Hello everyone … well, whomever may still read this. Activity has been low those last few months, and it seems as if only recently we had some new players. Yeah, it’s the same old song – developers being busy irl and so on.
We’re still lacking a pixel artist, I’ve not heard a lot from Sheck those last few months, probably because I, myself, was busy as hell, first failing, then coping with having failed, then trying to get back on top and… guess what – yeah, I’ve made it. I have survived, I’m still here, I’m still twitching. I’m okay, don’t worry.
Thing is, I’m going to start a full time job in november, and thus my time will be severly reduced – again. I’ll have to see if I can get back to writing up stuff some of these days. Maybe a bit of feedback would be a motivator. Well, I’ll have to see.
Well, we did chat the other night, but it wasn’t long, and yeah, there was a huge pause. (Always keep in mind: I’m just a volunteer, like you, and mostly, my contact with Sheck is via IM)
Anyways, currently, I’m mostly theorycrafting subconsciously as to how the game could be made more appealing in general. I’ve seen quite a few good games that have an active audience without being graphical blockbusters or anything. PoA definitely has the potential for that as well.
If I manage to lay down some sort of document and present it to Sheck… it may have a good set of deep changes, and I may have to argue in favor of a complete wipe (as if anyone would really notice, currently) but… I do have a few ideas.
Ah, ok. Yeah, I know you are a volunteer like me and the rest of us, but it seems you are always the one on top of things moreso than others.
Good to hear that you have some ideas thought up. A complete wipe, if necessary, shouldn’t be a problem. I’m sure most of the players who’ve been playing for a long time now wouldn’t mind, as they’ve (like myself) have completed pretty much everything there is to complete in-game. Hopefully we can get back on track soon with improving Aethora.
the locked exits at the beginning have to go, if you cant find the right quest you cant get anywhere. even though i am adept at this game, my new parties still get trapped in hopitan when i cant find the quest to unlock the west gate.
the ai needs a massive overhaul on what abilities to use, spending all their action points buffing rather than attacking, and using crap abilities rather than “double blade attack,” or other powerful abilities makes encounters overly easy and gameplay therefore boring
there needs to be some endgame content, some superhard zone comparable to hell difficulty in diablo 2, that gives people that have conquered the midgame something to strive for. a dungeon off of hope, or a gateway into another plane to fight the shadow creatures would do nicely
the southern areas have almost no quest content currently as well
there are a lot of unnecessary city areas in cities other than hopitan that make finding quests hard
abilities need better documentation on what they actually do, and some need to be eliminated because they are pointless. or reworked. melee abilities which push instead of damage? no sane person would use that ability.
it took me about 30 hours of gameplay before i figured out what abilities like “acrid smoke” actually did
also, i have detected no noticeable difference between using high damage weapons vs low damage weapons. damage/hit seems mostly based on your ability level relative to their passive ability level, with the weapon not a significant factor. therefore, basic fists are the best low level weapon, with speed 70, for 2 hits/round. “Upgrading” to iron knuckles would lower damage/round by 50%, since you go down to 1 hit/round.
my high level party only cares about what ability is used, and weapon speed. so i have 2 guys that do 2handed strike with a staff, which costs 18-19 action points per hit, 1 guy with weighted bludgeon, 1 guy that just casts performance boost 2times per round for extra action points. I can get 9 hits per round with this setup.
the problem is that given weapon damage is insignificant to actual damage, which invalidates 99% of weaponry.
1) Locked exits. Noted. Main problem may be that people will get lost in a sandbox-world and not find missions at all, and/or run into foes they can’t possibly defeat. We have to do more than abolish locked exits, we need to have a more clear distribution of difficulty. Likewise, a change in missions that makes the requirements easier to meet, which has been in planning for a while, should help alleviate the problems caused. Moreover, perhaps a notification when missions are available? But I think that’d take away too much from the exploring aspect.
2) Yeah, AI overhaul. Planned, down to pretty clear lines along it may happen. But it’s coding. And it’s not easy. And it’ll still be AIs.
3+4) Endgame content is a pretty moot point, sorry. It’s content that takes exactly as much, if not more work to create, will be played by a significantly smaller group, and likely not be played for much longer. Why? Because barely anyone will play it if it doesn’t offer feasible rewards, and because those rewards will make it easier in turn, and this will make that high-end content boring after a while, as well. And sorry, I’m working 10 hours a day, 6 days a week – so an endless supply of ever-newer, ever-higher, ever-harder endgame content is out of the question. Southern areas are our endgame, currently. A lot of people never make it to the southern areas anyways, and I’m sure no promise of quests to be done will change that.
