"Sandbox" campaign style and maps without plans

I'm looking for advice and input before my campaign gets going.  In the last year and a half, the gaming I've been doing has introduced me to new games and styles of play, and really made me think about the whole "railroad" vs. "sandbox" issue in gaming.  Just to explain, railroad play means that the GM encourages a more linear play and directs the players along a storyline.  Sandbox play is open, allowing the players to go about the world and make their own choices.  Obviously it's a continuum, and there's more to it, but that's the basic idea.  Growing up on D&D, I really only knew about a game that heavily encourages railroad style by its nature, especially as played and run by teenagers.  But this last year, in playing new games, I've moved much more toward a sandbox mentality. If I'm going to run a game with a sandbox mentality and I don't where my players where land up in my world (and I'm doing Song of Ice & Fire, which gives me Westeros to play with) I've got a challenge in having to slap together maps on here when I get to a combat or a warfare scene in my game.  Playing in person with a wipeboard, this was easy, I could just doodle a map.  But this is going to be a little more difficult to throw things together on the fly on here. Does anyone have experience and advice on this?  I want to give my players freedom to move about the world and not just push this them into contrived situations, but I don't want to make them sit around for a half-hour at a time while I draw a map for combat.

My best advice would be to create 8 small "combat maps", done as follows. Two wilderness (different types, desert, forest, whatever), two camps (tents in an arid region, wind-shelters etc in a humid one), two city ones (one in the south, one in the north maybe?) and two indoors ones. Once you have that you can pull some maps out of thin air. And if that's not enough, well, you can always make more. Eight is just my example, since in my experience that's actually enough. (though, to be frank, I tend to make a new one after the game to replace whichever ones I've used)

You could also put some different features on the GM layer of your maps then turn some off and on for different encounters, just so it's not the exact same battleground every time.

How do I get to all the stock features/terrain to paint my grid with now?

Good point Brett. I think I will use that trick.

There is a large element of railroading in sandboxing. If players don't head to town X and instead head to town Y, I'm sure the adventure you had planned for them will just reappear with the serial numbers filed off. Why waste material? As long as the players maintain the illusion of free will its fine. If they don't do something now, repackage it slightly and let it appear again later. The maps thing is a good idea but I' do it as a selection of encounters, which can possibly be spun into scenarios if you re stuck. A good start is the book 'Hellfrost Encounters' available as pdf on Drivethru. These are designed for the Savage Worlds Hellfrost setting, but they are pretty good and can be easily  re-specced to your own gameworld. 

Sandbox style is the exact opposite of railroading... wtf... sandbox style requires the GM to be far more prepared with extra encounters, dungeons, and such. Or else be very handy with creating stuff on the fly.

Bunnies, that assumes a traditional adventure, dungeon focused RPG, Song of Ice & Fire doesn't have that structure.  The narrative structure is completely different from traditional sword & sorcery RPGs so I can't just switch the name of the town of the dungeon and interchange them.  I'm not worried about getting stuck.  I'm worried about too much fluidity in my storytelling.  I have absolutely no problem with that.  I'm worry that I can't keep up with the demands of making maps that match the right settings my characters find themselves in a world that I'm buying into that I haven't made, that I'm taking as it comes.  The elements of the story will fall into place.  My only worry is that the maps look right so that they reflect the terrain features for the combat and warfare scenes when we get there.  And mats that just show grassland or mountains aren't going to do that, because I need to define the unique characteristics of the situation - but which I may not be able to play for, giving me the time I need to draw them up.

Personally, I'd say stick with the whiteboard approach.  Keep maps abstract, only filling in the key parts that are absolutely necessary at first.  If players ask for more details, fill those in at that time, not beforehand.  If you are doing 'dungeon delving' type maps, keep the grid on, and use the polygon-tool + snap-to-grid to enable relatively quick room-drawing.   Spend a little time outside of game learning the shortcuts in Roll20. I do think that with some practice (especially with the keyboard shortcuts), you'll find that you can whip up a new map in a similar amount of time as you would on a whiteboard. There's a little 'help' menu on the tabletop itself that lists most of the shortcuts available.

Balladeer, I played around with just drawing on a whiteboard, and you're right, this does make more sense.  I'm not doing dungeon delves, so most of my maps will be fairly simple maps of terrain features, the occasional castle and town.  I'm really not very good with graphics software and I'm not going to force my will on my players, so I think this is good advice.

