The Cost Of Moddability

[@PaintsAreOp]

Making a game moddable is hard and costly. But you know what's even more expensive? Creating anti-tamper systems that make sure the game is not moddable under any circumstances.

[@JediMB]

Yeah, that's easily the most annoying obstruction. It's understandable if a developer doesn't want to actively support modding, but for single player games the least they can do is to not actively make things difficult for no reason.

[@irAbyssion]

My mind went straight to Styg the creator of Underrail. Although to my understanding the anti-modding stems more from how insane his game is coded. Which I personally admire with how spiteful, entrenched and stubborn he is with his community and games.

[@gobogoo2329]

do devs normally make their own anti-tamper systems though?

[@old_scaly]

⁠@@irAbyssiontoo bad he’s also a homophobic, racist bigot. Though I guess that explains why one of his guiding principles is spite.

[@PaintsAreOp]

@@gobogoo2329 No, but even when you use something like Denuvo, you still need your own dev time to put it into your game.

[@pelicano1987]

I'm in heaven to see a video from Tim about a question by me. I grew up with Fallout and many other IPs by or worked by you. Thank you very much.

[@finesseandstyle]

There's another hidden cost of moddability when you do have it in your game. You must be a lot more deliberate with updates and how you decide to do reworks. Updates frequently break many mods that rely on a specific implementation of that version or they just need to match the new version number in the config. Bethesda did this many times when they released like 4 different definitive editions and many random pointless updates (like adding creation club into main menu) years later that broke many mods. If this becomes frequent, you're at risk of alienating modders, which is the last thing you want for all the effort you put into making modding possible

[@Dirty_Dumb_NAFO_Scum]

Do single player games really need that many updates? No, they don't. Maybe a few updates past 1.0, but after that and it's literally corporate vindictiveness.

[@markustambet4098]

if youre playing on steam, you can literally just roll back the game to the version the mod worked on

[@finesseandstyle]

@@markustambet4098 Good, seems like it works for many games that use mods but it's not universal.

[@JMww2]

@@finesseandstyle yeah warno is a good exemple the dev change the game files every or so update... because it maybe just table for the dev but for the modders it's text files and it isn't as easy as renaming a column you need to fix files with python script that does gatekeep modding somewhat otherwise it's a time sink

[@FatherAlduin]

Hey Tim! Very insightful video, and I wanna stress how helpful it is as a modder when a dev includes a proper error log and exception handling. There are games out there that have absolutely no form of error logging accessible to modders and it makes debugging them an extremely painful experience, and when you contrast the modding community of those games with the community of games that have much more in-depth logging and better accessibility in general it really shows! Have a good Christmas Tim!

[@torginus]

I kinda miss 90s and early 2000s games with their interpreted scripting languages, and in-house level editors, where moddability was a given - since the scripts were interpreted, you could easily just replace any piece of game logic with your own. Nowadays, everything seems to be compiled and sealed in a black box. Modding in those days had another amazing advantage - it gave amateurs exposure to the complexity of real AAA-games. These games often had million-line code bases, and learning to wade through them, and understand how a project of such complexity can be built up incrementally was a huge eye-opener for me. Something that's certainly very difficult to learn if you start game dev by yourself from Unity tutorials and work your way up from there.

[@gargamellenoir8460]

You know that there are more moddable games than ever right? You need to stop looking only at AAA games.

[@UlissesSampaio]

@@gargamellenoir8460 Maybe so, but at that time, the biggest AAA multiplayer games were highly moddable. It was glorious.

[@__-fi6xg]

the fuck are you on about, there is more mods than ever before, just use the internet dude

[@nairocamilo]

​@@gargamellenoir8460Not the point, I think

[@MrSnivvel]

Idk, I've seen more games use Lua as a scripting language, and expose it to modders, than compared to using their own in-house created or compiled binary-only language.

[@prplfleur]

ok Tim but.... why dont you just do this?? IF(not moddable) THEN (is moddable = TRUE)

[@Cheesyonmytoasty]

might as well be `is moddable = true`

[@bam_bino__]

this is an infinite loop

[@virionspiral]

​@bam_bino__ not a loop since it's just a one time conditional, but if it was like a while loop then yeah it would be infinite since the condition keeps being reset back to a negative.

[@buriedbones-nh9xr]

@@virionspiral It is an infinite loop, granted if its inside a while or Update method And i dont know any language that could compile this In c# for example this if statement would look like if(!moddable) { moddable = true; }

[@virionspiral]

@@buriedbones-nh9xr that's what I had said that it would be if it was in an actual loop but it's not. And any compiler would compile that code. It's a logical error not a runtime an error

[@MrJekken]

Another reason a lot of companies don't do mod tools, or do it far less than they used to, is they often view mods as competition. Rockstar is a good example, where even though they never purposefully made their games moddable, the games had extensive modding communities and with the old GTA trilogy, they DMCA'd the work of modders and de listed the original versions of their games. Bethesda and Valve have the right approach and it benefits them long term even with the cost of adding those features because their games can have longer tails and relevance, they can monetize and even publish certain mods in collaboration with mod authors and it allows them to have a high aggregate talent pool for their in house engine that they can hire for without needing to do too much on the job training or tank the cost of switching to an OTS engine.

[@Verchiel_]

Agreed High priced as the mods may be that are published on BGS games, i think that's a small con relative to how much, i believe, good, it brings. Providing a monetary incentive to developers, possibly allowing them to learn modding further to the point they're actually viable candidates for hiring, which Bethesda has openly done. My really only concern is whether or not it devolves to stolen mods being put up behind a paywall, which is the issue that happened (almost certainly still is) with minecraft marketplace. At the end of the day, so long as the company behind the game isn't actively fighting the community building the game obtusely for purpose of limiting modding potential, or straight up taking down mods directly purely out of any perceived threat as dumb as that may sound.

[@BubbleoniaRising]

As I’ve gotten older I’ve found that I spend about 80% of my time engaging with games that support modding because I enjoy that aspect of gaming far more than the alternative. I still play a handful of new games each year, but the majority of my interest is in the maker community.

[@gargamellenoir8460]

"Is making a game moddable easy/costless" isn't the real question, of course it's not. The real question is, what are the benefits. Most publishers seem to treat it as nice to have/courtesy to the players, assuming they're not actively hostile to the concept. But modding can be a multiplying force for a game's life and therefore sales. It turned vampire bloodlines, fallout new vegas and KOTOR 2 from crippled games to the true legends they were meant to be. It made Doom, Baldur's Gate 1&2 and Skyrim virtually immortal. Modding is so good for games that it created a vicious effect on Bethesda when they released crappier and crappier games just so the community can fix it in their stead. And then trying to monetize the mods and get a piece of that action! Making a game moddable isn't a courtesy, depending on the genre it can be a huge investment.

[@darmok3420]

skyrim is in no way the only moddable TES game, it's also made Morrowind and Daggerfall immortal and Oblivion to a lesser extent.

[@blueskull7898]

Just think about what that game “Gary’s” would be like if there were no mods

[@paulie-g]

Now imagine you're not an AAA studio that knows for a fact they'll sell enough copies to create a modding scene. Because AAA devs don't need to watch a video eli5ing mod support. It is *very* challenging for indies to do and far from certain that it is actually worthwhile vs creating more content or paying for more marketing. If a game sells 10k copies (and, for a first-time indy, that's often a raging success), there likely won't be enough modders to create enough added value simply because modders like to have at least a bit of an audience.

[@gargamellenoir8460]

@@paulie-g And yet a lot of indie games ar emoddable as frak. Look at Minecraft, made by one dude. Or Project Zomboid by a tiny studio. "Moddable" isn't a binary proposition, a lot of creators just open the methods and let the community figure it out. Or simply take some time to document how the scripting files they use work (which is good to do even internally) and let the fans do the rest. Modability is a spectrum.

[@jorge69696]

Another example is gun skins for counter-strike. The previous games had community made skins that were free to download. Valve then took that away to replace it with their own paid skins.

