Assuming a sandbox game environment that changes over time (by means understandable to the player), what's the right way to autolevel the game world's entities? I understand that auto-leveling is a feature easy to do wrong. If, for example, the world levels lock step with you just because you level it easily invalidates the whole concept of improvement. It can also create a powerful disincentive not to level in order to keep enemies at lower levels. What if, however, the world auto-levels unevenly, quickly in some areas and in others not much at all? The idea would mimic the development of nations / kingdoms, with some faring better than others and, by extension, having more or less to offer a leveling character. Some auto-leveling would be in response to player actions. Massacre a town, for instance, and the army might show up. But what constitutes an army would vary by region, with some offering ill-trained militia as a threat, others seasoned guard, maybe others mercenary squads that varied wildly in threat level. In a more linear game this would easily be a bad idea as you could not guarantee the challenges a player would face, and thus the experience. In a sandbox game, however, it would be up to you to fight or flee based on your estimation of your abilities. Thoughts?

I like the idea of different areas changing strength at different times, it contributes to the sense of a dynamic world. Maybe the dynamic world and the auto-leveling could be married even more prominently together.

Could you have the character's leveling cause the world's leveling as a byproduct? For instance, you "level up" through a distinguished career in one nation's military, and as a result the other factions in the area either ramp up their militaries to keep pace or are overthrown or supported by the ones that have kept up, ensuring that you'll always have worthy foes wherever you are. If you're a kickass space pirate and start beating everyone down, you'll have bounty hunters and other shady dudes out to get you, and even your buddies might consider selling you out for the paycheck.

That way, the world gets tougher as you get tougher, but the balance of power doesn't stay the same. It's not like the tanks from Country A get more HP and the tanks from country B get more HP, but rather Country A develops a new tank, possibly using data that you secured for them, and country B responds by beefing up its air superiority. If you take measures, as the player, to prevent country B from having the research, resources or infrastructure to upgrade its airplanes, then they just stay weak, country A invades them, and then a corrupt general goes rogue and starts a new junta there using country A's tanks.

Dragon Age Origins used this idea:

In each area, there are a range of enemies from very weak (green) to exceptionally strong (red).

While at a lower level, players will try to retreat from or avoid red enemies while taking on easy enemies. At a higher level, players can switch to hunting red enemies.

E.g. In the slums of a city, you might meet Petty Thieves (level 1), Pickpocket (level 2) etc or Bandits (level 5), Gang Leader (level 8). So there are plenty of action for players regardless of level.

The problem I have always had with auto-levelling is that if it is implemented simplistically then you have a situation where peaceful, starting point areas are suddenly full of thieves that could take out a small army (but not necessarily the player). Likewise 'the most dangerous dungeon (TM)' is full of really quite friendly monsters if the player is low level.

What I imagine works quite well is carefully tested auto-levelling maximum and minimum caps so that for example a peaceful town will be surrounded by level 1-5 thieves (even if the player is on level 20) and the more dangerous areas are full of monsters with a minimum level of say 15 so that they are not really accessible (or survivable) for low-level players.

It does mean that there will be a mismatch between the player and the characters he/she will encounter in certain areas but the advantages are that it adds a strong progressive element to the game (that can be linked to the storyline) as more dangerous areas open up as the player gets more powerful. It also helps maintain the level of game immersion with dangerous areas actually dangerous and common pick pockets not running round peaceful villages with top class knights armour and several magical swords. From experience I enjoy a game more where I can predict how difficult each area is and act accordingly, while there is nothing more annoying than knowing that whatever I do, however much training and levelling I complete, the monsters in a particularly tricky area will always be 5 levels more powerful than me.

To cut a long story short I would try (and am gradually trying) capped auto-levelling.

assume a world without the player in it. Characters will be leveling all the time, but the overall distribution of levels doesn't change. When one guy gets to level 10 some other guy of level 10 either leveled up to 11 or died. Every time someone dies a new level 1 guy takes his place.

Quote:

Original post by Iron Chef Carnage Could you have the character's leveling cause the world's leveling as a byproduct? For instance, you "level up" through a distinguished career in one nation's military, and as a result the other factions in the area either ramp up their militaries to keep pace or are overthrown or supported by the ones that have kept up, ensuring that you'll always have worthy foes wherever you are. If you're a kickass space pirate and start beating everyone down, you'll have bounty hunters and other shady dudes out to get you, and even your buddies might consider selling you out for the paycheck.

Nice! I've been thinking of stuff like the pirate / bounty hunter feedback loop, but had not really broadened it to ranking up in a faction. I like the idea that you could trigger an arms race or forge an alliance. What I think will be tricky is that the very idea of accomplishment suggests things go one way while auto-leveling seems to need them to go opposite.

For example, if you're a great general that has won wars you would rightly expect that you've made the game world safer and your country stronger. It might steal motivation to see a strong alliance rise as a result of you succeeding in a struggle.

But even if it's a paradox I think the novelty and feedback would make it worth doing.

Quote:

That way, the world gets tougher as you get tougher, but the balance of power doesn't stay the same. It's not like the tanks from Country A get more HP and the tanks from country B get more HP, but rather Country A develops a new tank, possibly using data that you secured for them, and country B responds by beefing up its air superiority. If you take measures, as the player, to prevent country B from having the research, resources or infrastructure to upgrade its airplanes, then they just stay weak, country A invades them, and then a corrupt general goes rogue and starts a new junta there using country A's tanks.

