And caths are left with difficulty doing well as EITHER an xp or value based planet.

There's no "gamble" in any of the attacks mentioned above - other than the "gamble" of whether it gets defence or not. And whether a planet gets defence or not is not related to the planet's value. Hence attacking a large terran/cath with appropriate ships is no more of a gamble than hitting a small one.

[furball]

I was making a more generalistic comment as opposed to one focused on this round. Certainly in the past it was more of a gamble to attack a planet larger than you, given the increased possibilities of defending against the majority of attack fleets. The specialisation that we see of current planets certainly would not be reflected in earlier PaX-era rounds.

Without XP rewarding upwards attacking, fewer people would attack the bigger value players because they're more likely to have enough defensive ships. Team-ups on these players wouldn't be so rewarding due to a lack of XP for the attackers.

The gamble is not the alliance defence - the gamble's based on the target's ships. Any CR Cath can hit an Xan with only a few ghosts. If they do that, then any ghosts in defence will probably be frozen as well, as the Tarantulas have plenty of spare freezing power. In contrast, the Cath could hit a higher-value (and probably higher-roid) planet, freezing all their ghosts - but not leaving much room for extra Ghost defence.

This is the gamble - and without upwards XP, many players just wouldn't bother taking it on.

[Banned]

In my opinion, Sid's assessment is fair and mostly correct. However, I think that looking back on the last few PAX rounds, there have always been similar situations. One of the reasons this round stands out so much is that so many people have chosen that as their primary route. Not only are there 3 alliances with a majority of XP players, but demotivated value players are suiciding their fleets in order to be able to join 'the fun'. And since value players rely on other value players to gain large amounts of XP, this has only exaggerated the situation.

Other reasons are the new alliance size limit, the stats for this round*, the amount of time investment required to play for value etc.

The one that stands out the most for me, however, is the combat engine. I don't see how this engine does anything but encourage the situation Sid pointed out (where some races have almost free roids on certain races). Last round it was Tzen vs Vsh, and I'd like to point out that this round the damage ratios are a lot better for the defense, but the problem remains. As praised as single targeting is, combined with the radical differences between the four races it leads toward an inevitable imbalance. I don't think it's fair that you have to choose race X to have a chance at winning**. Or that there's an inherent advantage in choosing the minority race.

Since Kal left PAteam they've closed the development process much more, so I don't know what they have planned for next round, but I hope that an improvement to the combat engine is in the works, rather than a bunch of old school game modules. If we wanted a PA clone we'd all be playing ********** or ********.

I have some vague ideas about how I think XP would be improved that include some sort of diminishing returns. The current formula make the differences between hitting someone at the bashlimit and someone 2x value too large.

* Which are more attacking than I would have liked. In hindsight I know a few changes I would've made, but hindsight is easy.

** Cath in r12, 14 and 15, Zik in r13.

[Synthetic_Sid]

Hind-sight's a wonderful thing when it comes to stats. Until PA team decide to tweak current stats instead of doing a total rewrite of them each round I doubt we'll ever see properly "balanced" ones - and each round will remain effectively a "play-test" of a new set of stats with the results of that test then being ignored to try out a totally new set the following round.

[ComradeRob]

I'm pretty sure* that back when I started playing (in round 3) Planetarion was officially described as being in 'beta', with the implication that some final version would one day be arrived at. Has it ever officially come out of beta testing?

* It was over 5 years ago so I may be wrong

[Appocomaster]

Originally Posted by XelNaga

Well, all ideas here have a lot of disadvantages (abuse, more imbalanced, etc etc etc), but how about this:

Stop the XP cap at 10. You simply don't get more XP from hitting bigger planets, you only get less if you hit smaller planets. This would make the xp whoring completely useless, since it wouldn't help at all to keep your own value down.

OWNED?!?

Yes it would. It always helps to keep your value down. It means that others get less XP from you, if anything, and you get more XP, if anything. It's an interesting idea for solving the problem, but it takes away the whole idea of bravery. It does have it's merits, and would certainly stop the problem, but I'm not sure it's the right solution.