[@saffral]

One thing I appreciate about Bethesda's engine is that while you have no access to code, the entire game is set up like a database with the executable reading what the game will be from it. The plugins (ESP files) are treated as database queries that update, add, or remove entries in the master database (ESM files). It does limit the true potential of how far mods could go; though I'd say that's been less of a problem than some might assume. And with the script extenders making code modifications to add in new script functions there's a lot of freedom. The database approach also means that dozens or even hundreds of mods can all be loaded simultaneously without the end user ever needing to do anything beyond copying and pasting the mod files into their data folder.

[@grungehog]

Well the plugins still need to be sorted and enabled in the form of a loadorder.txt The plugin limit is the typical 254, but through new plugin types Bethesda managed to circumvent that limit, I've read load orders exceeding 1k plugins this way before.

[@NovaFinch]

It's why most Bethesda modders hope that Bethesda never changes to something like Unreal. The ability to have hundreds of mods installed without having to patch together every single one is massive for allowing players to customise their game however they want.

[@saffral]

@@NovaFinch I'd expect that they could still make it work with enough time and money put into recreating that system or creating a new one (they already created it on top of NetImmerse/Gamebryo after all). But it's one more cost to switching and one more thing they'd still be maintaining instead of using the new engine's data structure.

[@darthryking]

Although the Source Engine has tons of problems, being inherently moddable is one of its core strengths, IMO. The way the engine is designed, pretty much all Source games (even the official Valve games) are considered by the engine to be "mods" (usually of HL2), which means that with a few exceptions, for the most part every Source game is automatically moddable just by creating a new "HL2 mod" and changing its mounting configurations to point to whatever content/script/config/asset files you want.

[@SuperFranzs]

Is that inherited from the Quake engine?

[@mryellow6918]

this is the reason half life 2 with raytracing exists, because portal rtx, is just half life.

[@LeadHeadBOD]

​​@@SuperFranzs mostly yes, idTech 2 followed a very similar workflow, although Source of course comes with nearly a decade worth of tech advancements.

[@UlissesSampaio]

As a gamer of the late 90s and early 2000s, many of my favorite "games" were mods. So, I'd love it if AA and AAA made modding a thing again. However, a big issue nowadays is that mods often decrease earnings: multiplayer games tend to sell skins, which conflict with mods. Man, there were so many great Half-life mods back in the day. All free of charge.

[@UlissesSampaio]

Hell, Counter-Strike's "The Dust" map, one of the most popular ever, was made by a teenager (there is an interesting No Clip documentary about it). Nowadays, I see fewer options for players to participate in making the games themselves, which is a bummer: nothing beats the "player hivemind" in creativity. Just see Skyrim and now BG3.

[@LDiCesare]

Very correct. Compare Civilization IV mods and Civ VI DLCs for example.

[@Helperbot-2000]

@@UlissesSampaio well counter strike as a whole was as half life mod too

[@stuartmorley6894]

A lot of AAA games these days are moddable. Look at stuff like Cyberpunk, Elden Ring etc. A lot of developers don't even seem to be setting up the games to be modded and yet people are still using external tools to inject code. Look at Nintendo games. They are now modded to death but Nintendo spends an insane amount of effort striking down modded content, tools to mod games. Capcom games are heavily modded too but they spend a fortune trying to squash the modding scenes for their games. The issue nowadays seems less like games aren't moddable than it is that certain publishers institutionally hate modding because they can't control the end product. You only have to look at a game like Skyrim where they let you mod anything but retain the right to get rid of mods that would break terms of service. All that's happened is that the huge amount of sex mods are all on LoversLab instead of Nexus or the Bethesda mod platform. Just any kind of odd thing you can imagine. A bunch of publishers look at stuff like that and just nope out.

[@lrinfi]

That they "decrease earnings" is obviously the prevalent, industry view at the moment and, I think, it's because the _industry_ can see only the financial "benefits" of doing much of anything other than "increasing revenue" and/or establishing "recurring revenue streams." It has absolutely no sense -- as human beings do -- of such "benefits" as community goodwill; healthy relations among human beings; and other such immeasurable, utterly nonquantifiable "benefits" to humanity and the world at large. Perhaps "the anonymity of the corporate economic system," to quote David Loy, largely prevents such benefits (along with their polar opposites) from being seen? Quanitfiability unfortunately rules supreme just about everywhere and in every human endeavor imaginable at the moment while quality largely lies fallow. To say there's an imbalance there would be the understatement of the year. Perhaps just a case of "we see things not as they are, but as we are," but the Automatron DLC for Fallout 4 was quite a bit better than many of us think, imo. If we recall, "The Mechanist" could only see the statistics of "hostiles killed" streaming back into her lair. Poor Isabella. She had no idea whatsoever what her robots were actually doing in the Commonwealth any more than many industry executives and "HQs" have any idea what their robots are actually doing in the Commonwealth. Why do I get the perhaps mistaken impression some developers were blowing off a little steam there? :) Whatever the intent, I (for one) thoroughly enjoyed partaking in it...for what that's worth.

[@Steelion69]

The funny thing is. Moddability is generally viable as a result of game development. Designers want to tweak and balance things within their games - especially if they lack a whole lot of Programming knowledge, and programmers will want to fix things broken in the soft code quickly. You don't need to mess around taking hours to compile a game; just edit a value, save the file, restart the game, and hey presto! now you can check if your fix or balance change is viable. Honestly, XML really was a game changer and pretty much is a key reason why modding has become accessible in the last 20 years for many games, whilst still having minimal performance overhead. Of course there are limitations since XML still depends in pre-defined data generally located in the compiled source code...and that's generally where modding ends. New items, and entities, etc - maybe limited to a hard cap for the sake of optimization. But that new content must still strictly adhere to the compiled code's behaviors and functionalities. It's very rare to find games that have scripts available in a soft-code environment (IIRC the Stalker XRAY engine is one of the few, which is part of the reason modding is so common for that game) so if a modding community is successful enough, or the right, crazy talented people show up, and if there are certain exploits that can be utilized (i.e the game uses Miles Sound System, unless disabled by a programmer it will load any .ASI file located in the game's main file dictionary) The reason why Bethesda games have been so moddable the years since Morrowind is different however, I suspect that the ESPs and ESMs came about because Bethesda back in the day compartmentalized areas, and aspects of the game until they were finished to prevent unnecessary data conflicts between different area designers. Then when all the necessary content was done (or they needed to push out a demo), they would merge the ESMs/ESPs (into say Fallout3.esm) But such a system would also be handy for future expansion packs, and patches; so they kept it. Then when support for Morrowind finished, they released a stripped down version of their in-house tools because they realized that the simplicity of the ESM/ESP data structure lended itself very well to mods.

[@Dirty_Dumb_NAFO_Scum]

This is exactly why Operation flashpoint/Arma was so moddable.

[@Chris-lc3wi]

If you've been through the deep end with their modding tools, you start to see traces of this ESM/ESP approach just being a byproduct of a custom source control system. Collateral functionality, so to speak.

[@wesp5]

Great video! So did Troika intend Bloodlines to be moddable? Because you exposed all the items and disciplines in normal text files that could be edited at will, but at the same time, some things were hard-coded, like the actual weapon or discipline code. Still all the Python scripts and all the dialogue files were exposed too, and without this, things like the Unofficial Patch would never have been possible! You also included a lot of cut content in the game files that shipped. Was this for us modders or did you hope you could use them yourself in future expansions and patches?

[@CainOnGames]

I wasn’t involved in those decisions for Bloodlines, as I came onboard two years after it started. I don’t remember, but aren’t those exposed files just how the Source Engine worked?

[@wesp5]

@@CainOnGames I don't think so. Everything is packed in the HL2 games as far as I remember...

[@SaberVS7]

An interesting conundrum in the modern context is that the ease of access to tools (Particularly _financially_ ) is that a lot of the Talent that would've been making Mods in the past is now funneled straight into making their own games instead; And a lot of the people who do make mods for modern games often springboard into crowdfunding making their own game soon enough anyway. Myself, I've thought about making mods for Rimworld or DwFortress, but the logical-half of my brain pretty much immediately shuts-down such ideas as "That's time that I could be spending on my own game instead".

