18 months on : FG vs. Roll20 vs Foundry

[HywelPhillips]

Hi All,

Like many of us I went online for my gaming in the first COVID lockdown about 18 months ago.

I started with Roll20 because it seemed to have the lowest barrier to entry. I picked up Fantasy Grounds Unity and Foundry pretty much as they launched.

I wrote up some mini-reviews a year or so ago, but a lot has changed on each platform in that time so I thought it might be time to update my thoughts on the various platforms.

For context I play two weekly games on Roll20 still, play in one "beta test" game with a GM polishing stuff up for his pro games on Foundry, and run all my own games (two or three a week) on Fantasy Grounds. If you are new to VTTs and especially if you are new to running games rather than playing them on VTTs, I hope this post will be of some use.

The majority of my experience especially as a player is with 5E; your mileage may vary for different rulesets.

First up:

Roll20- The Good

Roll20 has changed the least over the last 18 months. To be fair that is in part because it already delivered its mission statement pretty well - it's the easiest of the platforms to get going on.

As a player TBH Roll20 delivers the vast majority of what you need, so long as your GM puts in the work to set it all up for you. It has a full-featured character sheet for 5E with an excellent charactermancer for automating character generation and levelling up. It's not quite as slick and D&D Beyond, but it's still clear enough for relative newbies to level their characters up unassisted. Most player tasks are now handled with automation without having to delve into the unsupported Wild West of Roll20 Macros; you'll get a tick box to add your barbarian's damage when they are raging or your rogue's sneak attack, you can add bless and guidance pretty easily so long as you and the GM know what you are doing. You can add boxes to the character sheet to help you keep track of everything from session to session, like the hours left on a soul coin powering a vehicle in Descent to Avernus or the number of uses of your healing light power left that day.

You can buy all the D&D rules on there, and their marketplace has plenty of other rules systems with reasonable implementations on the players' side, and there are acceptable community character sheets for pretty much every system out there.

In terms of maps and tokens it lets you import external files for maps, add grids and lighting and line of sight. It's relatively simple and functional, but if you want Roll20 to understand your game system AT ALL you have to start adding macros, and that rapidly degenerates into a world-class mess of unsupported rumours and hearsay. Even something as simple as showing token health bars requires the GM to adjust settings accordingly, because the maps-and-tokens part of Roll20 by design is intended to be system neutral.

It can play background music for you but only from a static playlist.

It has no map-making tools internally and a simple four-map-layers model (map&background, objects&tokens, GM info, walls for dynamic lighting). As with much else on Roll20, it is minimal by design and system-neutral.

It has pretty video effects for spell effects but by default the GM has to activate those by hand.

It's looking-for-game calendar system is simple and good and since it has a critical mass of players, genuinely useful for making contact with people and finding games to play.

Where Roll20 really scores is its simplicity and low barrier to entry. If you are new to online play and don't have an existing group it's pretty easy to recommend as the one that will get you going most quickly- buy the D&D Starter Bundle for $80, find your way around the Icespire Peak module and the Player's Handbook, and you can be running a game by Saturday.

It runs in the browser, which means players don't need to install anything. It's performance is pretty good, although mobile versions are still in development and I wouldn't hold your breath. Again, nice low barrier to entry. The game is always up so players can check out their character sheets etc. between sessions. On the downside, it's out of your control. For a while they had a lot of server issues and sometimes it was plain not possible to log in or it was very laggy for reasons unrelated to your own machines. But that seems to have improved a lot recently, it has been months since I had a disrupted session.

Roll20 - The Bad

It has built in voice and video. Which plain DOES NOT WORK. I have never once been in a group that has persisted in trying to use it past the first week; in one-offs there is ALWAYS someone who can't connect. Everyone switches to Skype, Zoom, Teams or (the vast majority) Discord. Don't believe the sales pitch, avoid heartache and start on Discord instead.

