Tierlisting the BEST FEATURES for your game 

my next game will be a game where you stand on a train that is driving around a lake, and you are fishing this lake, while standing on the train

It's important to keep in mind that things like rebindable controls and gamepad access is about accessability not just like a comfort thing.

I know trains were a joke. But "Gimmicks" would have made a good category. Gimmicks like adding train can litterally take your game from a generic indie horror to a cult classic, as with the game you mentioned. Adding the grappling hook to just cause 2 took it from a generic action shooter, to a beloved classic sandbox series. It's an easy way to get your game to stand out and can end up being the games entire identity, as with the portal series being an extreme example.

Fun fact: Former Blizzard director on Vanilla Wow Mark Kern has said he puts fishing in literally every single game he makes- just because his wife loves to fish in games

Can confirm, trains are an S tier feature

Spoiler alert: All of it is scope creep.

About testing networking with multiple instances, I got news for you guys. This is actually done at big studios. I worked at Ubisoft and at Larian and we totally debugged multiplayer issues with two instances of the game running. One of the game instances always ran at a poopish fps but it was not used for debugging latency, just for debugging other stuff that could happen in MP and, believe me, a lot can happen that doesn't happen in single player mode. For instance, on Divinity Original Sin 2, we had VFX issues that were exclusive to MP mode due to sync issues. And, Might & Magic : Duel of Champions at Ubi just couldn't be debugged without running two instances as it was a free-to-play trading

Great list as always. If you ever do a sequel or remake of this then maybe some of these things could be added: - minimap (or actually a map overall) - color blind options - easter eggs/secrets (altho maybe this is kinda too similar with achievements as it is something extra players can do if they want to) - DLC/expansions - (animated or video) cutscenes - voice acted dialogue - fast travel - customization (main character, cursor, pet, base/house/camp, font, text box color & shapes... etc) - autosave/checkpoints (i feel like manual save is maybe too obvious for a tier list like this but autosaves and/or checkpoints shouldn't be overlooked)

As a left-handed, rebinding can be a huge dealbreaker for some games, especially shooters where you need to access quickly a lot of keys. We don’t have access to the same keys because our thumb is on the left side of the keyboard

Unpopular opinion, accessibility stuff like localisation, controllers, rebind, colour blind, subtitles, tutorials, multi platform, etc are all S tier. For nothing other than, it forces you to have rock solid design. Even if the player doesn't use them, it makes the game better.

Best he mentions trains again in this video, as they turn up in all the other scope creep vids.

7:09 Redindable controls are very important for the countries that don't have the same keyboard layout. Bc otherwise you need to add a new language to your keyboard and switch to it just to play a game. The casual player will probably refound rather than searching how to install another layout

Listening about Trains while implementing a fishing system in my game haha, you guys are awesome. You have been consistently providing so much value to the game dev community. Wish you all the best! 💛

Hi I am gamer from Denmark. You talked about location in Danish so I would like to give you my perspective. Unless you are making games for kids in Denmark then I would say that you should not bother with location. Most gamers here prefer to play games in English since past location in Danish has been so bad that most of us dont even bother to change from English to Danish.

I hope you keep your word, because now I'm going to add trains to my... chess puzzle game. Gonna be back for that free marketing in ~four months from now.

CONTEXT: - should be noted that the Tier-Ranking is affected by the game engine used, as per difficulty of completing the feature. - Godot, for instance, makes it VERY EASY to create key rebindings and controller support. And, by virtue of trying out what works with the controller + key rebinding, it opens up your mind on what is and should be possible in the envisioned game. By contrast, online multiplayer is likely much harder on Godot, than on Unity/Unreal. - For this reason: some features are S+ tier if easy to plug-and-play, but C-tier if their development is hard as a result of the framework used.