[@jaedavas3050]

I'm a big fan of Lua which makes it easy to just expose engine functionality. You can statically compile LuaJIT which is tunable and runs incredibly fast. Another thing ive been experimenting with is data oriented design, where it is common to use existence in a table as an alternative to if statements. Then, you have system functions run queries on tables, often in parallel when possible. This has its pain points, but the flexibility and ease of modding is wonderful. Id be curious to hear your thoughts on these kinds of approaches, and the pros/cons of making your own engine in general. As always, love the video! Thanks!

[@Me__Myself__and__I]

LUA is a nice and handy scripting system. Haven't had occasion to use it for quite a few years though. Think I'm going to reacquaint myself. I've built a game engine from scratch in the past. My suggestion is don't do it. I had no choice when I did it, now there are quite a few very good and inexpensive engines to choose from. Building an engine requires massive time and a lot of skill & knowledge. Time spent reinventing the wheel that you could instead acquire and then invest that time into the actual game itself. Plus, and many don't consider this, the existing engines are tested, supported and known to work! A custom engine would need to be extensively tested on a wide range of different hardware and OS combinations. Do you have time and money to build out at least dozens if not hundreds of different computers and test the system on all those? Doubtful. Existing engines are already tested and known to work on many different co binations.

[@faintsmile3279]

@@Me__Myself__and__I what about the learning experience you get from making an engine if making a game is not a goal for you and you are purely focused on learning? is it worth the time to make an engine from scratch or its better to focus on learning how to make a game instead of an engine? reason i ask this i always thought knowing how an engine is made is going to speed up your development because you know how the tool you are working with is functioning underneath.

[@Me__Myself__and__I]

@@faintsmile3279 If your goal is to play around with tech and learn stuff, then do anything you find interesting. Because the goal isn't to produce something good & profitable, its just to play around with tech. Now if you want advice, it depends on your skill, experience, knowledge and what your goals are. If you want to be a game developer - don't waste time building an engine. Might sound impressive, but a potential employer will be more interested if you built a few small games as a learning experience in the same engine they use. Engine development is hard. It requires a lot of skill and knowledge to do. Are you REALLY GOOD at software development? I'd say less than 1% of professional software developers have what it takes to build a game engine from scratch. And most won't ever land the kind of jobs where they would use the knowledge and skill learned even attempting it. Most game engineers do their work on existing engines (even if those engines are custom). For companies that have custom engines its usually a very small team of the very best engineers that actually build the engine. FYI when I built my engine it was challenging and I already had tens of thousands of hours of software development experience. Newer engines are even more complicated than they were then...

[@sarahshroom]

Oh wow. I was in the "surely making a game moddable has gotta be somewhat easy, so many of my favourite games do it" camp. I had _no_ clue it was such a thing. Very thankful for this video topic, thanks Tim!

[@Me__Myself__and__I]

Depends on the skill of the dev(s) making it moddable and it being done from the very start. It doesn't have to be all that much effort, but I get the impression many games try to bolt-on moddability later which is excruciatingly expensive and time consuming, see my top level comment for details.

[@picblick]

I'm so very happy you share your thoughts and experience! It's a mystery to me how you have not way more views.

[@Chris-lc3wi]

Hey Tim, Fallout mod author here. From my (and the larger BGS Fallout modding communities) experience, documentation is not critical as long as you expose your game data to modders and, if your game has scripting, an API ref chart for your script environment. If your players are dedicated enough they'll figure out how things are done simply by looking at what you and your team made and how they made it. Bonus thought re: data validation and fault tolerance: Modding on Creation Engine is weird in that way. In some subsystems you can feed it complete nonsense and it'll happily chug along, just gracefully handling all the problems the dev introduced. You can make the scripting VM completely nonfunctional by punching holes into the base class structure and the general "run around and shoot stuff" gameplay still completely works. This goes to the point where actual mod users become unaware of the game logic level problems they're introducing by combining badly made or conflicting mods because there's no "hard" problems like crashes or ingame error messages that make them aware. Working on Unreal these days it's almost comforting that when I do something stupid to the engine, it'll make me aware with a crash.

[@jundersplunkett2365]

I started learning 3D art back in 2016, started modding Fallout in 2017, I have no desire to work in game dev, its a hobby for me. PS, in the game making space, your my hero, thanks for all your contributions to RPGs, your passion shows, something that feels like a lot of companies are missing nowadays.

[@Kenionatus]

In my very limited experience of programming a little bit for an open source game, the _can_do() -> do()_ pipeline isn't just nice to connect external code to the game, it's also nice to make the code internally more extensible in a very clean fashion. It's an over 10 year old code base by now with probably thousands of contributors and I'd say that's one of the pilars that keeps the code readable.

[@Dullahan3470]

I always figured that game data was often exposed and modifiable for development purposes moreso than explicit modding support. It felt like most games were accidentally moddable and when PC gaming waned in the mid 2000s they started to deliberately lock down and prevent modding for DLC reasons. (Prevent pirating of DLC and/or player made skins competing with paid cosmetics). A lot of times back then the DLC would be specially encrypted even if the base game wasn’t. I know some games like Company of Heroes where they added more explicit modding tools and support but actually made the game much harder to mod - instead of being able to override everything you can only modifiable a shrinking list of explicitly allowed things.

[@iokuu]

On some level, modding is about being able to make changes to the game without recompiling it. The more the game exists in purely compiled code, the less modifiable it is. And there are definitely advantages to being mostly compiled! It's faster, after all. Computationally, and conceptually. Conversely, being moddable also means being able to iterate without compiling, which is useful for game development as well as modding. There's a push and pull.

[@Me__Myself__and__I]

Isn't PC gaming larger than it has ever been? Modding competing with DLC is almost certainly a factor. I imagine quite a few games intentionally do not support modding for this reason. It is a legit consideration from a business perspective, though personally I'd probably always choose to include modding support. I do expect a lot of games are in fact "accidentally" moddable. I think a lot of game teams find it a lot easier to have configuration data read in from text files somthat non-programmers can work on it. Which (probably unintentionally) then lends itself to moddability. Such games often don't directly allow for custom scripting and other more advanced modding. But somemetimes there are kinda standardized ways for modders to add that capability for reusabile engines like Unity.

[@Dullahan3470]

@ It is today - it wasn’t in 2005-2010 when modding started to really disappear. There was a significant resurgence when Steam started to reach critical mass in 2010-2012 and developers had to respect the market again.

[@MorbidEel]

There should be a distinction made between moddable and "supports modding". Anything on PC that isn't doing DRM stuff should be moddable but only some supports modding.

[@FlawedFabrications]

Easy moddability was always a key goal for the little games I used to make because I assumed that the easier it was for someone else to change it, the easier it would be for me to make it. For instance, JSON files made it incredibly quick and easy to add or test things. Sure from a technical standpoint it might have been more performance-friendly to have all those statistics and so-on directly in the code instead of being read into the program, but I didn't need to worry about performance and being able to make pretty significant changes by opening a plain text file is really nice.

[@trucid2]

Being able to configure things without modifying code is just good design.

[@chrisdistant9040]

@@trucid2doesn’t always work though. Eg you usually can’t implement via JSON new logic or ai behavior that isn’t already supported by the game, etc

[@trucid2]

@@chrisdistant9040 Sure you can. Code is just data, and you can load it through JSON or XML. Some games like Space Engineers even allow you to write programs in-game. Though whether you want to allow this is a different question. I was talking about tweaking game variables and such through a config file rather than having them hardcoded. That's what I was calling good design.

[@chrisdistant9040]

@ lol you have to do a lot to make something like that happen though. If “it’s possible “ is your standard then every game is already moddable, just hack the binary bro

[@penvzila]

"just". Lol. That's always what a good idea fairy leads with. "Why don't you just". There have been times however, when we had somebody who didn't really know what they're talking about ask a question like that and it ended up being a really good point and we made the change.

[@nairocamilo]

Like Mythbusters!

[@metarenegade]

“Just” might be the most harmful word in the English language.

[@Rodhern]

Indeed, most times when I encounter "just" it has a particular meaning. It means, "This thing would logically require a lot of implied offshoot work. 'Just' do the change in complete isolation (for now). Please don't worry about the stuff that breaks! I will be back asking for the entire rewrite soon enough. And that rewrite will be quick because I used the magic word 'just' in my initial work item request."

There is an exception to every rule of course. In general though (and in my experience as well), Mr. Cain was absolutely right here.