Originally Posted by sniborp

bravery factor is fine, maybe increased buts thats just me.

only change i would make is change score = xp*60 + value to xp*50

That'd just change the equilibrium point at which XP wh0ring was viable slightly. I don't think it'd help much, to be honest. Most people wouldn't even realise, knowing how few people read the manual and the changelog

Originally Posted by Synthetic_Sid

The sum total of the relevant stats is that it's not possible for a terran to play an old-style value/roid-based planet - but they're an ideal race to xp-whore with. And caths are left with difficulty doing well as EITHER an xp or value based planet.

and the ease at which defence is possible, surely?

Originally Posted by Syn_Sid

There's no "gamble" in any of the attacks mentioned above - other than the "gamble" of whether it gets defence or not. And whether a planet gets defence or not is not related to the planet's value. Hence attacking a large terran/cath with appropriate ships is no more of a gamble than hitting a small one.

using lots of ifs and maybes, you can connect a planet with higher value to a bigger alliance to a higher chance of defence.

Originally Posted by Banned

In my opinion, Sid's assessment is fair and mostly correct. However, I think that looking back on the last few PAX rounds, there have always been similar situations. One of the reasons this round stands out so much is that so many people have chosen that as their primary route. Not only are there 3 alliances with a majority of XP players, but demotivated value players are suiciding their fleets in order to be able to join 'the fun'. And since value players rely on other value players to gain large amounts of XP, this has only exaggerated the situation.

Yes, XP is almost a victim of it's own success. As more people switch, there's fewer targets for the rest. But once almost everyone is gone, there's no one to gain XP from.

Originally Posted by Banned

Other reasons are the new alliance size limit, the stats for this round*, the amount of time investment required to play for value etc.

why the new alliance size?

The one that stands out the most for me, however, is the combat engine. I don't see how this engine does anything but encourage the situation Sid pointed out (where some races have almost free roids on certain races). Last round it was Tzen vs Vsh, and I'd like to point out that this round the damage ratios are a lot better for the defense, but the problem remains.

This round apparently galaxy defence is more viable than alliance defence.
The combat engine doesn't help, but having a T2 on all ships at full damage would cause a lot of overtargetting and push things too far the other way.
Last round I did break Xan, but it could have probably been avoided.

As praised as single targeting is, combined with the radical differences between the four races it leads toward an inevitable imbalance. I don't think it's fair that you have to choose race X to have a chance at winning**. Or that there's an inherent advantage in choosing the minority race.

Yes, Cath have generally been possibly too strong as I tried to make them viable without XP whoring, which means they're generally winners at the top end.

Since Kal left PAteam they've closed the development process much more, so I don't know what they have planned for next round, but I hope that an improvement to the combat engine is in the works, rather than a bunch of old school game modules. If we wanted a PA clone we'd all be playing ********** or ********.

We've got a few new coders, and though sending the NDAs and getting them recieved by Jolt has taken more time than it should have done (and is still not complete in all cases) we're planning on making a far more functional, multipurpose combat engine that should be able to replicate every round but R10 which had the disturbance stuff.

I have some vague ideas about how I think XP would be improved that include some sort of diminishing returns. The current formula make the differences between hitting someone at the bashlimit and someone 2x value too large.

the general change I was considering was something of a similar form to before, score = value + y*XP^n, 0<1<n.
basically, it'd give more score for an increase in XP when you had less, but as you got more you'd need more and more to make the same increase. I'm just knocking together a few graphs in excel to upload to demonstrate the difference.

[Banned]

Originally Posted by Appocomaster

why the new alliance size?

With fewer members, it's less likely that there's someone with the correct type of defense available, and fewer of them available in total, so the same amount of incoming that might have been coverable last round, will be uncoverable this round.

The combat engine doesn't help, but having a T2 on all ships at full damage would cause a lot of overtargetting and push things too far the other way.

I never advocated that.

Last round I did break Xan, but it could have probably been avoided.

Yes. I think that if I'd been motivated to work off those instead of making new stats we might see balanced stats this round. Unfortunately it's difficult when the two stats sets are made by people with somewhat differing visions

Yes, Cath have generally been possibly too strong as I tried to make them viable without XP whoring, which means they're generally winners at the top end.