On the GM's side it is a whole ton of work to run games this way. It's organisational tools for GMs are very primitive- essentially just a flat list of maps you have to flick between by hand. You can copy and paste groups of tokens from one map to another, but there's basically no organisational structure there to help you out at all. Running pre-written modules which you've purchased is... OK. But the moment you want to start customising stuff and making notes, it does nothing to help. Most GM's I play with run stuff with the physical game book by their side and post-it notes. Which feels primitive beyond belief to me given that there is a computer sitting RIGHT THERE and computers are great at that sort of thing.

The system-neutral nature of the maps and tokens means that the GM has to do all the work of book-keeping, from comparing an attack roll to AC through doing all the math for hit points for NPCs. Players can do it for their characters but some GM's handle that too, which is plain masochistic as far as I'm concerned. Compare two numbers to see if one is bigger, then handle hit point subtraction... by hand... when there is a computer sitting RIGHT THERE?

I got annoyed with running games that way after only a few weeks. I'm impressed by the stamina of the GMs I play with who still run everything that way - personally I want a whole lot more of the drudge work and organisational work to be handled by the VTT.

Roll20 macros are powerful and one improvement they have made is a better section for finding useful scripts and improved documentation. Ironically, given the low barrier to entry on the player side, I find the barrier to entry the highest of the three when it comes to adding automation and macros/extensions. Want to give tokens a red/green/amber aura to track HP, rather than relying on the bar? You can, but you'll still be combing forum posts to find them or trying to figure out how they work. For example, on the front page of their recommended scripts is

"TrackerJacker -- Graphical turnorder tracker, as well as status/condition marker tracking". The entire documentation for that reads:

TrackerJacker is a cousin to TurnMarker made by The Aaron. Its logic however is quite different, and more rigid in function to simplify the most common use-cases. Like TurnMarker, it uses a graphic image that follows beneath tokens to indicate who's turn it is. The only animation TrackerJacker supports is the spinning graphic which is on by default (to disable it you'd need to edit a simple script flag). The real function however of TrackerJacker' is to track statuses and durations with an easily accessible graphical interface that's intuitive.

Any the wiser? Me neither. Oh, and it was last changed in 2015 and I noticed at the bottom it was in Category: Discontinued API Scripts. Which suggests maybe it should be moved from the list of recommended scripts on the introductory page, do you think? This is symptomatic of everything in Roll20 that gets at all technical. It's still a mess, even if somewhat improved from 18 months ago.


They are developing things, but the pace of change is glacially slow compared with FG or Foundry.

You are limited by server space which is a pain and which ends up mandating a paid subscription for the GM. Dynamic lighting requires a paid subscription from the GM. API scripts and macros require the top-level subscription. It doesn't take long for this to add up to more than the purchase price of Foundry or FG. Everything on their marketplace is a bit more expensive than on other platforms, too.



WHAT I THINK IT IS GOOD FOR

Getting going today and running a commercial module for your friends by the weekend
Running a rules-light system, stuff with a lot of theatre of mind, running low-level 5E D&D where the recordkeeping overhead is tolerable
Running an ancient or obscure indie rules system where the GM does all the work like they do at the tabletop - you'll probably find a decent community character sheet already on there
Running commercial modules with little to no homebrew input
Finding players
Low barrier to entry for players, too


It's the lowest common denominator, in a sense. I know plenty of people who are perfectly happy with it and use other things (Discord, D&D Beyond, paper notes, Google Docs, Dungeondraft) to supplement its shortcomings. I have a soft spot for it as it was my entry into VTTs and I've played some absolutely super campaigns in it. But running them? It's not for me.

WHAT I THINK IT IS REALLY NOT GOOD FOR

Homebrew campaigns
Organising your material as the GM
Crunch-heavy systems
High-level D&D or stuff with lots of combatants. At 10th level, battles take up entire 4 hour sessions and OMG the work for the GM to track all the conditions! Those little token markers in the corner only go so far.
Integrated audio-visual experience - it's all on the GM to trigger stuff by hand
Using extensive external assets (maps etc) because of the server space
Long-term it's very likely the most expensive option

[bmos]

Yeah, I'm a player in a Foundry game and a Roll20 game and run a game in Fantasy Grounds.
If it wasn't for the people I play with, I'd have dropped that Roll20 game ages ago due to the janky VTT experience.
Foundry is way better, but the automation for PFRPG is so lacking (you can't even target attacks and resolve against AC) that it's a real letdown.
Fantasy Grounds doesn't have the prettiness Foundry does, but the game flow is so much faster/better and having official content makes it so easy.