[@erikfjeldstrom8779]

My first experience with modding was with the Command and Conquer (and Red Alert) games. I was amazed at what could be done with simply changing a few data points. I was also learning programming at the time in high school and started realizing the immense complexity of making a game, and how my brain doesn't work that way. My hat's off to all you game devs!

[@spudd86]

I've seen a lot of indie devs say mod support has saved them work in the long run by making developing DLC easier. I would also expect making item definitions mainly data driven would have benefits outside moddablity. For example letting non-coder designers tweak things. To me it makes sense to design for modability even if you don't intend to deliver it. At least at the base design level, since as you said adding it later is very hard and it also let's you use the mod system yourself for DLC or expansions.

[@HeinerGunnar]

It's kinda unrelated to your comment, but your last sentence reminded me of how modding works in Mass Effect 2 and 3 inverted to that: modders hijacked the system the game uses to read DLC, so as far as the game is concerned, all your mods are simply DLCs not made by Bioware

[@tamathacampbell4985]

Thanks for all the useful and actionable game dev and design information you share... these videos are stuffed full of great stuff.

[@Spiderboydk]

Great video. There are so many misconceptions about moddability out there.

[@disky01]

It can be extremely worthwhile though. So many of my favorite games lean into modding, and if they didn't, I might not even be playing them. I've put almost 1,200 hours into Mechwarrior 5 and I've purchased every DLC for the game, but I cannot imagine a reality in which I would play it without the mods I'm running. It just wouldn't be the same game. Same with Skyrim, Cyberpunk, Baldur's Gate 3, Doom and the Bethesda-era Fallout games. No disrespect to the devs, but without mods, to me those games feel incomplete.

[@zhulikkulik]

Imo, Bethesda just relies on modders at this point and Starfield showed that people aren't okay with that anymore. I've never been a fan of their games in general, but I've tried and failed to understand how can people play something like Fallout 3, 4 or Skyrim without major mods. And it's especially bad in 4 because the vanilla game is literally unplayable by my "standards" - bad story, bad writing, bad roleplay variety, bad gameplay. Just a bad game start to finish (didn't actually finish it, never got past killing Kellogg). Throw in some breakable doors, ammo and stimpak weight fixes, item weight and inventory capacity fixes (i'm yet to find a "system level" fix for infinite containers), lockpicking and hacking skips (iirc they allowed to try to silently roll and open/unlock if your perk matches the requirement instead of that minigame that gets boring when you unlock your third lock), XDI with silent protagonist, prkf, damage rebalance, mod that brings back criticals - and you still need to fix story and writing. So far the only mod that does it is FoLon and even that has issues. I am surprised that it's the same studio that made Daggerfall which ONLY requires compatibility and stability patches (basically just Unity version with some options turned off) to become an enjoyable game.

[@Solus749]

Depends on what the game wants to be, if the game wants specific story/char beats they lock those to be unomoddable. If they want a specific combat experience they lock those parts of the combat to be unalterable. At that point you as a modder is actually breaking down what the dev intended the game to be. That is before we get to potential mods breaking engine data flow causing softlocks or memmory leaks. As Tim said those checks for mods slow down gameplay and gamedata reading. Baldurs gate for example don't need mods it is complex enough as it is and a mod attempt done wrong can break a long chain of code. Bethesta games or BUGthesta games as I like to call them don't really care if you mod the shit out of them as their code is spaggetti code allready. DOOM games, especially the first ones have map editor and even cheatcode structures built into the game base so the data flow don't get broken anyway. That means that you can add a whole act and rebalance enemies, ammo etc to your pleasure without breaking the engine. I say rpgs in general don't need mods ( no I don't count skyrim/fallout 4 as rpgs btw ) but shooters, racing games and a few more genres can have room for them. Why I don't count skyrim as a rpg is simple....you can be part of the thiefs guild and the assassin guild inner circle and they are in a invisible bloody war. You shouldn't be able to be a member of both to be frank and that isn't the only situation where rpg aspects break ( your choice SHOULD carry weight ..aka some decisions will lock you out of other ones). As for why fallout 4 is a horrible rpg is simple, you only get yes, dickish yes, no or ignore and your choices don't have hard checks until the endscenes really...which makes it fail as a rpg. Jrpgs are different in that they allow choice sbut of a different kind while preserving story. Also a reason you VERY seldom make your own char in jrpgs as your char is intervowen in the story they want to tell. Japanese devs also resist modding for another reason, why buy a game if you have no intention of playing it unmodded, is their work worth that little? Basicly put your desire to mod something can be seen as critique of their ability to make a quality product.

[@theobell2002]

I can definitely see myself playing Cyberpunk without mods because the game is actually good. Unlike Skyrim.

[@Solus749]

@@theobell2002 Baldurs gate 3 don't need it either if I am honest, it is deep enough as it is. RPGs in general don't need modding as it tend to fuck up story and combat balance....again exeption for BUGTHESTA fallout series and skyrim who I don't consider rpgs.

[@PedroGomes-cx7ku]

@@zhulikkulik This idea that "Bethesda just relies on modders" for their games to be successfull just isn't true. Skyrim's golden age was from 2011 - 2016, and most of the player base was on console and back then there wasn't mod support on consoles. There were entire websites dedicated to talking about Skyrim (and early Fallout 4), like The Skyrim Blog/The Tamriel Vault, where people shared character build, discussed lore and story, shared stories etc. KZbin channels grew to their first millions sharing character builds and talking about lore too. Even Starfield is by far their most stable game - its issues are elsewhere, and there's no mod available that has been abe to do what I wanted from that game

[@sieda666]

Thanks for the insight as always! Modding is such an important feature to me in games, I'm thankful when devs go out of their way to facilitate end-user modability. All of my favorite games are moddable to some extent, from Bethesda titles like Skyrim, Fallout 4, Starfield, to Paradox Strategy games like Europa Universalis and Crusader Kings, and then other moddable sandboxy games like Dwarf Fortress, Rimworld, or X4:Foundations. I've put hundredss if not thousands of hours into each and its thanks to mods.

[@PedroGomes-cx7ku]

While I would love for all RPGs to have the level of modding that BGS games do, this video made realize that they've been working, building and iterating on their engine and modding kit since the early 2000's with Morrowind - and even then, starting with Skyrim, they stopped releasing their mod kit at launch and had to take increasingly more time to release it with each game. It's one of the reasons I'm so skeptical/reluctant when I see people asking them to switch to UE - I imagine the work it would take for them to adapt the engine and to release a modding tool as extensive as the GECK/Creation Kit would be mind boggling, and I fear it would take some time before it was on "equal grounds" with the tool/engines that they've worked on for decades.

[@MrSnivvel]

BGS (and others who roll their own game engines) would have a higher upfront cost in converting to an OTS game engine but have a much lower long term cost, but there's also the costs of pride and the will to do such a conversion that's hard to financially calculate. BGS's forte is definitely not in making a game engine.

[@saffral]

BGS' engine is a fun example because of how modding can be set up. It's all one giant database and mods will add/remove/modify the data in it. A load of engines are built around "levels" or "scenes" that can be mostly separate from each other. Like if you look at creating a level in Doom you get a WAD file, if someone else wants to change that they need to make changes to that WAD file and then distribute the entire thing which then wouldn't be affected by someone else's changes to the level. But because Bethesda uses a database, one modder can create a level and then another modder can change things in that level and release only the changes, and those changes would still be made even if another mod changes things or if the original mod updates. It's more flexible and robust than overrides tend to be. Each of BGS' DLCs are effectively just mods that operate the same as community made ones, so it doesn't just benefit the modding community. Side note about the modding kit at launch, with Morrowind they were rushing at the last minute to get the CS out the door, it was so late that it couldn't be included on the game's CD and they had to create an entire second CD just to include it. I'm unsure about when Oblivion's CS released or if it came with the game or had to be downloaded, but Fallout 3's GECK also released over a month after the launch of the game. The tools have been lower priority for them to release than the game itself, and trying to get both ready for a simultaneous release is more work than I think they want to do.