For real though, there's a takeaway about the trains thing. Trains are very evocative, much in the same way fishing is. A train evokes freedom, exploration, new horizons, being transported to the unknown. Fishing evokes peace, contact with nature, the sound of birds and streams. Other examples of symbolic visuals are sunsets, god-rays, floating dust particles, seagulls, smoke/steam etc. It's valuable to understand your game's scenery and ambience and how such imagery can add to it. A-tier for sure. Related to this, is a thorough soundscape, but this is S-tier. Take the time to appreciate how your game world sounds, not just if the music is banging or the punch sound is punchy, but how does the world sound with music off and nothing happening? Does it evoke the right feelings for that scene? S-tier without a doubt

01) Translations Useful. Look at what your genres audience speaks and what they accept on other games. 02) Controller Depends on the genre, though the Steamdeck changed that a bit. Some genres are now considered controller-games like any sidescroller and even the average third person actiongame that does not include shooting guns. I would look at what the average person uses to play for the genre. 03) Rebinds Unitys new Input has "key position"-bindings, so my WADS are where there should be on azerty keyboards. However, for keyboard control's its a topic with a lot of emotional weight since people are used to their way of controlling games. For Gamepad, it's probably fine to go without rebinds for a lot of games. Toogle VS hold is also both a matter of taste and accessibility. 04) Graphic settings people like vsync The term "some people like vsync" is strange. On some systems, games without vsync look like shit with all the flicker and screen tearing. So a game without vsync is not an option. vsync on by default is probably fine for some genres, for others like action and especially shooters, the audience will be very very very mad. Also, don't forget to test the timing both with and without it and at high and artifically low FPS. 05) NG+ The completion rate of games is between 30% and 60%. Are player's who finished it once and wish to play it a second time the biggest issue you could solve with developement time? There's a reason this is often delivered by patch. 06) AntiCheat It's not as easy as just anti-cheat. For multiplayer games, especially player vs player and with matchmaking, it's not done with an anticheat solution from the asset store. Theres a reason why the big multiplayergames use both server logic and kernel level anticheat. Cheating in games is pandoras box. Don't bother for singleplayer, and if your players form the party by inviting their steam friends they can filter who they play with. 07) Achievements It's a nice way to add progression and goals to a game, and it's a few function calls. 08) UnitTests I'm not a fan of "Unittests take so much time thats why i dont do them". For some functions, writing unit tests is less work. My save system has unit tests, simple because with those i can test a lot of aspects of the system without having to start the game 4364 times and recreate the test situations over and over and over again. 09) Workshop Requires a healthy game to begin with. If you have a game that benefits from that, you know it. If you have to ask the question, you don't. 10) Demo YES-Tier. We can't win by paid advertising or hype. We need to convince. But holy shit, do not release crappy versions as demos. There's a difference between a playtest and a demo. My experience with NextFest 2024 has been kinda sad as i ignorelisted half of the games i played. 11) Multiplayer Multiplayer is S++ or F--. The game is either a multiplayer game, in that case it is not a discussion, or it is not. In that case, don't waste time. A lot of good story games of the last 25 years had their story cut short because corporate demanded multiplayer. Remember all the fun rounds of deathmatch we had in SpecOps:The Line? No? Me neither. 12) Branching Narrative I don't see that as a feature that can be added or removed after a discussion. It's either part of the design or not. There's not "we make an RPG, do we add branching narrative?". That's a different game. Important: Branching narrative is not the same as player influence. You can give the impression of a world that reacts to the players actions with very little work. The Deus Ex 1 toilet dialog comes to mind. Famously the tests of Dishonored generated the feedback that the game was "too linear" because players did not notice they had options, so they played differently and all thought that was the only way. 13) TRAINS Choo-Choo Charles sold. S++. 14) Difficulty Settings Depends on the genre. The difference in ability between players is massive.

Translations are good if you can afford it, because of the gaming market in Asia if you need to pick, its worth translating them, you open your game to a huge market

At first I found it funny how you added "TRAINS" category, but then I realized that in the game I'm currently working on, I plan to add a level that has a train (sort of) in it because I really wanted to have a train in it and yeah, I do like when a game does have a train of any kind 😅

For me personally, rebind key is S tier, nearly must have in every game. I'm familiar with some keybind setup so I always want to check key rebinding to change around. Even if after that I use game default key binding, that's my choice, and it makes me feel comfortable when playing. For example, in 2D platformer, thanks to Megaman X4, I'm familiar with setting Z:Dash, X:Jump, C:Attack. So in other games, like Hollow Knight, I need to rebind to that setting in order to perform the best I can. In FPS/TPS, except WASD to move and R to reload, other keys Q,E,Tab,1,2,3,... I usually switch them to see what fit me the most. Even in some strategy game, the key binding might be really annoyed, especially when they dont have option to rebind.