Asaram's idea is probably the best for this type of game. I ran a sandboxy game in 3.5 DnD and it ended up well, but the players ran into issues where they wouldn't take enough hooks to finish up quest lines.  Creating maps wasn't an issue, but I tend to keep a COMPENDIUM OF ENCOUNTERS on hand, with different (and hopefully interesting) terrains pre-mapped that pair up with the type of encounters that will be abound.  I don't plot into every fine detail on the map, I merely copypasta the encounters that i expect to happen (based on storyline) and leave a few out and available in case the players decide "This quest sucks, let's go find a new one!". From there you can pretty much BS your way into making them believe you have an entire world set up before them. In my compendium, I have varying types of beasties that pair well with the terrain at varying CR's, so using this, you can have many encounters off of the same map (though I'd reskin it or switch up some structures before reusing said map..).  Sound confusing?  Just in case, here's an example of what I do, with a page from my book. Encounter:  Gate Guard Synopsis:  Ranged combatants rain fire down from atop ramparts and watchtowers while melee units prepare to strike fast and hard at anything charging through the gate into their encampment, which the party seeks to get into. Combat Variables:  - Ranged units could include a divine caster compliment to keep them healed/buffed and cast wind walls for added support to defend against enemy fire.  Could also include a supplementary arcane blasty type with area effect spells to keep party's melee from the gate. -Gate can be set on fire by party, or as last desperate attempt by defending units to deter the party. -enemy melee units ready/hold action behind the gate to strike anything getting through. Added challenge = include slits for arrows or polearms to fit through to keep party off the gate. -enemy melee units supported by a pair of stealth classes capable of striking from flanking positions or from the shadows behind the gate Party Friendly Variables: - The forested/rocky terrain provides enough cover for one or two sneaky types to stealth past the view of the guards to reach the base of the gates or watchtower -Knowledge (engineering) check provides players with the realization that one of the watchtowers is poorly built and with enough damage to one strut, it will come crashing down, opening a portion of the gate -there is ample cover in front of the gate for ranged combatants to snipe from, or for melee to advance behind. -Knowledge/social skills/disguise: the players know that the bandits beyond the gate are frequented by black market merchants who come to visit every tuesday. The players can utilize this to their advantage (naturally) Next I have whatever beastie/npc's that could fit here in this place. In this case, it's perhaps a bandit camp's front gate or maybe an encampment for hobgoblin mercs, or even the front gate to a mysterious castle. You decide, it's your sandbox. We'll consider this a hobgoblin encampment, since the party heard about some hobgoblin raiders pillaging small farming towns for food. Using some ideas for the different variables, I place a few archers upon the watchtowers, a shaman type of hobgobin for support there, and a good collection of melee guards behind the gate, just chilling out till some hapless heroes decide to mess with them.  I don't know how the game of thrones system really works, but with DnD, you can adjust the challenge rating of creatures through advancement, and NPC's can gain/lose levels.  If my party is around level 5, then i'll make the hobgoblins around that level of skill, or perhaps more skilled if the night is young and more blood needs to be shed... be it PC or NPC....   I have tips for adjusting monster/npc encounters, but let it suffice for now to say that you must simply tailor these encounters to the party and game style you're setting up.  Loot should be another step, but that all depends on what game you're playing too, I guess. Hope this helps. I tend to not make much sense when typing though, so if anything's unclear, just ask and I'll dispel your confusion.

Mark.... First, I would keep in mind my following perspective:  "Even in a Railroading game there are sandbox elements controlled by the Players.  Even in a Sandbox game there are railroading elements controlled by the players."  I hope it might help in some way. Second, I agree with Balladeer.  i especially agree with keeping things abstract.  I am trying to use that myself in my game. Third, I believe what Bunnies, Atticus, Brett, and Asaram have typed are related and could help you still.  i think the key is still abstraction.  obstructions on the battle field are similar across many types of terrain. Fourth, I suggest also keeping the player's plans for the future of their characters in mind.  It could help direct your focus to steady your story. Fifth believe it or not,  I believe you may gain more insight into this matter if you research about video game development.  I have seen many similarities between the two types of development.... Take it as you will.  Hopefully something will help your efforts.    : ) If you have any questions feel free to message me.

I know you're running DnD and not scifi, but have a look at some of the sandbox tools in Stars Without Number (free!) . Some of the best advice in the GM's section on running sandbox style games.  Plus, it has a ton of tables to aid sandbox GM creativity.  You could easily adapt these tools to a fantasy game.

I'm running a game that's completely different from DnD.  Song of Ice & Fire is a completely different system with a completely different narrative focus.  Anyone who signed up for signed would be completely disappointed with my games.   "Dungeons" are rare to non-existent.  The setting is low fantasy; it's the setting of the books and the HBO series.

Off topic. Out of curiosity, what is the mechanic in Song of Ice and Fire for pushing a Stark out a window? In DnD 3.5 I believe its Bull Rush.

I believe it would be an opposed Athletics check!