[@saffral]

@MrSnivvel Yeah, a big upfront cost and then 5% of all future sales. How much of their current sales do you think goes into engine development and how much would go into customising Unreal? I'd wager than Bethesda has made a heck of a lot more from their programmers who worked on Skyrim than they'd have made from paying 5% royalties to Epic over the past 13 years. Of course if you're expecting to publish failures or less successful titles 5% might be a viable trade-off, but I think there's a few reasons why Bethesda got rid of most the middleware they used to use, and increased profit margins is one of them. The real cost of maintaining their current engine is probably more visible in training new hires in its use instead of the engine(s) they learned in school, or in trying to make a multiplayer game in it. Because no matter how much the engine was originally designed for MMOs, Bethesda had not put effort into maintaining it's network functionality over the years.

[@MrSnivvel]

@@saffral There's open source game engines like Redot/Godot that have no royalty requirements.

[@saffral]

​@MrSnivvel Absolutely, but it doesn't exactly solve the issue of them needing to build their own engine and toolkit on top. That's what they did with the last engine they licensed too. The only reason fans ever give to switch engines is "ooo, Unreal shiny" or reasons with minimal evidence like "the engine is the source of the bugs in their games".

[@bradandsarahibbard1831]

Wonderful video, Tim! Thank you!

[@drew_echo]

Another issue with tools is licensing. Either the company may have to pay per seat which means they can't distribute it to modders, or you can have poisonous licenses where you can't even release an Autodesk / Blender / Adobe plugin because it would contain _knowledge_ of some proprietary architecture (like Havok or some licensed file format) even though it doesn't contain any of the licensed code. And then you have to deal with maintenance: Bethesda released a 3DS Max plugin for working with their mesh files for Skyrim, but it was built for 3DS Max 2008 or some other old version which no one can acquire / uses anymore (and also has the same problem as before where modders have to go buy 3DS Max). One way to make modding a lot easier for games is to distribute the header files (but not the implementation) for your game like you will get for Console and other proprietary SDKs. It makes it a lot easier for modders to mod games without that much cost to the company, though lawyers and pencil pushers who don't understand source code will usually object. CDPR did something similar for Cyberpunk 2077 and released the memory addresses of key functions that are useful for modding for each game version, but it wasn't as comprehensive as modders would have liked. There is _some_ work though like making sure you're careful not to include anything sensitive in your headers and separating out any licensed code you're using that you can't distribute. That said, like Tim mentioned at some point you're just being nice. Bethesda didn't give modders anything for C++ mods with direct access to the engine, and the modding community reverse engineered a lot of it on their own. But, given the difficulty, you're not going to see that happen for games that aren't very popular / using routine engines. However, not providing modders enough tooling can also hurt the company: Bethesda added a bunch of mod monetization to capitalize on their modding community, but that changed the binaries just enough to break a ton of mods, so a large amount of players have been downgrading their game to keep playing with their mods. If Bethesda had put a bit more into what they exposed to mods, they might have been better able to get people onto their new changes + monetize mods better.

[@realizewave1513]

This sounds so ridiculous as someone who mods non-Bethesda games. The difference between what Bethesda simply exposes and what most companies do is night and day.

[@Prorock23Rus]

@@realizewave1513 May you list examples of other "most companies" that do that "night and day" difference, just my curiousity

[@Me__Myself__and__I]

This is easier than most software engineers make it for a simple reason. Being good at writing reusable code, frameworks and making systems modular and configurable is a SKILL. Sadly that skill is rarely taught to developers and most don't put any real effort into getting good at it. Which is why doing this for most devs is hard and time consuming, substantially more so than it needs to be. This is probably my #1 area of expertise. I've written more reusable code and put at least 10x more time into making software modular and configurable than any other dev I know of. So, from my perspective, this really isn't so difficult or time consuming. But sadly that is not typicaal. Example. Once upon a time I wrote a game engine from scratch for a startup game company. I built that engine from the ground up to be modular, extensible and configurable. Sadly it never made it to release because they ran out of money to pay the 3D modelers and art staff. But modding that game would have effectively been the exact same as what we were doing to build the game. The extensibility would have been exposed and I actually consider that as having taken ZERO time or money to accomplish because I built these capabilities for the dev team but in a way that they could have also have been used by modders. So moddability was basically free as a result of my designing the engine to be modular and extensible for our own use. Tim is 100% correct about doing this EARLY! The engine and/or game must be designed to enable this from the very beginning. Trying to make an existing software system (game or anything else) modular and configurable later on is EXTREMELY costly and time consuming. There is no way around that. But building systems to be modular, extensible and configuration driven does not need to be time consuming or costly if done from day 1 by someone skilled at it. I have so much experien e in reusable software that writing anything to be reusable/modular is effectively free for me now. It is how I think. It would literally take me longer to build something that wasn't because I don't even know how to think about code as non-reusable. The vast majority of all code I have written for many years is reusable. Its a skill and IMHO the most valuable skill for a software dev to have. It is a skill and a mindset, similar to how being good at OOP requires a different perspective. [Side note, OOP often makes reusability and modularity easier] Once you get good at writing reusable code it will multiply your productivity as you build up a library of modular, efficient and known good code. It can also mu,tiply the productivity of entire teams. My libraries and frameworks have been credited for saving teams many man years of time and effort. Its a VERY useful skill that I recommend all developers invest in.

[@Kenshkrix]

Yeah if it's the plan from the start then the work is straightforward. If I had to guess, many teams avoid this approach because they probably aren't experts at these skills on top of how this approach starts up more slowly and takes awhile to get to the more visible/fun parts of development (where you can actually see the progress happening and play a prototype). Slowing down the startup but saving time/effort in the middle and late stages of development doesn't spark enthusiasm or passion the same way early prototypes and hacking stuff together asap does, while the pain of the 'move fast and break things' approach doesn't happen until later when it would also be a giant pain to go back and refactor everything. I see a lot of devs/teams post something like "we'll look into modding after EA/1.0", which is sort of like saying "We'll look at the foundation after we've built the house".

[@Me__Myself__and__I]

@@Kenshkrix Agree 100%. I've seen that first hand many times. I know better due to experience but the majority of developers & managers don't want to put in the up front effort because they don't realize who much difference it makes. Because most don't have the specialized skills and experience...

[@HenrikoMagnifico]

My cousin is a huge fallout 1 fan, it's so cool you made it!

[@evanjellydonut9461]

been modding games for 20+ years (including bumbling through reverse engineering audio formats). Nice to see this topic from a developers standpoint.

[@en7622]

Great timing considering all the drama about halo mcc and infinite mod community drama.

[@mopugnothee]

I have never seen anything friendly moddable like the Aurora Engine, nwn was a treasure for modding, I also think it's easier to make moddable an engine developed by the studio itself instead of adopted by.

[@MFKitten]

It all depends on what you count as "moddable". Sometimes the game just has all the assets accessible and you can replace them yourself. If you want "intentionally moddable" like the Source engine where the tools all were released with decent documentation etc, then obviously that's more work!

[@635574]

Even things like reading values from a text file need specific code to process all that. GTA has a lot of text files, and one huge handling file for all vehicles(everiting is explained there), but not as much modability for gun values (just XML file nothing is explained), it had the same system since gta3. Finding why sonething crashed was not happening last time I was a dumbass GTA moder, before GTA5 came out and I wasn't able to run it.

[@Ulfric42]

I was designing a MMORTS(still hasnt been done to date) I was wanting to create a mod market. Essentially people could sell approved mods. So the creator wins, the company wins, and the players win.

[@veraxiana9993]

Normalizing my sleep schedule finally paid off, I got to a new tim video within the first half hour of its upload! Lol

[@Tom-f4d6l]

I was under the impression that making things moddable was generally a good thing since it required tools to be better (more data driven, scriptable, more useable, built to higher standards, etc). Even if developers didn't want to license the engine; having a bunch of brittle tools reliant on other systems seems like it would be an issue. Like if a team stops or has to pivot because 'Dave' is taking a long weekend (or worse yet he can not take a long weekend) or the whole studio stops because jira went down; that seems like it would be a long term liability. I know large teams do work like this but it makes me wonder; Are developers making games because of their tools or despite their tools?