5) If we’re going by necessity, we could simply make a list of encounters you want to fight and not even bother with a map at all. Or with quests. We need a world that is at least somewhat believable, and we need nodes to be able to base our quests on. The missions are, in part, about exploration, and a lot of people would find it pretty boring if everything was easy to find. I think we should have more “spading” in the game, as this seems to be a very fun activity to quite a few people. Anyways: If we only made spots as necessary, it’d boil down to one node per town, with a list of quests to check, or a network like we have, with each one pretty much being good for one or two quests, and once we add something new, players would instantly recognize it by a new area having been added.
6+7) Abilities are being reworked and rebalanced completely. That’s been on the list for months, now. Thing is: We want and need to encourage tactical combat, not maximizing raw damage output. About documentation… that’s hard. Because we don’t want to give too much information, either. Information should be short and small, and that means that if we detail everything a skill does, you get a mathematical figure to describe something that is supposed to be a combat action in a steampunk/fantasy game.
8-11) Yes. Items. That’s my biggest reason for suggesting a wipe. We have an increase – especially in speed – that is too fast, too great. And speed is much much more useful than strength. Likewise for armor. Some solutions include: Make damage/armor more prevalent in damage calculation, reduce drop rates of items generally, lower the total speeds, especially the increase in speeds on weapons… everything already discussed.
Now a few things to add:
A) We need an active, stable community. New players register, log on, write hello in chat. 48 hours later, a reply comes, and the new player has stopped caring 36 hours ago. We need means to encourage or enforce players to be active in game regularly. In most games, the community you experience in every day life and game, is made up of normal players. The developers can’t be animators in chat and forum.
B) We need a multiplayer-experience. It’s an online game, and yet, the only things we have is chat, forum and a market to trade. We also need some sort of ingame-economy, to make trading interesting and worthwile. Again: Reducing drop rates and introducing interesting rare drops may help. But we need more, possibly Coop-play, or even PvP.
C) We need activities. Stuff that doesn’t take as much work as a mission to create, but that will keep players occupied a lot like a mission would do. Repeatable missions could help, such as a regular tournament in an arena, with named specialized groups of enemy fighters to compete against in a K.O.-System. Or an event system, altering complete areas according to variables that are changed, such as elemental surges, bandit uprisings, void rifts…
D) We need to make combat more engaging. It needs more depth. The main questions should not be “Who does how many damage how often per turn?” and “Will this ability increase the damage/frequency of attacks of this guy enough to warrant using it over an ability that does damage itself?”. Solutions could be partial resistances or immunities, requiring specific attacks, strong defense abilities that can be broken, combo attacks or situation-specific/positional attacks that are very powerful, area-changing attacks on large maps (for example freezing a body of water to pass over it), and generally attacks that vary more in what they do and what shape they do it than how much they do. This would make more actions valid. Interesting could also be lasting area-based attacks, special abilities like invisibility, burrowing, or summoning helpers, or stances that completely change how a character is able to act (different base regenerations, different actions possible). Skills that are learned as unlocks could also make everything more interesting.
The list could go on. (More possibilities to compete with other players than simple ability ranks, possibilities for some sort of premium features to create revenue…)
Problem: Most of this isn’t as simple as it sounds. It’s easy enough to find problems (players don’t get attached to the game very easily), it’s a lot harder to find the cause (there is no active community to speak of), to conceptualize a solution (how can we encourage players to be online more often/regularly, and actually talk to new players) and then to implement it.
Smote – tons of great feedback there. Thank you very much for your insights.
I really wish I had more time to work on the game, but unfortunately, Life has a way of intruding. I work a full time job as well, and I have several other side projects. That being acknowledged, I don’t want to just abandon Aethora altogether. Sometime in the next couple of months I hope to release the source-code under an open source license.
I just want to take a second and address a couple of those issues you discussed. As Kran said, we recognized that the variety of battle actions got out of control, and many of them are duplicative. We worked on outlining what types of strengths and weakness each skill should have. We worked up a spreadsheet that would distribute actions in a balanced way across multiple levels; right now there are a few actions in the live game that require a certain level before you see them – in the new system, there are only a small number of actions that do not have the level requirement. This creates a setting where every so many levels in a skill (or combination of skills) you discover a new action, rather than getting hit with a bunch of actions that you don’t understand all at once. We put a lot of effort into the re-balancing of actions and got maybe 80-90% done, but then lost our momentum.
