I think this is usually the starting point for many many people trying to make their first level for multplayer. Specially when it's arena style. I have to admit that I when I first tried to make levels this is what I was thinking, multiple paths with no distinction between them. I could even argue this is related to "artist's block" because you have a space but have no idea how to fill it. Are you trying to do too much at once? Or cannot focus on any single point? I'm offering some insight in an attempt to solve the problem of having too many routes.

One temptation is to think that more routes equals more possibilities. For a multiplayer map it can mean less strategy. How so? It's about focus. If we are talking about an objective-oriented gameplay, having too many routes in and out creates too many openings. Too many positions to defend. Having fewer routes forces the remaining routes to be more meaningful and having less possibilities makes the gameplay more focused on stronger tactics. In turn, even the more experienced player can play to the level's strenghts. Too many routes makes each tactic weaker because you are forcibly spreading the players among too many routes.

What is the main idea behind this layout? It's about having multiple choices and freedom to go everywhere. Players can choose to go lef, right, up, down, whichever they like. Sounds great, right? Yeah, it feels enticing to have the freedom to go this way or the other way. However, this is poor design. You are not focusing on any particular point. First the whole design is symmetrical as if all areas are equally important and equally accessible. With a symmetrical design like this there is no incentive to play strategically. Experienced players would want to play to their own strenghts in a level, according to what the level brings to them. 

Let's break up the symmetry a little. Now we have larger squares and smaller squares. With this we have created shorter corridors and longer corridors. Some narrower spaces and larger spaces. That's enough to bring in some tactics to the gameplay. 

Now that we have an assymetric design, let's add two points. For now I'm skipping distances or tactical placement.

What we are doing is adding strategic elements that make the level have more depth, more gameplay elements, more strategy. I can make a comparison with magic cards. Newbie players tend to make choices based on the appearance of cards. They judge the power of the cards by the most eye-catching properties, one of which are numbers because it's easy to say that 10 > 5. However, in magic we have many more variables which adds up to the complexity of the game. A card that does more damage may have drawbacks, costs more to be cast and/or only be able to deal damage under specific scenarios. From the point of view of designers, the card brings that complexity when it's intended to be played by more experienced players. How would this same concept translate to level design? I'd say that an inexperienced designer would think much like an inexperienced player. 

A poorly designed arena would bring up a lack of focus and lack of depth, much like an inexperienced magic player would not have the experience required to see the complexity behind some cards and some strategies. The first image in this page shows how having everything symmetrical is poor design because there is no focus on any single point. In terms of magic how can we have this very same issue? One example: an inexperienced player would likely think that if half the cards in his or her deck are offense and the other half are defense, 50/50 would make a perfectly balanced deck. Not so much. In magic there is never one strategy wins all. They (Wizards of the Coast) have different types of cards, spells, colors and they can't make every type of card, color and strategy perfectly balanced. Even if by some miracle they could achieve it, you can't expect that the player base is going to be spreaded evenly among all types of strategies and decks. In the case of a symmetric level it's natural to expect that the action is going to be spreaded evenly among all the space in the level. Probably not, because players have human brains they are naturally going to use some routes more often than others.

Suppose that this design represents a CTF level and A and B are the opposite bases. We have 5 routes and each one has a different shape. Another mistake that I must have made very often back then was to think "I want to give more choices and each choice is tied to some different playing style". Now we have too many routes combined with the desire to embrace as many playing styles as possible at the same time. Say that we want one route to be shorter but more dangerous or risky. Another to be longe but safer. Another to require more jumping and running skills. Let's do some math, multiply the numbers (of routes) and end up with 15 or 25 combinations! That's one of the ways I mistakenly though about making levels back then. Mathematics is a tool but not to be used like that! Less is often more. In the figure above I'm not even counting intersections between different routes!

How does Magic handles different playing styles and routes? In magic's domain what we have is different formats. First, the decks have a certain number of cards, with 60 cards being the minimum and no maximum size. Second, we have multiple formats and each format have its own pool of cards. Some formats allow for a larger pool of cards to choose from than others. Most players, even the beginners, know that they can't fit everything into one deck. They have to choose some strategy and a limited pool of cards because there isn't room to have multiple strategies in the same deck. In level design this means that one level can't embrace every type of player at the same time. However, a game comes with multiple levels and each level is going to embrace some subset of players. Going further and yes, a game can very well suffer from trying to embrace everyone at the same time. This is some discussion left for another page at another time.

I don't have much to say about how to handle this in terms of design. I'd say that if you have a lot of ideas, one of the best tools is to write them down. When you look at a list of ideas you are probably going ot notice that some are more complex, some are larger, some are harder to accomplish, some should take more time to make, etc. That's the starting point. You are probably going to notice connections between them and some connections are going to feel more interesting than others. Some other connections are going to feel less enticing or not much appealling. Citing Mark Rosewater's own lesson about restrictions breeding creativity. Can you spot which ideas are essential to one level? Can you highlight which ones can be left out? To give some practical exemple: a CTF level focused on vehicles wouldn't play very well if it tried to embrace narrow corridors and many different traps to account for many different types of players.