My ideal dungeons: Radial Dungeons v2.0

Rungar

Radial dungeons

In this model players usually start in a central hub and the adventure extends from that central hub out into many arms. Rather than a linear trash-miniboss-trash-boss structure we instead use the central hub and task oriented global objectives. This is best shown by example.

Dead by dawn haunted castle

Players start this dungeon at midnight and have to work their way through a zombie infested courtyard to a broken down door. Once they enter the door to the central area they can repair that door.  The zombies will keep attacking the door for the duration of the dungeon and players must either keep the door repaired or fight the never ending zombies. Players would have to search the grounds for the wood required to repair/brace the door. The castle has 3 floors: Upper floor, main floor, and the basement. These floors are filled with undead enemies, traps and spawners to deal with.

Once inside there would be a helper npc to explain the dungeon and a little time to formulate a strategy as there would be a number of simultaneous tasks to complete. All tasks link back to this main room and keeping the door repaired and surviving until morning is the first task.

The second task is to find, bring back, and keep alive the survivors who are scattered randomly (i.e. different each time you play) throughout the castle.

The third task is to find a certain book hidden randomly (different each time you play) somewhere in the castle.

The fourth task to find the vampire lord boss (or possibly other boss), who is in the basement (always in the same place, not necessarily the same boss), and defeat him.

There may be other optional tasks and not all tasks may present each time you run the dungeon.

 

Players then have until dawn to complete all these tasks and finish the dungeon to the best of their ability. Since all the tasks are simultaneous players need to figure out how to get it all done with the resources they have and will be rewarded for their ability to do.

Reward conditions/global objectives:

The more tasks players can complete the better the rewards. None of the tasks are particularly hard but doing them all at the same time can be challenging. 

Wooden chest: At least one player survives until dawn.

Iron chest: one player survives and at least 5 survivors are saved or the book is found or the vampire boss is defeated.

Silver chest: one player survives and any two of the 5 survivors, book or boss.

Gold chest: one player survives and 5 survivors, the book, and the boss.

Platinum chest: one player, 10 survivors, the book and the boss.

Gem encrusted chest:  all players survive, all survivors saved, the book and the boss.    

If all players die the dungeon ends in failure and will have to try again. This is to move away from the “throw yourself at it until you win” choreography style game play.

So not only can players lose a dungeon but winning a dungeon is completely based on their skill & ability to work as a team to complete the tasks. This moves some (but not all) of the focus off of purely twitch based combat to a teamwork model. Players may have to split up to complete all these tasks and use non-combat methods like sneaking, swimming, climbing etc to avoid problems rather than just ram through everything.

The variable reward is the same as an integrated dynamic difficulty selector. Using global objectives, dungeons such as these can be modified at any time to include additional optional tasks to keep the dungeons fresh. They need not be static since you can employ a number of optional tasks that may or may not appear each time you run the dungeon. So sometimes you might get the "book" other times you might get something else. Other times the book might be a second boss. The optional objective model allows for a degree of variability that can be added to over time to keep it interesting.  

The surprise visitor: this is an enemy from another dungeon or place in the world that players might be familiar with that can randomly show up in the dungeon. If you manage to kill this character you get a chance for random loot from where this visitor has come from. This is over and above the objective/rewards structure from above and is completely random and may not happen at all. The visitor can spawn anywhere and at any time or not at all.

 

This is just one example of how a radial dungeon is superior to traditional linear dungeons. The goal is to get players talking to each other again and avoid rushing to the end which is pretty common in modern game dungeons. This also tones down the choreography aspect a little bit as the main boss is not necessarily the main focus of the dungeon anymore (though it could be). Once the tool for creating objectives and rewards is made the dungeons need not be as static anymore which means they can evolve with the game since it would be pretty easy to add new optional tasks on the fly and or remove obsolete ones. 