[HywelPhillips]

Foundry is no longer the cool new kid on the block, it's been out for a year now and we can have a better idea of the direction they are heading.

FOUNDRY - The Good

That direction is lots of hackers making cool extensions/modules, with a core development process that concerns itself with supporting that first and foremost. The number of system-tweak modules on their system outweighs the number of adventures to run, which I think is illustrative of those priorities. It's a code-twiddler's paradise, not least because it uses the ubiquitous Javascript.

Foundry is versatile - you can run it on your own computer, or set it up on a cloud platform, or pay a modest fee in top of your one-off Foundry purchase to run it on supported hosted servers (eg at the Forge, which is where the GM I play with regularly hosts his game). It's cheap for the one-off purchase.

It runs in the browser, so no heavyweight install for players, although GM's have to install and manage packages either locally or on the cloud-based server.

It's PRETTY. It feels like Roll20 with the prettiness turned all the way up. Add in a few modules for automated spell VFX and it's gorgeous - I know it is a bit of a gimmick but it IS cool seeing your warlock shoot off purple Eldritch blasts at your target. Crucially, unlike Roll20, this is amenable to automation. One of the lessons I learned with VTTs is that GM's are carrying a heavy cognitive load, so the frills are much better left to the machine to trigger or you'll forget.

Playing with a group of experienced players I would say that its user interface is more familiar-looking than FG's, but more complicated that Roll20's (naturally enough, because it can do more than Roll20). With us not playing every week, we all forget how to do stuff in between sessions and anecdotally I'd say that's just as prone to happen in Foundry as in FG. It doesn't seem to happen much to people on Roll20.

It is now possible to have a significant degree of automation in Foundry. The 5E game I am playing in does the to-hit checks for you when you have a foe targeted and make an attack, rolls damage, takes it off their hit points. You can set it up to handle conditions and the like. This is not core functionality, but added in through modules (and I'll have something to say about that in the negatives section). It's a bit flakier and less comprehensive than FG's automation for sure but it works well enough that I'd definitely choose it over Roll20 for that reason alone were those the only two choices.

They have a decent marketplace now and some companies are putting out nice commercial rulesets and adventures (Alien and SWADE are pretty good on there). But since what automation there is is usually implemented by third-party modules, out-of-the-box these products contain little more than Roll20 ones - it's all on the community to develop more automation support.

In GM organisational tools it is also intermediate between Roll20 and FG. You can pin stuff to maps (and indeed I much prefer its selection of icons for doing that to FG's pins). It's got more possibilities for map and scene organisation and so on. But it is all developed in a system-neutral way - which means it doesn't really "speak" RPG at all. It doesn't know the difference between a feat or a spell, and as a player this neutrality can often end up with you having to use a lot of mouse clicks to do something. It's so system-dumb that it doesn't provide anything like Roll20's charactermancer as far as I can see. There is a solution for 5E involving scraping or roll integration with D&D Beyond, for which see negatives. Our group has devolved into letting the GM do all of that work, which isn't ideal, but Foundry doesn't provide the hand-holding for players that FG or Roll20 does.

It definitely plays faster than Roll20 for large combats if you have a bunch of modules set up to handle to-hit/AC/damage and help automate conditions.

The UI for stuff like light placement is very slick.

It has drag and drop for a lot of stuff like character classes and inventory items and spells.

Audio is really good, including static playlists and ambient noises triggered by proximity on the map. If it had hooks into Syrinscape that would be absolutely killer and very easy to set up. (There's probably an extension for that).

Video FX and animations and stuff like animated map support and Dungeondraft import are great. Credit where it is due some of the extensions provide spectacular stuff painlessly. Still the one to beat in this area.