I don't think I'd say too strong. They have to be that way to be playable. People who don't get big planets with Cath get continual incoming, but the good planets eventually get enough ships to block out most attackers.

the general change I was considering was something of a similar form to before, score = value + y*XP^n, 0<1<n.
basically, it'd give more score for an increase in XP when you had less, but as you got more you'd need more and more to make the same increase. I'm just knocking together a few graphs in excel to upload to demonstrate the difference.

I don't think that adjusting the score formula is the way to go. I assume you meant 0<n<1?

[Appocomaster]

Originally Posted by Banned

Yes. I think that if I'd been motivated to work off those instead of making new stats we might see balanced stats this round. Unfortunately it's difficult when the two stats sets are made by people with somewhat differing visions

well ... *shrugs*
my R14 stats were meant to be the basis for my R15 stats. They changed quite a lot.

I don't think that adjusting the score formula is the way to go. I assume you meant 0<n<1?

good guess
http://www.appocomaster.co.uk/XPScore2.htm has the current setup, score = value + 720*xp^3/4, and score = value + 8500*xp^1/2 (ok, I can *see* how large the scores will be early on with initiating roids, but still...).
Both are equalised so that they all give the same value at ~ 20k xp, and after that the fall off starts. Most of the top planets this round have ~100k xp, but I doubt they'd get anywhere near as high in a new round.
The only worry with this sort of thing is that Zik have no limitations, and so mid-end round could possibly win. Zik have always been a late round race though...

The other issue which I forgot to comment on was the idea of diminishing returns (I turned my computer off 3/4 of the way through my post, so hit post as it was rebooting).

Blame it on my maths degree, but the basic idea was taking the "bravery factor" to be 10*sqrt(targetvalue/yourvalue), or even more extremely sqrt(10*targetvalue/yourvalue). not sure about the capping, but that'd surely reduce XP, even if we capped it at 20 still.

edit: corrected the values, the first one was *slightly* incorrect. my memory is going

[Appocomaster]

Originally Posted by ComradeRob

I'm pretty sure* that back when I started playing (in round 3) Planetarion was officially described as being in 'beta', with the implication that some final version would one day be arrived at. Has it ever officially come out of beta testing?

* It was over 5 years ago so I may be wrong

No.

[XelNaga]

Originally Posted by Appocomaster

Yes it would. It always helps to keep your value down. It means that others get less XP from you, if anything, and you get more XP, if anything. It's an interesting idea for solving the problem, but it takes away the whole idea of bravery. It does have it's merits, and would certainly stop the problem, but I'm not sure it's the right solution.

What most people here seem to forget is that if you attack a larger planet (higher value that is), he is very likely to have more roids than a smaller planet (low value that is).

So you already have the benefit 1) of getting more roids and 2) of getting more XP since you already get more roids. It would still be interesting to attack bigger players, as it always was, even before XP was introduced.

But what WOULD change, is that you don't need to keep your value down artificially, so you can get more XP. As long as there are planets from your size, you get the maximum cap and it would be useless to be smaller. And if there are none, that means you are number one, or are amongst the best, well then a little disadvantage certainly will help to keep the game interesting.

This way, there wouldn't be any reason to adjust the XP formula, as XP whoring as a whole, attacking only bigger planets to get the max cap that is, would be rendered useless, and XP would have the effect it was supposed to have: encourage the normal, oldschool attacking, while still trying to keep the fleet alive and value up.

And everyone with a sole no just demonstrated that they didn't even think about the consequences for more than a second.

I'd really be interested to see some, if any, founded reason why this would have any negative effect? What probably irritates most people is that it just doesn't sound right, but that doesn't mean it can't work out right.

[Gate]

In my opinion, XP is very good. It encourages attacking, and helps keeps rounds fluid. It's a great bonus, but it is an annoyance to other players when someone with 300k value ends up packing 8 million score. It's getting to a point in some ways that value play is worthless and so we could see more pure XP alliances in the future, and the end of wars as: what's the point?