Anyway, reorganizing the actions has another benefit; there is some new code in the development branch so that when the AI goes to choose an action, it gathers up the list of possible actions (ruling out any actions that would be useless – in other words, don’t put on a Fire-defense buff if none of the PCs are wielding Fire-based weapons; also don’t reapply an effect that is already on your target – that kind of stuff). From the list of possible actions, it chooses one randomly, but in the new code, the randomization is weighted, so that higher level actions are chosen more often. With the new re-balance, we’re ensuring that higher level actions are actually good, useful actions, and therefore they are the actions the AI would most likely want to use. For example (in the new system): “slash with blade”, the Long Blades action is level 0 and “double-blade attack” is level 20. When weighting, we give anything at level 0 a weight of 10. So in this case, the weight of “double-blade attack” works out to a 67% chance that the AI will choose that action and only 33% chance that it will choose a standard “slash with blade” attack.
About the “strength” rating of weapons and the “defense” rating of armor: You are precisely correct – the formula used on the live server favors skill levels over weapon/armor stats so heavily that it makes the weapon strength (for attacking) and armor defense (for defending) practically useless. This was an attempt by me to establish that in this RPG setting, skill level is more important than equipment, and it failed, because it was too far out of balance. In the development code, this has been changed drastically and armor and strength of weapons actually mean something. They weigh in as equally as skill offense/defense, however – the formula caps the influence of armor and weapons so that it doesn’t outweigh the influence of skill levels. In other words, if you are working with a really high strength longsword but your skill level in Long Blades is really low, you’re not going to get the full benefit of the weapon’s strength. And in just about any situation, Armor is going to be much more effective than it was before.
However, as Kran said – in order to roll this new code out with the best effect, it would make most sense to do a complete wipe. Which is why we haven’t done it yet – we haven’t committed to such a wipe.
And then there is also the problem of not having any new content – sure, we can balance the game mechanics, but if we still have the same areas that are not interesting (as you pointed out), then it’s pointless. Those extra cities outside of Hopitan were intended to be something a lot more special – they would have their own missions and things like that, but unfortunately, we have not had the time to design and build all those missions.
I would recommend specifying monster’ by type: support, attacker, range.
each has its own unique AI. support only would use high quality buffs/debuffs, like performance boost or defense raisers, or movement boost to a range unit. it would also use heals, if those are implemented.
if it turns out that passive boosts like “Strengthen” actually do something in the new code, then they could use that type of lower quality buff (basing qualitative quality rating on current system).
support would also use debuffs, preferably only using effective ones: like heavy boots and muddy mind, as opposed to ones that waste action points
melee units would not use any buffs at all, unless they are considering ending their turns at >75 action points. a melee unit should try to guard the ranged units/support, and stay slightly in front of them in formation. only when enemy is close, or ranged, will they charge and attack, and only with full action points so they can preferably do 3+ attacks. Also, if they are in active combat with someone, active combat defined by “was hit last round”, they will charge and attack. Also, they will attack if any of their comrades are being attacked.
The goal of this new ai is to get enemies to fight as a unit, rather than running in and dying one by one.
ranged units would use ranged abilities and actually move away from the party after attacking. ai would be as follows:
If enemy distant:
Movement <= 50 If ally is being attacked, move into attack range and attack. Else wait.
Movement > 75, action > 50. Move up to 50 movement points toward enemy, if such movement will put ranged unit in range. Then attack. Then retreat 50 movement points.
If enemy close:
Move to max range, away from enemy, and attack. If near edge of map try to find new route away from enemy.
If you want to get clever you can include some “chokepoint ai” where the ai actually uses map to its advantage. chokepoint detection/use does sound tricky to program, though
AI should also concentrate fire, as it does now, but not to the point where it ignores a potential target in favor of trying to get to an unreachable target. AI should retreat all units, if a single unit is near death, with rest of AI shielding injured unit’s retreat.
A possible implementation on ranged attacks could also be to only allow them with Line of sight.
I understand that endgame content has limited application to most players. But this game has an anticlimactic ending, where the top tier zone has u fighting farmers in farmland. Its just plain depressing, it needs an epic ending/dungeon.
Also, I see no reason for monsters that can double blade attack to ever use “slash with blade.” If someone was fighting with 2 weapons and they slashed with one, they would inevitably do something with the other as well, making it basically a double blade attack.
