The problem with higher star counts for harder demons is that the difficulty ratings of levels changes frequently. Players could beat a newly rated Easy Demon while it’s still rated Hard Demon for free stars and there would be problems for people’s star counts who have beaten a demon before its difficulty promotion/demotion.
They could just make it work like the non-demon difficulties, and levels rated as demons right before the update would just need to have their difficulty verified or changed.
3:46 I think Robtop said something about this, so if he was to make it so you get more stars for harder levels it would be more difficult since players choose the type of demon. And time to time demons change in difficulty to extreme to insane or other ones so it would be unfair for players we haven't beaten and to those who have before it increased or decreased in stars
i think it shouldnt be difficulty specific, maybe it could just be like for example: ICDX is 15 stars extreme demon but it changed to insane demon but still wont change (note: ICDX probably did not change back to extreme in real life)
7:54 Instead of the achievements tab containing buttons that contain each type of achievement, (stars, coins, etc) it could just show the easiest achievement of that type. For example, one of the achievements could show "Collect 5 Secret Coins" and once that achievement is completed, it would be replaced with "Collect 10 Secret Coins". This process could be repeated until all of the achievements of that type have been completed, checking off the achievement with a "✓".
I LOVE the one about "certain user level" achievements! There are so many user levels that are so amazingly influential to the community that, if they never existed, the community would already be dead. Examples: Retray Bloodbath Cataclysm 9 circles Demon Park Bloodlust etc.
The last one (sfx) can be very easily replicated using a touch trigger , touch triggered spawn trigger, and trigger orbs, and sfx will be in 2.2. i might add it to my levels
loved the first idea 1:18 how many things would make the community more engaged in the game I also loved the idea of having a challenge of 100 demons, I just think it's a bit too much, maybe 50 or 25 so they don't end up with all the demons in the game but obviously I want these things to be added to the game only in 2.3 or 2.4, we are already waiting too long
Idea: Geometry dash Items, basically something you can add into your level, that the icon can pick up and use. Like a sword, Or a pickaxe. (this would be possible with triggers, but making something like this as a default thing would make it easier)
Ideas for Geometry Dash 2.2: Temporary Offers: In these limited time offers there are icons for a number of jewels or orbs. Custom Achievements: The New Geometry Dash Achievements are: Demon Route: Get the 1st Time Machine Coin and the last Stereo Madness Coin. The prize will be a ball with the viewer of the characters of among us Ambitious Collection: Get all the coins of each level. The prize will be a Battery Robot. Icon Builder: There is going to be a new robot, the Builder. You will need the BUILDING PIECES. Building parts can build store icons WITHOUT BUYING THEM DIRECTLY! Building Pieces: At least 5 pieces per level can be found in All Levels. Normal cubes require 3 pieces, Rare cubes require 5 pieces, and Epic cubes require 8 pieces.
Man u r hella genius, the crazy part is I was also thinking of ideas and I can't believe that u already said the tow I was thinking, 1, stars gets bigger when going further in demons and 2, making preset's for icons Awesome video man
Guys i have an idea that in my opinion would be amazing here is the idea: (Its Large so if you are lazy dont read this) The idea is to create an artificial intelligence (AI) that automatically adjusts to clicks or taps in the Geometry Dash Level Editor, However There are several aspects to consider before starting the project Data Capture: You will need a way to capture the input data, that is, the clicks or taps that the user makes in the Level Editor. This can be through some type of event logging or by intercepting user input. Machine learning (ML) model: To train an AI that adjusts to user clicks, you will need a machine learning model. You can consider approaches based on neural networks, supervised or reinforcement learning, depending on the complexity of the behavior you want to achieve. Training data set: You will need a training data set that contains examples of user clicks and how the level structure should adjust in response to those clicks. This data set should be representative of different level scenarios and styles. User Interface: The AI will need to interact in some way with the Level Editor to make adjustments. This may require integrating AI into the game or creating a separate user interface. Evaluation and tuning: After training the AI, you will need an evaluation process to ensure that it is producing desirable results. It may also be necessary to continually adjust and improve the model based on user feedback. This project will require skills in programming, machine learning, and user interface design. As you progress, it's crucial to collect user feedback data to fine-tune and improve the system over time. Here is how it would work Initialization Phase: In the first steps, you use an initial set of data collected during interaction with the Geometry Dash Level Editor. Train your machine learning model on this data set to get a starting point. Implement Initial Feedback: Integrate a user interface that allows users to provide feedback on the quality of the generated levels. This could include options like “Like” or “Dislike,” specific comments, or even rating scales. Continuous Learning in Real Time: Implement a system that adjusts the model in real time based on user feedback. This will allow for continuous improvements as players interact with the system. Constant Data Collection: Design the system to continue collecting user interaction data and feedback during normal use. The more data you have, the more robust and adaptive the system can become. Continuous Improvement Phase: Periodically, train the model with accumulated