Screenshot 2021-08-29 at 14.30.02.jpg

Did I mention pretty? It has icons for all sorts of things, including the full 5E SRD.

[HywelPhillips]

Foundry- The Dark Side

So what are the problems? There are in my opinion some really big ones.

1) Take a look at that screenshot of a game in progress. Does it remind you of anything? It reminds me of Fantasy Grounds. Combat tracker, chat window with results, the system rolling and applying damage, adding conditions to foes. Thing is, that's all done by third-party extensions. The GM for that game is running a WHOLE SCREENFUL - 40+ extension to get it to do what FG does out of the box. They are stuck on a much older version of Foundry because in all the newer ones, one or other of these extensions breaks or interacts in unpredictable ways with other extensions.

They have achieved the "crumbling edifice" state of software development where no-one dares change anything because of unexpected snowballing effects making the whole system stop working. Given that the program has only been available for 12 months, that's quite an achievement. I want the automation, but not at the expense of being stuck with a very out of date system that I cant change until the campaign has finished for fear of the whole thing collapsing! It's really heavyweight on the GM in terms of maintenance, and that's even running it on The Forge which offers to do a lot of this legwork for you. If you want to extract all the potential functionality from Foundry, you'll be doing a lot of maintenance.

Oh and as I predicted in my review last May, their system is already littered with abandoned modules produced in a rush of enthusiasm by someone with time on their hands, then left unsupported. Its a community of enthusiastic code-heads, not a professional company supporting these extensions as products.


2) No D&D official content outside the SRD. While some companies like Pinnacle and Free League are doing great things releasing commercial products for Foundry users, the elephant in the room remains the fact that for the most popular system of all, you can't get the content officially. If you want even the PHB - let alone stuff like Xanathar's or Tasha's - you are out of luck.

Instead, they have a frankly bizarre workaround which involves using an extension to scrape all the content from D&D Beyond. So first you'll have to have all that content on D&D Beyond, then you'll have to run some hacker scripts to import everything to Foundry, and if it changes I guess you have to do it again? This requires you to support a lone hacker on Patreon, and it makes me uncomfortable for several reasons. First and most important I am a content creator by day, and I think it is morally dubious at best. MAYBE the D&D Beyond terms and conditions allow it but I'm pretty sure it isn't in the spirit. I'd MUCH rather pay to own the content unambiguously. I want to pay, because I want to support a stable ecosystem which supports my platform of choice, my game of choice and funds maintenance and support in the long-term.

What're you going to do if WOTC pulls the plug on D&D Beyond? This all has a whiff of piracy to me regardless of how they are skirting the actual legal T&C.

3) You want that? That's an extension. Or write some code.

I'm mildly critical of FG for leaving some stuff which I feel should be core functionality to be handled by extensions. But on Foundry this philosophy is all-pervasive. Even popping out your character sheet to a separate window (core Roll20 functionality and hardly rocket science for something browser-based) requires an extension. The Foundry Discord is reams upon reams upon reams of discussing what extensions to install followed by how to get them all working in the same version. This is a recipe for swift development but a disaster for anyone running a year-long campaign. I've been running my regular group in FG through some absolute convulsions like FGU going through and out of beta and the release of lighting and every week the PHB has been reliably there for my players and the commercial adventure module still works and wow now the maps have lighting! In my limited experience of Foundry everything breaks all the time.

Example - you can drag a weapon to a hotbar slot to speed up attacks, but doing the same for a stealth skill roll requires writing some code. Drag and drop functionality is skin deep.


4) The UI is more familiar at a glance, but it still requires arcane juggling the moment you dive deeper. Double right-click to target is not superior to CTRL-click to target. And because Foundry doesn't speak RPG, accessing stuff is often a bit obscured by the totally generic nature of the object model. For example, you cannot see what damage weapons do at a glance on your character sheet. You have to open up the details pop-up, which is a tiny icon and highly non-obvious. I would actually say that we have more plaintive calls for help accessing basic player stuff to the GM and the one technohacker member of the group for Foundry than we ever have for FG. So this much-vaunted UI is actually no easier in practice than FG's much-derided UI. Especially not once you've got 40 extensions installed.