Hi. I just watched your video, and now going to build my own full-feature dnd game with trains.

your video unironically gave me an idea on how to add trains as the level-change mechanic for my first game. Thanks guys!!

A thing to add about controller support, it's not just controller. It's basically accessibility support. Many disabled gamers out there with their own controller setups which is what allows them to game at all

That proud smile after the game's journalist difficulty mode joke 😂 p.s. I agree that trains are S tier.

I was actually thinking of a branching story based on relationship meters. Picking something good or bad will either add or subtract a float. During dialouge if a float is greater than some number you get that narrative. As long as you make the characters not entireoy rely on other dialogue or make it generic enough that it works for either you can whip out a whole new story quickly.. This works only for certain games though where every character isnt fully connected.

As a Dane, I appreciate the mention of our language, but I must admit, most of us do understand English as well, so definitely agree on that first point. Localization is less of a necessity.

one of the interesting things I noticed in translation is that french buttons aren't capitalized while english is for example "Save As" in english will be "Enregistrer sous" in french, where only the first letter is capitalized.

IMHO for faster passed games rebinds are essential. I've refunded games for lacking rebinds before and there is a reason why i don't buy bethesda anymore. The good news is, you can't get a shittier control system than bethesda. Also, there are people who don't have all of their fingers or may have a hand missing. Disabled people also have a higher chance of being gamers.

for testing network latency there's a bunch of tools, one of the easiest is WANem (though you can also relatively easily do similar stuff with firewalls), setup a small linux firewall machine between the 2 computers, and link a computer on each port. you can use a cheap sbc that has 2 ethernet ports (or wifi), that is affordable and it's still useful for a bunch of other stuff like as a small backup server and such. if you are really tight on budget you can use an old laptop from a recycling center or family and just connect one pc to the ethernet and the other to the wifi. recommendations would be: orange pi zero for wifi+ethernet (~15 bucks) or NanoPi R2S (~40 bucks) if you want dual ethernet you could also use an old pc + an usb to ethernet adapter. if you have an openwrt if you are on a 0$ budget you can always use a single machine and use virtual machines, but that can be more complicated to setup

Mod support really does sustain games but it needs an initial audience first.

Okay trains are on the implement list . My game is a twin stick shooter. How do I do this @marnix ?

Thanks guys, your videos are incredible. I'm a solo app developer in part time from about a year and have been learning a lot from you. I'm working with Android apps instead of games, but many things you discuss still apply. About the localization, I already consider more or less the same problems you discussed. I agree that google translate is usually a bad translation because it losts context. But did you guys tried GPT4 mass translation? Something like, you prompt GPT4 with something like: I have strings for a game in english, the game is like that, the mood is like that, etc. Here are the strings: .... Give me the strings translated in Spanish. I did it and GPT is able to keep format, JSON formats etc, keep the SCV tye justing adding a column, etc. GPT translator is much superior than google translator ones and it can keep the programming sintax. You can feed it at least 100 words at time easily, probably something like 1000. Adding more at once it is easier for GPT to understand context. In a different direction, can someone recommend to me a channel focused in solo dev/small team dev with the same approach as ''BiteMe Games"?

I agree, trains are peak content in all forms

Too bad accessibility is never mentioned for controller support or remapping keybinds. There are tons of people who can't enjoy games like we do because of a disability. If you want as many people as possible enjoying your game, I think controller support and remapping keybinds are probably a must have for most, if not all commercial games.

Trains are S-tier. It is known. It doesn't have to be a ~train~ even. It can be any sort of logistics system, but trains do it best.

Thank you! Great list! I will add the following(where possible): - the game to be DRM-free(increases the preservation nature of your game) - local co-op - modding API(via Lua or some toolkit you provide)

Difficulty settings is relative to your own game, and most importantly how you intend to "tweak" that difficulty. There is the lazy way and the dedicated way. The easy way is just tweaking numbers, which is the less costly way to do and allow players to enjoy your game even if they are not good at it, while the dedicated way is like adding more enemies, new enemy variants, new endings, more level branches or level variations the higher the difficulty is, which requires rebalancing a whole new game for each settings you want, but add a lot to the replayability of the game, and players will be eager to replay in harder mode to discover new things. In term of value, the lazy way is a A tier to me. Easy to implement and allow every players to enjoy your game whatever their gaming skill are, so it should be implemented everytime. Maybe not right during early access, but at least for your game release. The dedicated way is between C tier to me. It can meaningfully extend your game lifetime and enjoyability and raise your game to the "memorable" kind by the sheer quantity of experience you offer to the player, but it will require much more development which you make you go over budget easily. But that can push your game from "game that try to please everyone" to "game that offer an experience to everyone".