The only suggestion I can think of to solve this, is to keep XP as it is, but put a score cap based on your value. For every 1 score you get from value, you can get, for example, 5 from XP. You have all the XP all the time, but you only see the score bonus when you have the value.

So someone with 300k value can have a maximum score of 1.8 million for example. But if they have the equivalent of 6 million score in XP, every time their value goes up by a point, their actual score goes up by 6 points until they catch up. This would lead to value players having a far more consistant rank, whilst XP would be far more volatile and could perhaps be more interesting.

[cypher]

Originally Posted by Gate

In my opinion, XP is very good. It encourages attacking, and helps keeps rounds fluid. It's a great bonus, but it is an annoyance to other players when someone with 300k value ends up packing 8 million score. It's getting to a point in some ways that value play is worthless and so we could see more pure XP alliances in the future, and the end of wars as: what's the point?

The only suggestion I can think of to solve this, is to keep XP as it is, but put a score cap based on your value. For every 1 score you get from value, you can get, for example, 5 from XP. You have all the XP all the time, but you only see the score bonus when you have the value.

So someone with 300k value can have a maximum score of 1.8 million for example. But if they have the equivalent of 6 million score in XP, every time their value goes up by a point, their actual score goes up by 6 points until they catch up. This would lead to value players having a far more consistant rank, whilst XP would be far more volatile and could perhaps be more interesting.

imo this is better then it was atleast... however i think if people stack some res (or we get back to being donated in the end) then suddenly they can get 20 mil score in 1 tick by just spending in the end or by capping loads of fleet...

But atleast SOMETHING has to be done if pa wants to stay a game with active players and fun wars and all that sort of stuff..

i still think making xp only 33% or 50% of what it is now would be the easiest and fastest solution.

[Heartless]

Originally Posted by XelNaga

What most people here seem to forget is that if you attack a larger planet (higher value that is), he is very likely to have more roids than a smaller planet (low value that is).

So you already have the benefit 1) of getting more roids and 2) of getting more XP since you already get more roids. It would still be interesting to attack bigger players, as it always was, even before XP was introduced.

But what WOULD change, is that you don't need to keep your value down artificially, so you can get more XP. As long as there are planets from your size, you get the maximum cap and it would be useless to be smaller. And if there are none, that means you are number one, or are amongst the best, well then a little disadvantage certainly will help to keep the game interesting.

This way, there wouldn't be any reason to adjust the XP formula, as XP whoring as a whole, attacking only bigger planets to get the max cap that is, would be rendered useless, and XP would have the effect it was supposed to have: encourage the normal, oldschool attacking, while still trying to keep the fleet alive and value up.

And everyone with a sole no just demonstrated that they didn't even think about the consequences for more than a second.

I'd really be interested to see some, if any, founded reason why this would have any negative effect? What probably irritates most people is that it just doesn't sound right, but that doesn't mean it can't work out right.

Capping XP at 10 is pointless as that wouldn't reward hitting planets bigger than yours (you get 10 xp per roid when you hit a planet exactly your size). Therefore, capping xp at 10 per roid is non-sense and you'd hunt planets your value, but not planets bigger your value. In return this would also benefit those people which prefer hitting planets smaller their value as the people which are more encouraged get less advantage from their courage.

All in all I am personally more a fan of appocomaster's first suggested formula. The effect of gaining less experience per roid the more experience you have in total is pretty logical - after all, experience is what you gain from doing things, but the more often you do something specific the less experience it grants you.

That combined with a different set of stats should be more than sufficient.

[cypher]

i agree with that suggestion from appoco, although i'm worried you're gonna boost the "base" XP first..

[XelNaga]

Originally Posted by Heartless

Capping XP at 10 is pointless as that wouldn't reward hitting planets bigger than yours (you get 10 xp per roid when you hit a planet exactly your size). Therefore, capping xp at 10 per roid is non-sense and you'd hunt planets your value, but not planets bigger your value. In return this would also benefit those people which prefer hitting planets smaller their value as the people which are more encouraged get less advantage from their courage.

All in all I am personally more a fan of appocomaster's first suggested formula. The effect of gaining less experience per roid the more experience you have in total is pretty logical - after all, experience is what you gain from doing things, but the more often you do something specific the less experience it grants you.