[@romanshatalin7077]

Merry Christmas, Tim! As a software engineer, I completely agree with your points. I will only add that there is a trend of indie games designed specifically to be moddable at the expense of other aspects. Many of them go with 2D or isometric mode which also makes it way easier for modding compared to full blown 3D. Quite popular examples of that are Factorio, RimWorld, Project Zomboid which are all quite different genres. Less known game, Vintage Story, actually started as a mod to Minecraft but they got tired of dealing with updates breaking everything and built their own similar game which has amazing mod support. Still, far more popular AAA games have other goals but after enjoying the ability to overhaul a game as I want they are less and less interesting to me.

[@77Arcturus]

I love mods especially the incredible system created by Bethesda for their open world games. Its dreamlike how they have made it possible since MORROWIND to jump right into the guts of the game and change everything from world behavior to audio/graphics. I still remember taking the sound files from Bradley's rpg WIZARDS & WARRIORS and creating a talking demon sword in MORROWIND even changing things like the look of the sword while doing it. Probably the easiest editor to use alongside the old UnrealED for the fps shooter UNREAL TOURNAMENT. WadED was also pretty good but was a bit more like a CAD design program. All these editors especially the one by Bethesda taught me so much of the what, where, why and how of game structure where i found myself not even using editors years later because of knowing where everything is and what it does. This is another great thing of when Bethesda games first come out there is no editor out yet so everything has to be done on a manual level til Bethesda releases the modding tools/editor so its one more way of learning how things work. Pretty exciting and magical. Who would have thought way back in the 1990's playing THE ELDER SCROLLS II DAGGERFALL one day i would be able to transform my favorite games with aspects of my favorite books, movies etc etc. Thanks for the great video with information i had never taken into account making me appreciate even more both the base game and mods along with those who make such technical miracles possible. Cheers and happy holidays ☕

[@kmg9763]

Tim, Great video as always. Do you have a Good "I told you so" story?

[@johnmcconnell7052]

I tend to go look at tf2 it originally was a mod now look at frostpunk. Some games are designed for mod support. Then you bring up halo 2 dear god that game is held together by ducktape and prayers.

[@Pepperham04]

Then you have a game like STALKER, which is also held together with tape and prayers, yet people still mod the hell out of it.

[@thisorthat629]

​@@Pepperham04 tbf duct tape and prayer games generally fare better ir long lasting/in depth modding. personal theory 1. casuals enticed into modding due to needibg fan patches, then trying modding visuals, then quests, then gameplay 2. a lot of successfull duct tape and prayers/cult successes are tjode for reason ie they have that "special" sauce, which entices mof creators eg compare fallout 3 and fallout nv. fallout 3 has it's fans, but nv is another lvl

[@Luckmann]

Duct tape.

[@SethStoker]

Hey Tim, I don't know if you have any experience with this, but what are your thoughts on in-game items with real money? Team Fortress 2 has skins you buy from other players, the infamous Counter-Strike gambling stores, or some of the Fallout 76 item stores. Basically, player stores are from players, not player stores from the company. I don't know if I explained it well enough, but what are your thoughts? Also Merry Christmas! :)

[@BetweenTheBorders]

Example: I was the only one working on our engine in MonoGame, and I did want to make it moddable from the ground up. The ability for a script to call for a map file name was easy enough to do, but the error handling wasn't trivial. Every level, every script, every image was a potential point of either failure or abuse. But since this was data driven, the maps and events were moddable, but only to do the things that had op codes. Loading assets, on the other hand, was a huge pain. MonoGame expected certain data structures, so I had to load each and every file, process it, and store it in a data structure. I fixed the huge issues in this by saving the structure as an archive and loading the archive if available, but otherwise would rebuild and save the structure at runtime. But all of this was only available because I designed in a rather bloated way from the ground up so we were focusing on modding versatility (string names rather than enums, for example). Of course my way also opened the door to casual theft of assets, which wasn't optimal. We may have used an override system in the end, but the team falling apart is another story altogether. At the end of the day, it would've been pretty moddable, with the asset loading and Harmony to edit C#, but we'll never really know.

[@lennysmileyface]

Theft of graphics assets is pretty much unavoidable anyway since there are programs to rip all of it straight from memory.

[@BetweenTheBorders]

@@lennysmileyface Hundred percent correct. Unpackers and even AES decryption of assets are very common. I mean, you're letting code run on their machine, and it needs to decode whatever you've done. But no point making it as easy as looking in a directory.

[@bratttn]

great insight as always. Once you go moddable does it stay forever in your approach to the game's structure? Like kind of "you see a password you automatically store it in a secure data structure" reflex?

[@courier7049]

Another hidden aspect is when developers/designers are out of touch, like in one interview with Boyarsky and Cain when asked about modding Outer Worlds answer was something like "modding is great, but for OW it doesn't fit because it would mess with creative design of the game", now looking at story and characters design the game has is clear what was meant with that. In same vein why first version of Bard's Tale 4 (which was crowdfunded as I remember) had open UE paks, while in Director's Cut were encrypted. No one asks for developers to make first class support for mods, just that that they don't actively put barriers for changing the game.

[@doublejesusful]

I'd like to add that if a community wants to, they will find a way to mod your game whether you support them to or not.

[@monkian]

By the way you replied to the fictional manager about "just" doing something we can tell you have had that conversation more times than is funny. :)

[@zharyel9890]

No matter the cost, moddability is a long term benefit to the game and its reach across years, even when service updates end.

[@KryyssTV]

This is why Epic is building scene graph with native support for the Verse scripting language. At it's core this means that once UE6 launches it will permit every game to have automatic mod support because the devs can choose what parts of the API are exposed and that will determine what within the scene graph can be modified. Now, since scene graphs can contain scene graphs, a developer could just expose things like the mesh and materials of an NPC but not sxpose the API on the world. This is all done within the UE editor which means no custom tools are needed and creates the kind of mod support you see with Bethesda games running ObjScript/Papyrus.

[@slBrelaz]

Yup! I've been working on a project outside of work (custom engine) and it's taken me 3 years to add an intricate modding API that can reliably replicate itself over the network. Still need to update all the documentation for it though.

[@MartinWoad]

When I choose what game to play, I instinctively select good ecosystem over good baseline game. Don't get me wrong, I really like Minecraft, especially the Beta era versions, but it's not like I would play it if there was absolutely no way to modify anything about it. And even if I don't install any mods, I just get a lot of enjoyment from the fact that I am playing something that I can rather easily modify. Same thing with Skyrim, Fallout NV, Witcher 3 or any other. I cannot imagine myself playing something with a hard expiration date on the amount of content it provides, even if I feel like I would enjoy the core gameplay loop. I need to feel like I am purchasing a tool and not a commodity. That is also why I don't play console games at all. No modding is a no-go for me.

[@roberteltze4850]

And there are secondary costs particularly in multiplayer games. In World of Warcraft mods give players a lot of customization but they also allow gold farming bots. It also makes it harder for devs to balance raid bosses. I'm convinced it would be a better game without mods.

[@635574]

How the fuck did that game not die, they allowed modding in MMO?

[@MartinWoad]

That is why I hate MMOs and instead prefer an ecosystem of private dedicated servers with the server software released for public use. This way players can choose whether to play in smaller communities which are much more lenient about the use of addons or to play on more restrictive bigger servers which can restrict the API use or even come with some anticheat.

[@EnneaIsInterested]

OpenMW's LUA-dehardcoding is a fascinating example of a written-from scratch game engine replacement gradually switching from data-based moddability to method-based moddability, in essence, this is something an actual capitalist firm would NEVER DO, but OpenMW may well become the go-to open-source cRPG game engine of the late 2020's because of this shift.

[@yellingintothewind]

The answer is heavily engine/language dependent. If it uses an intermediate language (C#, Java), it is sufficient to not actively obstruct modders (or do so, and then publish the obfuscation map, which is how Minecraft works). If the game does _not_ use one of those languages, then it mostly comes down to if your internal tools are suitable for public release. A good example of this is the original Evil Genius game, where for fast iteration internally, they gave the game the ability to read overrides at runtime, and then at community request, published those. The problem is the low level modding you get in Java or C# simply by not obstructing depends on the symbol names still existing. This is the default in those languages, but not the case in C/++, where code gets inlined, optimized, stripped, and ultimately cannot be trivially turned back into the original form. C# and java games are de-facto pseudo open source, as you can get something closely resembling the original source code with off-the-shelf tools (at least assuming no intentional obfuscation is in play). Which brings up the simplest way to support modding. Provide the source code under a suitable license. This can be like Turning Wheel and some other companies do, where the engine is copyleft, and the game _content_ has a standard game license. Or it can be under incredibly restrictive licenses, so long as they don't interfere with unrelated projects and work. This _does_ require some additional effort, on the part of your community managers, in order to leverage the community into a beneficial form. Ideally you want this to result in suggestions and PRs that _reduce_ the workload on your core team, but that requires skills that many closed shops don't really have.