This sounds cool but I'm going to have to formalize it to make it feasible. Maybe it should be something like a series of possible impacts you can have at each level that ties into the global changes. At a lower rank you might get the missions that can, say, alter local balances of power but not screw up global stuff. As you rise, the gravity of choices becomes greater.

An example might be that of rescuing a leader. At a low level the leader won't be the president because they wouldn't logically give something so important to a grunt. At the highest level, though, it would be appropriate. The same might be true for a sabotage mission which weakens a country. At lower levels you're just eroding things so tougher NPCs can (abstractly) take them on, but at the top level you cause/win the war.

Quote:

Original post by Jon Alma The problem I have always had with auto-levelling is that if it is implemented simplistically then you have a situation where peaceful, starting point areas are suddenly full of thieves that could take out a small army (but not necessarily the player). Likewise 'the most dangerous dungeon (TM)' is full of really quite friendly monsters if the player is low level.

This issue still severely bugs me. The question I think that players end up thinking about (even if in the back of their minds) is whether the world revolves around them or whether the world is in place for them to meet/challenge it.

I'd prefer that if a region change you can somehow get an idea of why. If the streets are filling up with thieves, let it be because a mob boss has moved in, or the economy has crashed or something logical. Then at least maybe the player can do something about it if they choose.

Quote:

What I imagine works quite well is carefully tested auto-levelling maximum and minimum caps so that for example a peaceful town will be surrounded by level 1-5 thieves (even if the player is on level 20) and the more dangerous areas are full of monsters with a minimum level of say 15 so that they are not really accessible (or survivable) for low-level players.

I would be okay with this provided they're not the majority of places you go. Fixed locations can give you something to compare yourself with, which I think it essential to the idea of leveling.

The problem comes when you have to traverse those "expired" levels repeatedly. Morrowind comes to mind with its dramatic encounter music blaring and you, at level 50, turning to find that it's a lowly worm of bird that would have been scary at level 2 or 3 but is now just a nuisance.

One thing I might do is try to cook up a per region growth curve that can be modified by player actions. It would cause some places to always remain backwaters in terms of advancement while others always leveled faster than you.

(Although this really doesn't address the real issue, which is how to add valid gameplay in terms of high player level to low world level interactions. IOW, the great lord need to be able to have some meaningful interaction with what are now the peasants)

Quote:

Original post by Glak assume a world without the player in it. Characters will be leveling all the time, but the overall distribution of levels doesn't change. When one guy gets to level 10 some other guy of level 10 either leveled up to 11 or died. Every time someone dies a new level 1 guy takes his place.

This could be kind of funny in a sort of "single player virtual MMO" sort of way. You're leveling and the world's leveling around you, and maybe there's an element of keeping up.

It would be interesting to cast this not just in terms of heroes, but faction leaders. If the mob boss is somehow leveling while you're leveling there may be an element of strategic competition. This is sort of how Depths of Peril works, with NPC heroes even being able to steal quests and hirelings from you. I don't know if I really like the idea as it makes it hard to leisurely enjoy the game.

A more mission central approach might work better. You accomplish a mission, level and then the NPCs respond.

One problem with this is how do you get a sense of how strong/fast an NPC will level? Maybe it's again based on region. If you know a mob boss has taken over a backwater you won't expect him to level as fast as one that's based in a corrupt city.

Maybe some challenges come to you. Trade in expensive goods and attract the attention of "high level" pirates. Deal in narcotics and you're up against both superior narco police and competing drug gangs. Fly a cheap rust bucket grain transport and nobody gives you any grief but get an end of the line luxury ship and attract all kinds of attention (not all of it bad: easy to get audiences with potential business partners - but if your computers are hacked and while your sensors are blind half your cargo disappears because your computer expert isn't one that shouldn't be surprising).

The key is that the world doesn't react to your level so as to provide an adequate challenge. It reacts to your perceived level - representing strength and wealth invites tougher challenges, which the player may or may not be ready for, but which he has his own choices to thank for. So maybe it is possible to solve some crisis by showing up in an unmanned unarmed battleship and bluff-threatening everyone to back off or else.

But I guess this is pretty much what you guys have been saying so far, so... carry on. : )

Quote:

Original post by Wavinator Assuming a sandbox game environment that changes over time (by means understandable to the player), what's the right way to autolevel the game world's entities? I understand that auto-leveling is a feature easy to do wrong. If, for example, the world levels lock step with you just because you level it easily invalidates the whole concept of improvement.

That's the definition of auto-leveling. Which is why auto-leveling pretty much just sucks.

Quote:

It can also create a powerful disincentive not to level in order to keep enemies at lower levels. What if, however, the world auto-levels unevenly, quickly in some areas and in others not much at all? The idea would mimic the development of nations / kingdoms, with some faring better than others and, by extension, having more or less to offer a leveling character. Some auto-leveling would be in response to player actions. Massacre a town, for instance, and the army might show up. But what constitutes an army would vary by region, with some offering ill-trained militia as a threat, others seasoned guard, maybe others mercenary squads that varied wildly in threat level.

None of this is auto-leveling. Just leveling or powering up.