5) I'm sorry to say it, but the fraction of tribal asshats in the Foundry online supporter community seems to be higher than FG or Roll20. There ARE plenty of nice and supportive people out there, but there's also something of a rabid-follower-one-true-way-information-must-be-free-sheeple thing going on too. This may just be my mistaken impression but there it is, that's my experience. I found it rather off-putting.


WHAT IT IS GOOD FOR

Medium weight automation for games supported by publishers like Alien, SWADE, etc.
Pretty. Very, very pretty.
Automated pretty (much better than Roll20 in this respect; FG doesn't provide anything even faintly comparable in scope yet)
Simple ways to extend it yourself if you've got a bit of javascript
Wide range of community rulesets

WHAT IT IS NOT GOOD FOR

Maintenance reliability the moment you have extensions, which you will because achieving the awesome requires them
5E content, and purchased adventure content generally
Picking it up and running a game by Saturday (our regular GM spends HOURS AND HOURS doing prep. It's nothing like buying Icespire Peak and away you go)
The learning curve is significantly steeper than Roll20 and player-side automation is much worse especially for 5E
Hard to recommend as your first VTT as a result unless you are comfortable with GitHub and basic developer sysadmin chores

Foundry is a hacker toolbox, not a one-stop-shop VTT

[Jiminimonka]

Great reviews. Love the Foundry one especially.

[HywelPhillips]

Fantasy Grounds - The Good

Fantasy Grounds has changed the most in the last 18 months. FGU is now out, stable, and with some major extra features, most notably lighting and image handling in general.

This change has been managed without breaking everything. Scenarios first published years before FGU even launched continue to work. As new features have been added, the back catalogue has been updated (not universally, but still to an impressive degree). For free. This speaks very highly of the company's commitment to long-term support of their customers and the things those customers have bought from them.

If you are reading this and you've not run a game in FG, it's hard to express just much deeper the automation and system support runs compared with Roll20 and Foundry. FG consists of a core VTT engine to handle maps, lights, walls, tokens, images and so on. But sitting on top of that is the ruleset - and the ruleset actually understands the roleplaying system you are running. So unlike Roll20, the ruleset knows how an attack works in 5E and will automate it for you. Select your target, click on your character sheet, FG rolls the dice and tells you whether you hit. Roll damage and it applies it to your target. It understands what critical hits do and rolls the extra damage for you correctly. It understands what damage resistance does and halves it for you. It knows how temporary hit point work and handles that for you. It knows about half-damage from non-magic weapons. It knows about how prone works and if your attack can knock someone prone it'll facilitate you making them roll their save, and if they fail it'll let you apply the prone condition. It knows that melee attackers get advantage on you if you are prone, and ranged attackers get disadvantage. If you make spell saves it knows which spells cause half damage on a successful save. You can fireball a room full of zombies and with two clicks FG will roll all the saves, apply full damage to those who failed and half to those who succeeded.

As a result combat at higher levels in D&D is AT LEAST twice as fast to play in FG as in Roll20, even more so as the number of combatants increase. This was the grand finale big boss battle of my 18 month campaign last week - 12th level PCs, a vampire-lich homebrew, corrupted treants, vampire spawn, 12th level undead NPC spellcasters, and a whole barrel of minions. We did the lot in three hours. In Roll20 I guarantee that would have taken two or three full sessions. And as GM I could concentrate on WHAT the NPCs were doing it, rather than keeping track of them doing it. Finger of death spells, healing potions, multiple fireballs from both sides, all handled effortlessly by the automation.


Screenshot_2021-08-19_at_19.28.38.jpg

Note how pretty this is - this is a relatively plain map (Sunless Citadel) brought to life in a really cool way by lighting. Such a thing was not possible in FG 18 months ago.

Until you've experienced the reliable out-of-the-box automation that FG provides it is hard to convey just how much easier it was to run this battle than it would have been in any of the other VTTs I have tried. It has made us brave enough to aim for 20th level in this campaign, something we've not done for in-person games precisely because the record-keeping gets too onerous for the players, let alone the GM.