[@brianviktor8212]

While modding can be a great benefit to the game, I feel like it's only there to compensate lackluster parts of a game. The simplest example is Skyrim. There is a mod I use to make alchemy useful and better. Why? Because the standard alchemy is bad. There are mods to make combat AI better. Why? Because it's bad. There are mods to accelerate the start of a game by simply spawning somewhere randomly. Why? Because it's tediously long. These are all matters which the game could work on somewhat later. Instead there is like the 20th Skyrim edition which basically has no changes anyway. And of course plenty of bugs that remain unfixed by the devs, only by modders. I think there are mediocre games that become great if mods can fix its shortcomings. But it's like an open invitation to have the game be fixed by random people.

[@nihren2406]

The timing on this video is incredibly perfect with the rumors about the Elder Scrolls IV: Oblivion Remake. I've seen a lot of people concerned about the moddability of the game as it's either a vertical integration with Gamebryo on the bottom providing code, physics, etc. with Unreal Engine 5 on top for graphics, or it's entirely rebuilt in UE5. Both of which have their concerns. How does modding work when you have two incredibly different engines impacting the game? And by that same token, how deep can that modding really be? The same applies if it's exclusively Unreal Engine 5, just how much can the developers open that up to modding? How easy would that be to do on the level that the Creation Engine offers? An Elder Scrolls game without in depth modding on the face of it sounds like a failure waiting to happen because Bethesda games are associated with basically being able to make a new game within the mod tools.

[@AdioShu]

Hey Tim, what do you think of the decline of physical media in video games. Could you give both the perspective from you as a gamer and also as a game developer.

[@velorama-x]

The single biggest regression in gaming happened when games moved from a physical to a digital presentation. No digital game is as easily and completely moddable as every physical game in existence. A game is what/how players choose to play, not what game designers want players to play. Moddability is an integral part of the gaming experience (this is especially true for RPGs where physical games evolved as far as GURPS). Unfortunately almost every digital game these days comes with the equivalent of a straight jacket for players, albeit for various reasons: monetization (throttling player progression, selling customizations, etc.), tight budgets (as mentioned), game designer hybris ("players don't know what they want", "my vision is more important than your experience") and player gatekeeping ("only one difficulty"). Now, we can debate the economics of adding moddability as much as we want, but we should at least be able to agree on the chilling effects of patents: beyond the moddability of a single implementation, they also try to put a lid on the creation of game implementations with a modded but similar ruleset.

[@klausklebstoff8240]

That's a good video! I always wondered why not every games comes with a modding tool

[@Main267]

In some instances, modding is 'free', like with Unity Engine games. "Just" don't use IL2CPP and obfuscate the code, and modding frameworks already exist and are easily extensible to the game.

[@andypanz]

i already forgot when devs made their games moddable from scratches. But stalker 2 latest example of exemplary moddable game and they even provide tools yet.

[@ashuggtube]

“Illegal script code. See Dave.” Goddamnit Dave 😂

[@georgedeligan133]

That's kinda amazing that most actions in Arcanium were Command driven.

[@-C69-]

You mentioned the C# compiler… I want to create a game that converts Scratch scripts into C# or GDScript… the game is about a moddable robot, where the player can edit scripts to make the gun fire faster, hack door locks, deploy viruses (that serve the story)… would you recommend including a client-side compiler to achieve this?

[@FathDaniel]

Interesting insights. Another huge problem with making methods visible approach - how do you ensure mod doesn't run a crypto miner or hostile code? Basically what if mod is hostile?

[@CainOnGames]

That’s an interesting question, and one that we never considered until recently. If modders are restricted from internet access and can only write data into game save files, I imagine that would prevent a lot of abuse.

[@Me__Myself__and__I]

Two obvious approaches I can see. 1) The environment the mod runs in can be sandboxed as Tim mentions, limiting access to things such as networking. At least some sandboxing is probably desirable. 2) Consider it not the dev team's problem. People should be responaible for the software they choose to install on their own systems. While sandboxing attempts to protect people from their own choices it also limits opportunities. For example there is a mod for Cyberpunk 2077 that adds in multi-player support. That would be impossible if the modders couldn't access networking. But also, there are often 3rd party mod enablers for the common reusable engines that allow for things like adding scripting to games that don't natively support it for mods. These in theory have no restrictions (sandboxing) of what can be done with them. So even if official modding support had sandboxing it wouldn't be difficult for a malicious mod author to say "this requires you install the script extender". Ultimately when people choose to install software and mods on their computer they have to be responsible for those choices.

[@lennysmileyface]

@@CainOnGames Also probably don't let people run random dll files, Cities Skylines 2 found that out.

[@mark009vn]

My favorite past time is to open up a game and mod it, mostly to mess around or to experiment with different game systems and see how it would work with a pre existent base. I would say 7 out of 10 games I messed with never had modability in mind, some are definitely never meant to be messed with as they are made of ROM files where the only way to edit data is to use a hex editor. I would say modability is a buzzword, as long as the means to interact with the engine is available to you, that's all the modability a game needs. Fallout 1/2 is a good example because modding wise it is complete arcane nonsense, and it take the community to develop a lot more arcane engine modifications and tools to make it serviceable, but when you look past all of that it is still modable and people have made very impressive total conversions with it. Not to say I don't want games to be easier to mod but more that there is a difference between ease of modding and actually being able to wrangle the game to do what I want. The best thing devs can do for modding is to not care about modding at all and just focus on making the content creation pipeline of their engine/platform more robust, which is a direction many companies went down anyways as they continue to stick with their bespoke engines and making it easier to work with or using proprietary engines that are easy to modify by design.

[@AlucardNoir]

Modding is relative. I remember modding Caesar III back in the day and that was just changing a few values in a end user accessible file. Most mods for most Owlcat RPGs are just portrait mods. What counts as a mod is relative. Is the Dungeon Master Mode in Divinity Original Sin 2 a moding interface or it's own separate thing that comes with the game?

[@astralgazelle]

Hey Tim! In the era of advancing artificial intelligence, do you think it will pose more of a threat or be more of a help for developers when creating games? Do you think there's a risk that some positions might be replaced by AI? There's a lot of talk about AI-generated graphics, but my bigger concern is AI generating code.

[@CainOnGames]

Artificial Intelligence kzbin.info/www/bejne/bX61qZKfbsypjdk

[@JessHawk]

Do you think it's time and money that prevents Bethesda from making their updates not break the script extender on all of their games?

[@leafstormblaze9449]

Some of my favourite games have been moddable and not just bethesda ones, many retain community's far more than most single player titles which is a nice thing for your game, rim world, project zomboid, kenshi, darkest dungeon all continue to get played because or the communities which build stuff for them, while it is hard its definitely a good thing. Classic tombraider for example still lives thanks to the level editor it was given with back in the past and has 5000 plus custom creator content.

[@AyaKho]

Most of these games benefit from designing for moldability to start with. That factor alone reduces much of the difficulty. Not every game does this and not every game should do this.

[@leafstormblaze9449]

​@AyaKho understanding but I do for sure know I come back to these games more.

[@AyaKho]

@@leafstormblaze9449 I respect moddability and I would be lying if I said I do not appreciate its fruits. The teams of modders on Morrowind behind the Tamriel Data projects like TR and PT have effectively quadrupled the size of the game with content on par, as dense, and sometimes better than the base game. Yet I also wonder if gaming as an art is hindered by people chasing these hundred or thousand hour experiences rather than trying new things.