If implemented properly, the aforementioned ai changes could make Aethora much more interesting and tactically challenging in combat. If combat turns out to be too difficult, just lower skill levels to compensate, then there is that much more headroom for later advancement, and more challenging upper lvl encounters without the need to design a bunch of new content.
by the way, in regards to wandering off and dying, punishment for death in aethora is not that bad. Wandering into a hard zone and getting killed by high level monsters just gives players a future goal for self improvement. Meanwhile feeling trapped and frustrated make players not have fun and quit.
Just a few notes:
Line of sight: Considered it, nigh impossible to realize properly on cartesian coordinates. Dropped.
Endgame: Current is farmland, correct, with voids/farmers alternating. My idea would be to have only one group of them available at a time, depending on a “global state” that’s affected by player actions and some underlying routines. Before farmland, endgame was Deep Plateney. After farmland, it will be something else. Content comes slowly, but it does come.
AI: Easy to say what it should do, hard to implement it. Plus: If it is too systematic, it becomes exploitable again. I’d also rather have AI that can adapt to player choices to a degree – having specialized teams, again, leads to a race to find out how to abuse the given weakness most efficiently.
Actions (as mentioned in AI): The goal can NOT be to make the AI avoid “useless” actions, it’s avoid having those actions in the first place, preferably by having only useful actions, and not only by damage numbers. What’s the point of tactical combat if the player doesn’t have an actual choice besides “Use the best action” or “Use a less efficient version of the best action”? Could as well have random number generators competing for the higher value.
I agree with you. Every action should have a purpose, a situation that it is useful for. A player’s thought during combat should be to identify the situation and pick the best ability for it.
So lets say every action does have a perfect situation, where it is better than every other action, given those circumstances. Then the AI should have an algorithm that figures out its situation, then finds the best action, and does it. Since any situation should be possible, any action should be possible.
The assumption that every action is the best action for a set of circumstances is not currently true. Currently it is always better to double blade attack than to slash with blade. (Though I guess this assumes relatively equal blade speed.)
AI: Yes, it is hard. But challenging programming is the most fulfilling and enjoyable to actually program. It is much more fun to create a masterpiece than repetitively do the mundane.
Systematic does not mean exploitable. Chess AIs are systematic, and they have completely surpassed humans. Aethora AIs can be systematic and non-exploitable as well, if well-programmed.
One can also program a flagging system, such that any battle with comparable skill-levels between AI and human where a human wins in less than 4 rounds is flagged for view, to see how the human won so easily, and program a counter-strategy for the AI for that circumstance.
Plus, just because an upgrade might have some flaws, does not mean that it should be avoided. While it might be exploitable in the future, given that we are likely imperfect programmers, I can promise you that the current system is already very exploitable, and it would still be an improvement.
For LoS, you could make a simple one that is D&D style. If a line from the center of your square to the center of another square passes through 2 opposite edges of a square, then if that square is occupied line of sight is blocked. Lines passing along a main diagonal (corner to corner) are also declared as blocking.
For endgame, yeah, current is farmland, which is really weird. Why are some of the toughest enemies farmers? That doesn’t make sense! They’re farmers!!
The issue with not rolling out an update is not that we don’t know if it was an improvement, it is just horribly incomplete, and the people working on it barely have any time to do so. As for everything that’s coding – LOS (It does have appeal, but it would be very technical), AI specifics, flagging, whatever – that’s stuff you’ll have to discuss with Sheck. I’m not a coder or programmer. Whenever I write something, it’s either using what knowledge I have and applying common sense, or reiterating what we discussed as developers.
I can tell you one thing, though: Chess is a highly complex game, but one with a set and relatively even battlefield. The thing here is that the computer always has the same setup of pieces, more or less, but the human player can vary and adapt to it.
The farmland is high-end because it’s where we wanted to put the voids, which are the actual end-game foes, currently. ;)
Kranador, I could possibly try to make better images than the ones before. However, that was actually the first time I’ve EVER tried to make images myself, so I’m not sure if it’ll work. And, just to clarify, is this the only thing I could do to better Aethora currently? Just lemme know.
(I would also love you guys forever if PVP was implemeneted sometime in the future, heh.)
Currently, it’s the only thing that I know nobody who could take care of it, yes. You can also try to help with the action rebalance, over on the development server, but I don’t know how much there is to be done, currently.
As for PvP: I’m more of a coop PvE person myself, but I know the appeal of any form of direct player interaction (and for PvP, it also automatically becomes a viable endgame) – but difficulty-wise, it’s way up there with AI coding. Complex, time-consuming etc.