I completely agree with trains. I've been developing a sci fi farm sim and one of my original ideas for it was that there has to be a train

Revisiting this one again as it's gold. You guys dropping Dutch made me chuckle - I get tired writing my street name: 19 letters alloneword... In Japanese it'd be four characters. In English the same length but at least with some spacing to allow linebreaks.

I've developed a solo game with a branching narrative. I wouldn't say it's impossible, but it does require heavy tradeoffs and a lot of work(it took me about 1.5 years working with an established engine.) It's definitely a lot easier in a text based format, though. I doubt I could have done it with full animation and voice acting.

Regarding modding and Minecraft: Minecraft is pretty old now so maybe there are a few people that just don't know but Minecraft mods used to be made by decompiling and deobfuscating the game and adding your own code (in Eclipse... Wild times). There were mod loaders at some point but it all started incredibly primitively. None of that modpack, and modloader stuff you have now. So, if you want modding and you have that option, add modding that you use to make your game. Don't give users an alternative API. I don't know much about engines so I don't know how well that integrates with Unity or Godot but, like, if you want mods in Lua, make content for your game in Lua. If you use C# (like, Monogame or straight up OpenGL bindings. I don't think that works with Unity), make your content as DLLs you load at runtime. Then you can offer your users the same interface you're developing against. Still probably not a good idea in terms of time investment but at least you don't have to maintain a second API that you also need to keep in good shape.

I disagree on tutorials. They are a MUST. Not every tutorial has to be a defined chunk in the game with blocks of text explaining the game mechanics, but every game needs to teach the player how to play it somehow. Even just telling the player which button to press when it's first relevant and clearly demonstrating every core game mechanic at some point. It's bare minimum but I would still call it a tutorial because it is an extra thing the developer needs to keep in mind. I would even argue that tutorials are so essential that even a bad tutorial is better than no tutorial. So that one would be an S tier for me for SURE.

Control remapping is a must-have, IMHO. Keyboard has like 100 keys to choose from. Bonus points if you let people separate, say, use from get in. Or crouch from skip dialogue. Also some keyboards don't have some keys or their layout has some keys in different places. And qwerty azerty as you mentioned. Only if your game is literally menu based or uses like 5 buttons you can say "It's definitely not worth the effort" But at the very least - give people two alternative mappings. Like arrows, z, x or wasd, j, k, for example. Please don't make tutorials part of the game. And perhaps don't make them intrusive. Meaning, tutorials that lead you by hand. All those "this outlined bouncing highlighted shinny button is this, press that bright green wiggling button to do that" evaporate from my brain as soon as I've pressed them. I would very much prefer readable UI. Literally. "Inventory" and" journal" instead of random 64x64 image of something that's either create or a book. Even "obvious" things might be misread. Like, "there's an icon of blood drop. Am I bleeding? Why my health doesn't decrease then? Oh, that's some kind of counter for some mechanic that you explained me a week ago in a whole paragraph of text that i can't read again!" PS Thanks for the idea. My game will definitely feature two modes - game journalist difficulty and normal difficulty. 😂😂

6:15 I honestly believe rebinding controlers is S tier, I love dark souls games, elden ring sekiro whatever fromsoftware has, and I do believe I would've liked ds 2 as well, but when I saw the rebinding was limited to a set of their bindings I was instantly like, nope, I ain't playin this

I'm trying to figure out how I can add a train into our demolition derby game. . . Maybe an arena obstacle perhaps. Or heck even a quirky car design. :^)

A big branching narrative game usually doesn't need other features to succeed

For Indie Action Combat/Adventure Games... Custom Game Controls & Controller Support Color Blind Mode (for Character Customizations such as Color) Music/Sound Options & Visual Effects Options Simple Mode and Complex Mode for Combat Styles. Cross Saves and Cross Play (console dependant) English Language and English Sub first (for me, in America). Bots for PvP stuff and Offline & Online Support. Dedicated Servers. I think this is all the technical side that most players will interact with in "Settings" among other things. Now... As far as actual game mechanics/content goes? That's up in the air as the Hat of Infinity is endless. I mainly do Conceptual Design Adventure Games across all platforms as a hobby. So my list is what I consider as "Basic Necessities" when designing games. It is the foundation to which everything else is built on.