FG has the best organisational tools for the GM bar none. It provides the things you want to run a TTRPG - like encounters and parcels of loot which are made of items and can be dropped onto the party, and shared out amongst them. Every Roll20 group I am in has a by-hand party loot sheet which some poor sap has to keep up to date. The party sheet on FG does it automatically and effortlessly.

You can pin stories to maps, prepare encounters with pre-placed tokens without cluttering the map or the combat tracker. Foundry comes close here: in some ways I prefer their method of pre-placing tokens and selecting which ones to bring into the battle when it starts. But FG has extensions too and if you prefer to do it that way, there are robust extensions that do the job. There is still be something of the "abandoned extensions" syndrome but my experience is that the FGU developer community is much more committed to long-term maintenance and compatibility.

One complaint about FG has just been addressed - it used to be that you had to manage extensions by hand, which was a pain. The Forge fixes this and supports a much more Foundry/ForgeVTT like way of choosing, installing and downloading updates for your extensions. This is a significant improvement.

FG has the best marketplace for playable content (modules and so on). Roll20 probably has more systems, but FG has more adventures, and you can use more than one adventure in your campaign which is curiously convoluted in Roll20. The commitment to updating old modules with the latest line-of-sight and lighting has been exemplary. Material is properly licenced and once you've bought it you own it, in a local copy. So it's never going to just evaporate (like Foundry's D&D Beyond-scraping might).

Where the ruleset automation is lacking, particularly for 5E, extensions take up the slack (for example I use automatic pack tactics, automatic mirror image, automatic death resistance and the like just so I don't have to remember them). Unlike Foundry extensions which have to be very broad given the lack of automation out of the box, these extensions can be tightly-focussed and functional. Sadly conflicts and errors do occur, but they are usually resolved quickly by the authors. The most I had to do even during roll-out of major feature updates like lighting was to turn a couple of extensions off for a week or two.

A few extensions significantly extend FG's functionality - I'd particularly recommend MatteKurre's Syrinscape integrations (pin sounds to maps, and have sound FX triggered by stuff as you play - someone casts Charm Person, the Charm Person sound FX is automatically queued and played by Syrinscape). Portals is great for letting people explore the map organically and transport them to other maps. One-Click Druid or the Polymorphism extension handle one of the most onerous of 5E features (wild shapes and polymorph) effortlessly. Some of these I personally think should be core functionality (portals) and some stuff is still missing but overall the degree to which you can automate your game is truly impressive... without it feeling like a computer game.

Maybe it is just because I think the same way as the developers, but I find it a genuinely pleasant experience to fire up FG to do my session prep. I like to base everything on the map, and add annotations via pins. FG makes this very easy to do. Stories, encounters, npcs, image handouts to share, notes to give to players and so on. The spirit is the opposite of Roll20 and Foundry's system-neutral approach - FG tries its best to get into the nitty-gritty of each supported game system and provide the tools you actually need to prep, run and play it.

The level of drag-and-drop support is impressive. (If not always the most intuitive for new players - see the bad stuff post).

Oh, you can build maps entirely inside FG, facilitated by the art packs. Personally I think it is better used to customise base maps imported from elsewhere - there is a Dungeondraft universal VTT file importer as an external script that I very much wish would be incorporated as an extension or folded into core functionality.

I was running an adventure where they killed some owlbears. When they returned to the area, they found bloody trails leading away where someone had dragged the corpses away to animate as zombies. Adding this to the commercial module map was a simple matter of finding some nice blood spatter assets and painting them on the map. It took me a couple of minutes, tops. The impact was great - when their characters returned to the area the players discovered the blood trails for themselves, not via me reading flavour text. They freaked out and followed up the clues themselves, talking amongst themselves. FG's image tools are awesome for that.

[bmos]

Wow. This comparison is absolutely aces.
I like how it isn't judgemental and acknowledges strengths and weaknesses all 'round while accepting that different VTTs could be better for different games/tables/situations.