[@LethalBubbles]

personally, I would ignore it. they're gonna do it anyway. but if I wanted to do like Microsoft and Valve does, and commodify the creative output of people (who may not realize what they've signed up for in the EULA); I'd probably make a modding platform and claim copyright over all that data. Modability isn't always what it seems. What I do like though is the old school approach like WAD files. those are just normal files that anyone can get into and play around with. There's a great talk by Will Wright way back in the 2000s on the value of user created data. It is totally worth looking up. although, that's how you end up with really tasteless stuff. nude mods, maps of people's high schools and stuff, with your game's logo on it like a seal of endorsement. So tread lightly with mods if you don't wanna deal with all that. But like I said, if they really want to they're gonna do it anyway. But that's the difference between 1 really difficult community mod or thousands of things someone threw together without much thought. either can be good or bad things imo. If I did want to make a engine highly moddable, what I would do is make that the product I am selling and the game would just be the trojan horse for selling the engine. Like how Burnout was a tech demo for the RenderWare engine.

[@Jaqinta]

Hi Tim , First of All Merry Christmas to you and everyone . About this topic , like for example in first 2 Fallout games some of the games features can be reachable with Star Trek Scripting language , i might be assuming wrong about that , but would that feature helped to divide some of the works into other groups , so this shortened production time . But i think , that scripting language was already written before Fallout . So simply if Fallout Team didn't have such tool would that be logical to write it by scratch or maybe write a really simplified version so making this tool as more efficient as possible ? P.S: As for Orc picks picks a item and takes damage thing , rather than use it's actual purpose , Orc use to scratch it's back -1 hp but +1 relaxation maybe .

[@CainOnGames]

That is the pros and cons of that scripting language. True, it saved us time to use the ST scripting system, since Fallout scripters could get started sooner. But they also did some things with scripts that caused problems later, usually by checking conditions that were insufficient for the result, leading to events triggering inappropriately.

[@PedroMDIX]

I think is worth the extra effort, Fallout, TES, Asetto Corsa, are examples of single player games that are still being sold and being played by a lot of people because of their healthy Modding communities.

[@eugkra33]

This explains for Starfield is the way it is.

[@thedude7319]

11:11 so software dependancies ? It shouldn't be so difficult to re-route or has jira a reading/writing system that isn't open to be viewed ? Because I am viewing it as error, error data send to jira and not. Error has occured, jira what has happend and format it to our style

[@elitebrothers4256]

Hey tim, i've already posted this comment before in the outerwords lauch party video but i REALLY want you to see this (and i have no money to be a member, you know college stuff) here it is: have you watched the outer world's secret level episode? If you have no idea of what im talking about secret level is an animated series of different game ips like d&d, Warhammer, megaman and ofcourse outer worlds, it very fun. Love to hear your opinion. Ps: don't worry i won't be reposting this anymore.

[@CainOnGames]

I’ve seen it, and I was involved in several meetings during its production, so my take on it is biased.

[@elitebrothers4256]

@CainOnGames come on Timothy we love your bias, but i get it im not the one who has to read all the comments saying that's just your opinion after you just say that it's your opinion. Thanks for reading and responding your an inspiration Tim.

[@DiluviumEyesofThunder]

I'm just naively tinkering, but in my systems I data drive everything I can, data Items are stored in groups, and each group has configurable subclasses for relevant implementation which obviously default to generic behavior. Trying to take the "it just works approach", so that future me can narrowly focus on UX/design space.

[@SPTX.]

The way I see it, if you make it easy for the user to work with the engine, you also make it easy for the devs. I understand it doesn't make much of a difference when you're used to your tools anyway, but why go through the trouble of obfuscating your assets? It also makes it faster to form new hires.

[@thedude7319]

Oke somewhat deeper in, can I assume that if it has to check everytime it isn't linearity but exponentially equal or roughly equal to the degree of complexity (how much code and requests ?)

[@luisorbegoso7248]

Will you make a video about the "The Outer Worlds" short in the Secret Level series from Amazon?

[@Lemurion287]

The real question is not just the cost, but the value. Is implementing modding going to be worth the investment of time and money to do it. For Skyrim, the answer turned out to be definitely yes. For other games, it may not be. Personally, I have to admit that it's a huge driver of whether I buy a game or not. I mostly play single-player RPGs, and I can get a lot more play time out of a moddable game than one that isn't moddable. That makes moddable games more valuable to me, and so I'm more likely to buy them.

[@fearingalma1550]

With Unreal being such a popular engine choice now, couldn't devs just dump the project file to the community, assuming you don't need to rip out any other licensed middleware? I guess the main problem then is someone could just repackage and list the game again for a higher price, particularly if you're an indie developer without an army of lawyers to sue people that reuse your stuff without attribution. Sadly, the golden age of modding is over, but now everyone can just go make stuff with an OTS engine and stock assets rather than learning a game-specific engine with deliberately designed and modifiable parameters.

[@reecesx]

Number one rule of moddability: have "beta testers," "QA," etc leak your PDBs. Done.

[@hotflakestom]

This is a wonderful video. Not only tells us the thought and process that needs to go into providing modding but also the reasoning. Makes sense why many devs who do put in modding tools do it after launch. And so much more respect to developers who do.

[@Me__Myself__and__I]

No. You totally got the wrong message from this video! Adding modding later and espicially after launch is literally the worst idea. No one should do that, literally ever. Imagine building a car from scratch and then deciding to redesign and change out the entire transmission system, steering system and engine AFTER the vehicle is completed and working. Its a massive waste of time and effort. Such things need to be built-in from the beginning. Tim specifically mentions that adding modding support from the beginning, bofore a single line of code is written is the best time. Lets say modding support could be added to a game as the first step taking between 80 and 160 hours. To add that same degree of modding capability after the game was completed would likely take THOUSANDS of hours and still end up being worse.

[@BlackJar72]

I'd really like any games I make modable, as I started out modding. I doubt I'll actually do it any time soon -- it makes things so much harder and more complex, and trying to make games is already a whole lot on one guy whose still far from being really good at it.

[@thedude7319]

Oke, now I am curious. It would seem it shouldn't be too difficult. Since you need to have openings in which you can attach something if I take my engineering background into account with some restrains ofcourse . But how does software handle that openess

[@Kiyuja]

Making a game moddable seems like such a daunting task to me. Dont get me wrong, if I could, I would always make my game projects expendable by fans, I love the idea. But I have no idea how to efficiently expose the entity data while also making good code.

[@mikeystir3333]

is there a reason that the outer worlds didnt have a 3rd person toggle? its the one thing that hurts my immersion, i really wish i could quickly view my guy from the 3rd person while in town and switch to first person when i need to be able aim better or search a room. i was really dissapointed to see this feature withheld from the outer worlds and cyber punk it was like a small seeming thing that held these two games back ever so slightly from being as great as the 3d fallout and elder scrolls games. happy to see avowed will have this feature.

[@CainOnGames]

There were no third person animations for the player. You have to plan for and make animations for walking, jumping, climbing, using various weapons, and interacting with objects like doors, computers, and containers. Our budget was fixed and relatively small (compared to AAA games).

[@TheMichaellathrop]

I take it that constantly needing to run those checks is a good bit of why the Bethesda engine is both so slow and so likely to crash.

[@blahblahblahvblah]

ARCANUM: MENTIONED!

[@GarrethandPipa]

man Tim I don't know. While I haven't ever made a game moddable. I modded a electronic medical system to allow vendors to snap in different insurances it easily double the development time. I bet I had to touch 80% of the codebase after the fact. It is one of the most painful things I have programmed in 47 years and I spent years writing postscript drivers back when printers was the wild west and usb wasn't even a wet dream.

[@WastedPot]

It’s not a matter of ‘Oh if they just do this…’ It’s a matter of Skyrim still having 29,000 active users 13 years later, and that’s only the PC audience. You forego mod support at your own peril.

[@arcan762]

It is similar explaining why we can't "just add multiplayer" 😅 _This single-player RPG is nice... but what if we turn it into an MMO?_

[@dempa3]

Development time and budget set aside, are there other,say more commercial reasons why one wouldn't want a game to be moddable? Otherwise it seems like something that could keep a game alive and interesting for a long time .

[@CainOnGames]

As others have mentioned, both licensing agreements and converting the game into something harmful are both reasons to forgo moddability.

[@dempa3]

@@CainOnGames Many thanks for your fast and informative reply!