30:40 It may be a better idea instead of adding difficulty settings, add some accessibility settings. Like immortality, time dilation etc. Especially if you game is story based but have some insane challenges.

Watching this again, there's one "feature" I missed: Randomly and procedurally generated worlds / scenes / levels /etc. Maybe do a retake in a couple of months? 😉

graphical things that should have settings: motion blur screen shake flashing effects foreground/background fps limit (especially if your game benefits from having it running in the background, having a way to limit it to 5-10 fps is really useful) fov (for some games)

Localization is S tier Unless you’re a native English speaker

At our studio we almost got dead threats for not implementing an option to invert the X & Y axis in the mouse. I didn't even know people do that with first person controllers XD

Some of these might be better represented on a scale of "effort vs impact" instead of a tier list. Like "difficulty" for example: low effort, medium-high impact. That can help small devs pick and choose what challenges to take on.

Rebinding your keys should be a must/required in any game going to be sold. Accessibility features can make or break many people's ability to play your game. I'd highly recommend reading the game accessibility guidelines for any developer and their basic list covers easy-to-implement and wide-reaching things to add.

To be honest, the trains in Dorfromantik make me so happy tat they easily bring up my enjoyment of the game by 100%

Yay more tier lists, love these videos!

This was good advice, mixed with a great 2-man comedy act heh. For part 2, I'd add: + Character/skin customization [genre-dependent, A/B for RPGs, D/E for anything else beyond palette-swaps] Exception being avatar-gender - have grown up always having to play a boy in games, it's f*cking awesome now you can pick. + Furries 😅 [I share Marnix' love of trains, but they're a little hard to work into medieval games heh. But don't underestimate that furry market, either via skins, creature aesthetics, or pets.] + Sex appeal! [Gameplay-wise, this can be datesim elements, but also relationship/affinity metrics, or certain quests. Even if you aren't making a waifu-centric booby-game, having character designs that at least consider a sexual gaze (male & female, straight & not) is likely to have a positive impact, on your marketing & your fan-loyalty. Ppl naturally get more attached to characters, esp. good-looking ones.]

I can’t find a place for trains in my game as it takes place like.. 3000BCE. Would Wagon convoys do? Fishing is going to be a thing for sure

Where would you guys place editing a "text" file for non-critical but very nice to have graphics settings and remapping controls

planning for these features will make it way easier to implement if you want and usually make your game easier to develop in general for controller, in a menu heavy make your menus support tab navigation (i.e. use tab to switch between buttons and space to press the button, if you want to support controller you just have to take your list of buttons and find the closest in that direction), and outside of menus, have wasd movement. though some games like rts may be harder to support controller on, but look at similar games with controller support to see how it can be done. for localization, put your text in variables that are loaded from a resource file, not only it makes localization way easier, you don't have to recompile your whole game just to change a typo or a bit of dialogue same goes for keybinds, have the keys be loaded from a config file, if you want to add a configuration menu later you just have to do that, otherwise you can just have players who want to change keybinds modify the config file

The German translation looks like someone just ran it through Google translate. For example "Scales" where translated as "Wagen" (balance scales) instead of "Schuppen" (animal scales)

azerty/querty option is absolutely necessary if you don't have rebindable controls tho

Solid discussion especially the part about multiplayer. Server based multiplayer was an obvious no but it does make me question even at what point self hosted peer to peer multiplayer becomes worth it in titles similar to Stardew Valley for instance. It worked for that game (and similar ones in the genre) but is it worth it if your game scope and expected sales are much lower? Other titles in say farming sim genre like Harvest Moon were just fine without it (albeit the classics are rather dated at this point).

About trains... I would say, there are a few things that would make any game a little bit better: * Trains * Pets * Fishing There are SO many people that would be much more interested in your game INSTANTLY ;-)

i translated my first game (Karina Katana) in 10 Languages, but that was possible for me to do in a week because i dont have that much text in thr game

I would say controller support is most critical in genres where analog inputs are standard. In rough order of importance: Flight sim > racing > shooter > RPG > anything top down and macro scale.

