Ideal game map size

[awergh]

This is both a game design question as well as a question of what various different games have done.

What would be considered the ideal or typical level\map size for a game.
For my purposes I am considering a game divided up into rooms\screens (mode 0) and counting the total number of rooms. This works best in a side on or top down game without scrolling but I'm sure it would be interesting to know how big a game can be for an isometric perspective.

Looking at maps on cpcpower of a selection of different games I have this list (hopefully I didn't miscount).
20 Manic Minor
30 Caves of Doom
32 Impossible Mission
48 Monty on the Run
57 Technician Ted The Chip Factory
59 Jet Set Willy
60 Jarlac
79 Auf Wiedersehen Monty
85 Goody

I ask the question because I am currently thinking about the next game to write and it is helpful to think about how big the game should be. 
Note: This is a comparison between a game made up of 13 single screen levels and a game made up of the 9 (3x3) screen levels.
13 Fitzroy and the Missing Chair
81 Fitzroy Dives Deep v1.0
90 Fitzroy Dives Deep v1.2\v1.3

It is surprising to me that the final version of last years game is bigger then all the games I listed earlier (there must be some bigger games though). I'm generally of the opinion that any memory not used is wasted so you should have as many levels as you can possibly fit. The only unfortunate thing is that my last game was a little bit too difficult so while it had 81 screens (9x3x3) I suspect most people wouldn't have got through most of the game. (At least it wasn't as hard as Profanation 2 (never did get off the first screen ))

[Nworc]

I found 180 screens per level in Turrican quite comfortable, while 1008 screens in Nonterraqueous were a bit over the top.

[SharkusMaximus]

To take a phrase usually associated with another context, it's really not the size but what you do with it that matters. It's all very well having a huge map but if there's nothing for the player to do in half the locations or there's no variety in the gameplay or visuals, then why bother? The context of your game is also important to consider, as the size of the map needs to make sense in terms of its environment. If it's set in a haunted mansion, then it would make more sense for it to be fairly compact. However, a vast, sprawling space station is conceptually believable.

One game which I think gets it bang on in terms of size and gameplay is Super Robin Hood, which is fun to explore, with lots to do on each screen, but which doesn't outstay its welcome by being too big. In contrast, something like the second Magic Knight game is packed with empty screens which serve little purpose and which contain nothing to do apart from walk from one side to the other. Starquake is fun but I've always found it far too massive with too little variety, something which makes it intimidating and which only the most dedicated player is likely to complete.

[andycadley]

Yes, context is everything. I remember an interview with the developers of Burnout Paradise where they mentioned that the proportions of the city layout are all entirely ridiculous, but it works so much better as a game when there are long straights and sweeping bends. It would be a terrible map if you were instead walking around the city.

[reidrac]

As said, the type of game and the context matters.

IMHO, having lots of screens only makes sense if the game has a maze component. I know I have worked hard in my games to get more and more screens on 64k, but it was because Golden Tail was short (again, in my opinion) and I wanted to push further my 64K & single-load games.

In those cases, around 64 screens is the sweet spot for me.

[SharkusMaximus]

Pacing is also an important consideration. If you're going for something fast and action-packed, then multiple short stages work better to give the player some kind of respite, even if it is only a few seconds, as it's hard to maintain a constant fast pace for a long time. But for something slower where exploration or puzzle solving is more the focus, large maps make more sense as they give a greater feeling of freedom. A good example of the former approach is Equinox, while the Dizzy games typify the latter.

[awergh]

I think I should have expected that the answer is that it depends but you can only figure these things out if you ask the question  .

To take a phrase usually associated with another context, it's really not the size but what you do with it that matters. It's all very well having a huge map but if there's nothing for the player to do in half the locations or there's no variety in the gameplay or visuals, then why bother? The context of your game is also important to consider, as the size of the map needs to make sense in terms of its environment. If it's set in a haunted mansion, then it would make more sense for it to be fairly compact. However, a vast, sprawling space station is conceptually believable.

One game which I think gets it bang on in terms of size and gameplay is Super Robin Hood, which is fun to explore, with lots to do on each screen, but which doesn't outstay its welcome by being too big. In contrast, something like the second Magic Knight game is packed with empty screens which serve little purpose and which contain nothing to do apart from walk from one side to the other. Starquake is fun but I've always found it far too massive with too little variety, something which makes it intimidating and which only the most dedicated player is likely to complete.

Making sure rooms are not empty and that every part of the level has a purpose is a good point. I would have to admit that Fitzroy in the Dark had mostly empty levels (not that you could see them    ) and while I had an interesting concept it never felt super compelling to play.

I found 180 screens per level in Turrican quite comfortable, while 1008 screens in Nonterraqueous were a bit over the top.

I suspect without ever having played Nonterraqueous that it repeats rooms or reskins them to get its clearly massive size .

As said, the type of game and the context matters.

IMHO, having lots of screens only makes sense if the game has a maze component. I know I have worked hard in my games to get more and more screens on 64k, but it was because Golden Tail was short (again, in my opinion) and I wanted to push further my 64K & single-load games.

In those cases, around 64 screens is the sweet spot for me.

Having a maze component would fit the game that I am thinking about but I think I could achieve the player getting lost without having to design a maze. I'm not sure a maze is the most exciting type of level to design or play. There is probably a fine balance between confusing the player just enough but not too much that they get frustrated.

Pacing is also an important consideration. If you're going for something fast and action-packed, then multiple short stages work better to give the player some kind of respite, even if it is only a few seconds, as it's hard to maintain a constant fast pace for a long time. But for something slower where exploration or puzzle solving is more the focus, large maps make more sense as they give a greater feeling of freedom. A good example of the former approach is Equinox, while the Dizzy games typify the latter.

For my purposes I am probably going to be more puzzle then action but I don't know until I actually start building levels how it will play.
I suspect that some of the game will benefit from taking your time and other sections will force the game against the clock.
Of course the concept I have now may be nothing like the game I submit in November but that's OK, if I can flesh out my ideas earlier it will hopefully be easier to build .

[Sykobee (Briggsy)]

Mazes can be quite compact if you wrap them or have teleport functionality to move around the maze when hitting the screen edge. Other useful things, depending on the game, perhaps there is a logical way to solve them, or clues available elsewhere in the game.

A Manic Miner screen is very dense, functionally, and the player is small. Other games with bigger characters would make each screen a multi-screen experience but with the same amount of gameplay.

Some games are massive just for the sake of it. Times Of Lore was a bazillion screens, but in the end the action happened in a few select areas with a lot of walking, although that did add a sense of scale and continuity to the game that worked in that case.