That combined with a different set of stats should be more than sufficient.

It would still reward hitting bigger planets. Re-read my post, that's the whole point I made. And Appocomaster's suggestion won't put an end to XP whoring.

[furball]

No, it wouldn't reward hitting bigger planets. It would only reward hitting planets your own value or slightly more. To quote from your own post, as you asked:

Originally Posted by XelNaga

You simply don't get more XP from hitting bigger planets, you only get less if you hit smaller planets.

Originally Posted by XelNaga

And Appocomaster's suggestion won't put an end to XP whoring.

We don't want or need to put an end to XP whoring, just to stop the massive reward that Terrans had this round from doing it! And that's solvable by a stats change anyway (which I still advocate).

[Heartless]

Originally Posted by furball

No, it wouldn't reward hitting bigger planets. It would only reward hitting planets your own value or slightly more. To quote from your own post, as you asked:



We don't want or need to put an end to XP whoring, just to stop the massive reward that Terrans had this round from doing it! And that's solvable by a stats change anyway (which I still advocate).

Thanks for rubbing it into his eyes furball

[Appocomaster]

Originally Posted by cypher

i agree with that suggestion from appoco, although i'm worried you're gonna boost the "base" XP first..

Oh totally, we'll make xp a zillion times bigger.
XP will be worth more if we do change the formula (720*xp^3/4, as I corrected it above), with it being worth 4x as much score at 100XP, 2x @ ~2000XP, and so on, as my table shows.
If players want to XP whore to 100k XP (as some are at, or higher than, currently) then it'll be worth 33% less. By 200k, it's closer to 40% less.

I think that it still makes XP a viable route, but by mid-end round it's not high enough to compete at the top level. This means that for a new or average planet who's not so active, playing for XP might work. For bigger and higher ranking planets, playing for XP is pretty doomed by the end of the round.

[XelNaga]

Well well I guess both ways would work, but in any way, xp whoring shouldn't be an option if you want to play in the top ranks...everyone who disagrees is an xp whore in the top ranks .

[Heartless]

Originally Posted by XelNaga

Well well I guess both ways would work, but in any way, xp whoring shouldn't be an option if you want to play in the top ranks...everyone who disagrees is an xp whore in the top ranks .

Oh right. And next step is that cheaters shouldn't be allowed to play at all because you shouldn't be able to get a top rank by cheating, and everyone who disagrees is a cheater in the top ranks.

Logic can be such a pain in the ass sometimes.

[jerome]

Originally Posted by XelNaga

Well well I guess both ways would work, but in any way, xp whoring shouldn't be an option if you want to play in the top ranks

why not?


PS dear god you must be shit at planetarion, how embarrassing.

[JonnyBGood]

Originally Posted by XelNaga

Well well I guess both ways would work, but in any way, xp whoring shouldn't be an option if you want to play in the top ranks...everyone who disagrees is an xp whore in the top ranks .

Learn to think.

Originally Posted by Heartless

All in all I am personally more a fan of appocomaster's first suggested formula. The effect of gaining less experience per roid the more experience you have in total is pretty logical - after all, experience is what you gain from doing things, but the more often you do something specific the less experience it grants you.

General sentiments of approval. This could need some shifting about for value bases, ie you get diminishing returns for roiding consistently at 400-500k value, a different set of diminishing returns for 600-700k and so on and so forth upwards. This counteracts the keeping value deliberately low tactic rather nicely and yet you can still get great xp if you pull off a fantastic roiding when you're at a pretty high value.

[XelNaga]

Originally Posted by Heartless

Oh right. And next step is that cheaters shouldn't be allowed to play at all because you shouldn't be able to get a top rank by cheating, and everyone who disagrees is a cheater in the top ranks.

Logic can be such a pain in the ass sometimes.

Who else would find that OK...well maybe those in the lower ranks, too.

[SpaceMonkey]

Just a suggestion, I think I've seen something similar before...

bravery factor = 10 * (target value / your value) - 5

possibly (re)move the bash limit and reduce max bravery to 15?

The idea being that XP still works, although not as well, and value players can retal at a price.