In defense of trains, I have an idea for a navel game, and I 100% want to add trains

Localization is different than translation. Localizations could include cultural considerations, like some reference or name or an icon might not mean the same thing to a foreign audience. Localization might also mean adjusting features, for example I worked on an app that used certain hand gestures (mobile device) and we noticed very poor performance in Indonesia so for whatever reason in that region certain swipes and gestures weren’t used so for that region we dynamically added buttons which made the experience more cluttered so we didn’t include those for outside that region.

I just had deja vu watching the train segment lol

Why would you drop Rewired in favor of Unity's input, though? It's easy to setup, convenient to use and has stuff like rebinding out of the box.

16:00 the tutorial is always the best part of a game, even if it is just text in a photo. how else do you know the controls of a game. (in some games i have died because the death key was not in the tutorial and someone was joking around.)

You need to explain this tier list to new users in your videos as I am confused. What happened to using MoSCoW for a list of Must, Should, Could, Won't have? Anyway, I did enjoy the range of topics covered. Great stuff!

Great tier list and discussion. I’m also thinking that difficulty might be an option to new game+ to make the game more replayable. When you play the first time, maybe you start on easy and work your way up. As long as the game is interesting enough to play more than once :)

26:15 if you want to make a story game in that way, it's like a tree, splitting every time there is a new option. to make it simple you can just make less options, but if you make a multiple choice game it's very hard to make it long because if you have 10 things that can be, and you want to ride more, you than need to ride 20 more stories. (i think it is fun to add 1 or 2 options to change the game, so than it would be 2 or 4 stories in one, but i dont even want to try more than 4 choices that can change the game.)

NG+ is something you do after release as free DLC if the game takes off.

You forgot one crucial "feature" - REGIONAL PRICES! :-))

With Unity/Mono, modders can likely do it via modding frameworks on their own. Consequently, modding can help with adding more language translations to some extent. Two hurdles are private datatypes and right-to-left languages though.

difficulty settings should be a I want to play the story/intended difficulty. you could relatively easily do it where it unlocks if you die 10 times in a row on a boss or something also tie it with achievements if you do a hard mode that is balanced but be careful with just scaling hp, if you have mechanics that do %hp they might trivialize the hard mode and stuff may brake with damage reduction (e.g. triggers on doing some amount of damage) also if you just scale hp and the player can stand here barely moving, you have bad enemy design (diablo 4) my main tip if you are making a game with rpg elements, make in-combat life regen hard. (out of combat life regen is fine, though that depends on how persistent you want the damage) a good example of a somewhat bad implementation is path of exile, the ease of access to copious free life regen means that to be scary, a monster has to kill you in one hit, otherwise you can just heal all the damage in less than a second. though on the other hand if you give free trickle life regen out of combat, that incentivizes the player to not play the game while waiting to get back to full life. note that if your intention is to make a polarized damage game where you are supposed to dodge or die, this is fine, but know that you risk making the only viable way to play be crowd control or high damage.

Great vid, Thanks. "If you're a Solo Dev...How do you test multiplayer?"...you should have that on a T shirt! Train people are funny...there is rationality to their obsession, but it doesn't help. Wife's Gran worked on rail in Canada...the whole family is train obsessed.

Must have features, IMHO: Subtitles Key remapping Graphics settings (at least resolution, windowed mode, toggles for motion blur and similar things). Volume sliders and colorblind modes. This includes designing a game such that even in b&w you can distinguish things from background and similar items from each other. For example, life potion is in a heart shaped bottle or has a heart on label, but acid/poison is in triangular bottle and has scull on label. Subtitles. Ability to skip everything that isn't gameplay. Not at once, but like skip cutscenes, voiced sentences, etc with a couple presses of a button or by holding a button. In UI-heavy games - text tips for icons. Like how in most OS's and programs you can hover over something and it gives you a little text box with the name of the thing or a description. Bonus points for moveable and scalable windows in your UI Subtitles

For disabled gamers, rebindable input (and controller support) is extremely important. There are some games I simply cannot play because I can't rebind input. Also, by having rebindable input in mind at the design phase, one makes sure that they think in actions rather than specific keys. The added benefit to that is that if one decides later on to add controller support or any other mode of input (microphone, steering wheel, vibration, air flow or whatever esoteric mode one can come up with), the groundwork is already done. In general, accessibility is easier and cheaper to implement the earlier in development one does it.