[HywelPhillips]

Finally:

Fantasy Grounds - Some Painful Truths

There are some negatives to Fantasy Grounds and let's get the big one out of the way first.

1) The user interface is divisive at best and hated at worst. This comes up time and time again from people who have investigated FG and decided not to use it. It's never that they think the automation is too much, or that their favourite system doesn't have good enough support. It is always that they found the user interface confusing, ugly, unintuitive, clunky, or old-fashioned.

Some of these comments are unfair because they haven't looked again at FG in its Unity incarnation. It's certainly much improved over Classic, which looked every bit like the 90's computer game a lot of people compared it to. New developments have been implemented with more modern UI conventions, the the drawing/image layer tools, which are using a pretty widespread set of UI conventions. I rarely see people commenting negatively on this part of the program.

But there are still a lot of things which bear that 90's computer game DNA, most particularly the context-sensitive right click menus. And the UI is scattered with minor annoyances (as in a recent thread about why sometimes you need to double-click to roll a d20 from your character sheet, and sometimes single click it). The module window is still broken, not showing you the bottom row and not being able to expand the window to do so.

A lot of this is tied in with the good feature of backwards compatibility that I praised above - rulesets developed 10 years ago still work, but that means you can't do a complete redesign of the UI without breaking them.

Roll20 isn't immune to such annoyances. (To read an ability you click it, but if you do that to spells you ping it to the chat. Or maybe it is the other way around. It should be the same convention whatever, players get that wrong at least once a session). I've already complained about some of Foundry's UI quirks.

So however much I enjoy using FG there is no denying that its UI puts people off. Different themes can help a lot here (SirMotte's dark theme for example which I'd love to use but it is too low-contrast for our group's middle-aged eyes).



2) It's very hungry on screen real-estate. Partly that's because it does a lot more (so for example the combat tracker window needs to be a lot bigger). But also because it eschews modern UI conventions like tabbed windows in a lot of places where they could be implemented. The chat window doesn't work like other windows, so you can't bring it to the front when the map is in your window background. Again, newer additions do better here but it's still struggling free of the confines of Classic in many areas.

Different themes can help a bit here too but only to a point.


3) Performance. It's remarkable that we can run combats with lighting and 30+ high level combatants and sound effects triggered through Syrinscape extensions and the works. But there's also no denying that FG does not make best use of the hardware, because the game can be stuttering and lagging whilst I still have 9 cores out of 10 sitting idle. There are workarounds - /imagequality and /vsync help people with lower-spec machines a lot - but in my opinion there's a lot of room for improvement here.

Let me not overstate the case. FG is perfectly usable. It's just that from the resources my iMac Pro has sitting idle, it should FLY. And it doesn't.

Some specific issues: search in the asset window is painfully slow if you have a large collection of maps, which I do. Whereas MacOS searches those same directories almost instantly. PC character sheet action tabs with long spell lists can be laggy especially on older laptops for players.

I really hope this is something that SmiteWorks will continue to work on.



4) You do have to download the software and install it on a computer. This is a barrier to entry for non-techie players compared with following a link in their web browser. Conversely, it means that everything is under your control, so it has compensations. In the unlikely case that SmiteWorks goes bankrupt, you'll still be able to run your games and have your complete library of purchases. For me the main annoyance here is probably the lack of offline module reading and the fact that players can't access the game while I'm not hosting it. There are workarounds (they can create their own campaign and import their character to look at it out of session) but it's not as good as letting them have their characters on D&D Beyond in parallel to FG.


5) The annoyance that the automation isn't complete. One aspect is that commercially-purchased material may not implement automation as fully as it is possible to do in the existing system. It does rankle a bit to purchase Rob2E modules with fully-coded effects to get the full benefits of automation for PHB, Xanathar's, etc.. I understand why: the time it takes to implement every spell with all its variant effects and actions is non-trivial. And you can do it yourself if you have time. FG say that they aim for 80% automation and I'd say that for 5E in particular it is already higher than that out of the box for FG-plus-PHB, way beyond the dreams of Foundry and something that Roll20 hasn't even imagined.