More Game Modes

To get the most out of assets created for dungeons they should have up to three unique variants of each setting. This also opens the door for multi-group dungeons where players can be either competitive or cooperative and other game modes such as solo or duo. Using the situational mitigation and a smaller group size (such as three) you could have dungeons where three different groups complete the dungeon at the same time and are rewarded based on the global objectives which can include additional parameters for the game mode.

If a dungeon had 3 simultaneous wings, 3 groups could run this at the same time but independently. Alternatively 3 groups could compete against each other. Using the global objective/reward system it’s just a simple addition to add tasks/rewards for these unique modes. It would be fun if you could see the other groups but not necessarily directly interact with them.

Three but variable possible

I prefer a group size of three because what’s meant for 3 is also challenging solo or duo players and efficient use of content. The situational mitigation model from my other posts also allows more dynamic groups in smaller numbers since your current role is dependent on the enemy you face, not a selection when you make your character. The group size of three is also more scalable when you consider content such as competitive, cooperative or other modes and three is also the easiest group to assemble in today’s modern casual atmosphere where everyone wants to solo. I think making this accommodation will bridge the gap so to speak such that content generation is as efficient as possible and will reach the maximum number of players for the longest duration.

However this type of system can accommodate more or less players than 3. There may be special conditions added if you use too few or too many players but you could allow it to do so to keep it flexible. For instance adding a 4th player might be possible but you might not be able to complete the all best chests. Likewise you can do the dungeons with less people but might not be able to complete all the tasks which is quite reasonable. You don’t necessarily need 3 players to do any one task but you only have so much time to complete them all so you could theoretically solo them but it might not be easy or efficient and you might not qualify for the best rewards. You can do it though.  

Quizzical

One worry with that dungeon design is that it would turn into players finding the most efficient way to systematically sweep through the entire dungeon so that you find everything, no matter where it is hidden.  Or are the locations of the book and what not visible on your map from the start, rather than having to run around to find them?

If it becomes an issue of, you'll likely "succeed" in some sense, but it's difficult to get the maximum success, then that could create some real challenge that is lacking in a lot of games.  You could get a triage case of, we can definitely get X, but if we try to get both X and Y, we might get neither.  So do we try it?

Another concern is that time pressure tends to heavily favor aggressive, offensive gameplay and devalue defensive or support.  A crowd control character who could enable the group to beat tougher groups of mobs than another damage dealer could be undesired if his lack of damage output means that the group can't clear everything in time.

But don't take this the wrong way.  I'm not here to say that your idea is stupid.  I'm glad to see interesting, new ideas proposed, and a lot of possible objections are solvable.  We might yet manage to turn this forum into a one about MMORPGs rather than one about NFTs.

Rungar

Thanks. 

By design there are random and static components. You cant have all random components but you need a few to keep people on their toes. 

the variable reward system allows lesser skilled players to win at their own skill level which will promote them trying again for a better reward. It also allows smaller groups or even solo to do the dungeons. 

For instance the only critical requirement in this example is to  survive until dawn. A solo player can do that. They will only get the basic reward but they can do it. 

the simultaneous task strategy also promotes an any means necessary (within the game rules) kind of thinking. Fighting will not always be the best strategy. You have to complete the tasks, not clear the dungeon. Sometimes avoidance will be the better strategy. Sometimes you'll need all three people go complete the task and other times you might just need one. As you say if you want the higher rewards you have to take risks. 

not all dungeons would be timed. Its just an example. Other dungeons might have other failure conditions. 

i made my own trinity for this model because the classical tank/dps/healer model is not well suited for this kind of gameplay. I use a situational mitigation dynamic role model which is basically your character is defined by defense. So mages can tank other mages, fighters can tank other fighters. This will lets players be more multidimensional while still being specialists which I think would translate to a better overall experience once players wrap their head around it.   

sometimes they will have to do the face tanking, sometimes the killing and sometimes the crowd control. I don't have a healing role. It is not necessary for this model combined with action combat. Instead players will have to rely on their archetype defenses and play to their strengths.  

With this system knowing you enemy is critical. Once you know your enemy you also know your role or when to run.