Perhaps if a language is too long, we can just make scrollable text box for every texts. But if you don’t add them from the start, it could be a big technical debt to pay.

Most game engines have a native rebinding controllers that are simple and quick to implement, and controller support. Especially Godot and Unreal are dead simple to implement and it is a huge plus for accessibility for users with disabilities.

Controls rebinding depends on the engine too. For example, if we're talking about Godot then it is an absolute no brainer because every control is just an action with a default bind and all your programming is done with action. So you can simply have the player press a button and it receives that key and puts it to that action.

I am glad you put mod support to the place where it belongs. Finally somebody says it, I am telling this people in game forum since years. Too many games were demolished by the community asking for it. Same for multiplayer. This features are a huge development sink and only pay of in certain situations or have to be planned from the beginning on ("you can't just add multiplayer"). Also multiplayer mode in singleplayer games caters to only a very small fraction of the playerbase(1% or even lower in terms of hours spent?) Achievemnts are something I wonder about. I'd have listed them much lower, because I prefer it if games integrate them in the game itself and not as a some kind of metagame. However, steam achievements might be something different if they are easy to implement and are relevant for "steam fame".

Yeah, trains really DO make any game more interesting, never thought about that!

I've seen multiple videos where you are talking about not implementing multiplayer. I agree, but what about and arcade game with a scoreboard that is synced to a server and where you are ranked against other players? Is this also part of the 'multiplayer pitfall' or does this fall under another category? P.S. Hi from Belgium 👋

About the party games. In reality if you have a local multiplayer in steam you can use remote play together for the online multiplayer and it works really well. It's kind of a saviour to indie devs but local party games dont sell that much anyways

Controller support and rebinding are important for disabled people. That alone makes it S tier even if not one fully able person uses it. Accessibility is always S tier.

In my opinion, branching narratives can be a good thing, though I tend to lean on them converging at one point is crucial. Something Like Fire Emblem 7 or Sacred Stones.

Multiplayer is a full subgenre. All the game need to be thought around it. You can not compare it to settings. Trains topic made more sense to me than multiplayer topic.

Even if it's just reading them in from a config file I would say control rebinding is important.

I listen to what you say and think "You guys are crazy. This is way more important than what you're describing." Then you give it a ranking and I think "okay, yeah I kinda agree with that." The graphics option discussion was interesting to me. My mind went straight to things like resolution, texture quality, AA levels, and was thinking these are not very important. Doubly so for indie games that never - or at least almost never - have graphics demanding enough to warrant those settings. I didn't even think about things like motion blur, FOV, and I would add bloom to that list. Graphical settings that when enabled actually make games unplayable in my opinion. What you are describing as NG+ is really just new difficulty settings. NG+ just means you start the game over with all the upgrades you got in the first playthrough, with increased difficulty being optional. Either way I don't see it as worth it, so agree with you there. And on to actual difficulty settings - I think these are generally not implemented well. First off, being able to change the difficulty mid-run is something that should be done more often. Like you said - you don't know how good you will be at the game at the title screen. Second, I think you should really focus on one default "normal" difficulty (side note - don't design around one difficulty and call it "hard" - that's stupid). Make adjustments from that baseline to make the game easier or harder, and don't use difficulty as an excuse to poorly design your difficulty curve and just say "well, the players have options!"

Maybe consider not every country has the same default keyboard, thats why action heavy games can't just assume qwerty or wasd are always used. I also played flash games when I was a kid where num keys werent assigned to the keys but to the characters which makes using czech keyboard just bad, we need to press shift for the numbers.

Good video. Good trains.

1:05 there are no wrong opinions

what about multiplayer but through the steam remote play function ? Like for a fighting game

6:10 that's not my problem 😮 that face shifting to cold chilly stare is scary AF

04:04 het is leuk als spellen ook in het Nederlands beschikbaar zijn, maakt het makkelijker begrijp baar. als het enkel weg is gedaan omdat de teksten te lang waren, waarom zijn de teksten dan niet korter opgeschreven? (sorry if you dont speak dutch.)