Also annoying are the things that are crying out to be automated because they are hard, which FG have on their list to tackle but which haven't been yet (like paladin Auras, targeting shapes, effects attached to magic items that trigger when you equip them). It's churlish to complain when FG does so much more than the other VTTs, but also hard not to want it all!


6) The find-a-group/game calendar side of FG is not as good as Roll20. There's a forum and a Discord but it is not as easy to get your first game as it is on Roll20.


7) No sound integration.


8) No animated graphics file format support, which means no cute VFX for fireballs or rays of acid like on Roll20 or Foundry extensions.

9) It has a 5E charactermancer, but this is not as user friendly as D&D Beyond or Roll20. The UI is somewhat primitive and confusing eg the paladin class will show up several times if you have PHB, Xanathars and Tasha's all loaded. On the plus side, it picks up homebrew or third party content on an equal footing, which is much less well supported in Roll20. I think Roll20 has some Kobold Press content, but it's not seamlessly integrated the way it is in FGU's 5E charactermancer. And D&D Beyond is WOTC content only.


Do these sound like minor quibbles to you? That's good, because that's how I feel about them. For sure I'd like them to be improved. But for me most of them fall into the category of fine-tuning. What FG does out of the box is bloody awesome when you get your head around it.


WHAT IT IS GOOD FOR

Running adventures with full automation and GM record keeping. 5E, Pathfinder, Starfinder, 2E, SWADE in particular the automation is hugely impressive
Structured GM development environment - you can do it all in FG in a way which is not possible in Roll20 or Foundry. In many ways it provides a lot of the functionality of things like WorldAnvil and map-making software all inside your campaign, and that's really hard to let go of once you've experienced it!
High level adventures
Running commercial modules with excellent system support
Playing together week after week
Homebrew - not just homebrew adventures, but homebrew classes, races, spells, skills, items, etc.. All are treated as first-class citizens by the system. It's really nice.
Supportive company
Supportive community
Backwards compatibility taken seriously
Creator support and legal issues taken seriously
Marketplace with very rich content

WHAT IT IS NOT SO GOOD FOR
Introducing scared technophobe newbies into playing D&D online. They need to at least be comfortable with installing software and running check for updates.
Running minority systems. I mean there are community character sheets and even rulesets for a lot of stuff. So it's not worse than Roll20 or Foundry here. It's just that the automation isn't there the way it is for the best-supported systems.
Playing games with people who find its UI horrible.
One-shots with strangers (especially those new to FG).


If a new player asks me which VTT to try first I always recommend Roll20.

If a new GM asks me which VTT to try first I would debate which to recommend. Roll20 for sure will get you up and going for your first session more quickly, and it might be all you need if you're only going to dabble.

But Fantasy Grounds does so much more, and is so much nicer to develop your own adventures in!

If you are at all serious about GM'ing, I recommend Fantasy Grounds. You will have to work at it a more at the start, because there is so much more than it can do. Compared with Foundry, you'll get FG up and running with automation much quicker and there's the brilliant option of just buying IceSpire peak and running that while you learn the ins and outs of FG. You can't actually do that with Foundry.

You'll curse me for the first few weeks while you get up the learning curve. Then you'll thank me for the next five years while you watch GM's struggle with all the record-keeping chores on Roll20.

In some ways my ultimate recommendation is this - when you're running a game in person, would you want to have the VTT running beside you? Roll20? No, why would you? Foundry? Maybe, for pretty. Fantasy Grounds? Yes, because of GM organisational tools. When I ran my first in-person games in 18 months on holiday for my regular group, I started running it without my laptop and halfway through the session I went and got it so I could have FG running beside me as I ran it, even though I wasn't using the combat tracker. It's a powerful GM support tool with a built-in VTT.

[Moon Wizard]

Thanks for the balanced reviews. Very good reading.

Regards,
JPG

[stephan_]

Excellent reviews, thank you. Will be my go-to when anyone asks about differences between VTTs.