data to improve its performance. This can be done from time to time to incorporate changing user preferences. Experimentation and Testing: Perform experiments and tests with different model configurations. You can try different hyperparameters, network architectures, and training approaches to optimize the system's ability to understand and adapt to player preferences. Performance Monitoring: Establish performance metrics that reflect the quality of generated levels and user satisfaction. Monitor these metrics to evaluate system performance and make additional improvements. Basically it will get better over the time but The AI would need to be tested so here are some ideas that would work for a Good Test For the AI: Rewards based on user interaction: Design a reward system based on the user's interaction with the generated levels. Rewards could reflect perceived fun, appropriate difficulty, or overall quality of the level. Assessment of Level Diversity: Consider metrics that measure the diversity of the levels generated. You'll want to prevent the system from generating levels that are too similar to each other to provide variety for players. Efficiency in Level Creation: Evaluates the system's efficiency in creating levels compared to manual creation. You can measure the time the system saves users and the resulting quality of the levels. User Feedback: Incorporates mechanisms to collect direct feedback from users. Survey questions, open-ended feedback, and rating scales can provide valuable insights into the user experience. User Acceptance Rate: Measures the acceptance rate of the system by users. This may be based on the number of times players use AI-generated levels compared to manually created levels. Adaptability to Changing Preferences: Evaluates the system's ability to adapt to changing player preferences over time. You can measure how the system adjusts its recommendations in response to new preferences. User Retention Metrics: Examine user retention metrics to see if introducing AI into the level creation process has a positive impact on player retention and engagement. I think this could be for 2.3 I dont know if its a good idea but in my opinion i think it is (I just asked ChatGPT for all of this, and pasted it and this idea just got in my mind and i just thought about it like a good idea)
6:49 I just want to mention, robtop doesnt need to make more levels, this game is entirely community based, Robtop provides the tools and we use them. Even if he created 100 more levels, the playerbase wont change that much.
3:20 the problem with giving more stars for top demons is that once these levels start going down in the list. Would they stay the same star rating or will each new harder top 1 get more stars than the other. Idk it seems like a good idea if robtop wanted to implement it but it probably needs tweaking
3:39 How about Easy demon = 10 stars Medium demon = 11 stars Hard demon = 12 stars Insane demon = 13 stars Extreme demon = 14 stars For the top 15 hardest levels, how about 15th = 15 stars 14th = 16 stars 13th = 17 stars You get the idea.
I know this has been out for a few weeks already, but I'm so glad you're back! I was kinda bumbed to see you gone for so long. Although, I get it. You had other things to do. :3
2:46 I think he could leave it as it is and then only show the likes and dislikes separately when you click on the icon (or something) because we should still be able to see them added up. by clicking the icon, I mean even just a simple “👍 2735 : 497 👎” would work (for me lol)
2:33 yeah i thought about that too bc you can just flex that you are good and have 1000 demons, that are only ez demons. But actually a guy with only 20 demons, but 10 of them are extreme is better then the guy with 1000 ez demons.
i think a possible reason why you cant see other people in versus mode is because it might distract you from your goal, which is to beat the level first. things like accidentally mistaken others icons for your own, and would just be a confusing experience. but tbh that should at least be an option.
3:00 That would be a problem, cuz the demon difficulty is decided by the community so they could just rate every demon an extreme to get the most rewards. Or trolls could do the opposite and make top 10’s rated easy demon
5:29 and also, add flags! English 🇧🇼 Spanish 🇬🇶 Russian 🇲🇩 Korean 🇰🇵 And more languages French 🇳🇪 Portuguese 🇲🇿 German 🇧🇪 Dutch 🇸🇷 Hungarian 🇸🇮 Italian 🇸🇲
The one with more stars for demons, would be kind of bad. People could just rate easy demons extreme to get max rewards. If RobTop could put his own custom star rather than always 10, then it would be better.
8:44 I think I got an idea for this. In Mario Wonder, the online mode in the game has other players around the world in an 50-20% opacity, so if they did the same with GD, it would be perfect.
Ok, I have an idea for the thing you said "The nightmare & Bloodbath is both 10 stars." Idea: demon - 10 stars easy demon - 11 stars medium demon - 12 or 13 stars hard demon - 14 or 15 stars insane demon - 16 or 17 stars extreme - (according to the pointercrate) top 1 - 30 stars | top 3 - 25 to 28 stars | top 5 - 19 - 24 stars.
1:49 this is more like making geometry dash the worst game ever fr now like XP? reaching new levels? like thats a non-epic feature fr fr edit: flw also forgot that robtop can fool everyone cuz he himself doesnt know when he will release it edit 2: if you want to make ideas for 2.2 it will make the release date be even longer, not even in 2023, in 2024
0:23 Absolutely. It would give new players a direct access to a far greater community of experienced players and creators, thus accelerating progress and connecting more people worldwide. The issue with this however is that the player base is largely composed of young individuals, and moderating the so-called 'Global chat' may become impossible. There is also a problem with what I call 'exposure': newcomers may experience insecurity or dread because of sky-high standards for rated levels and be discouraged from participating publicly in the community. Sure, it may filter out players that are not serious about this game, but it should be a platform for everyone, so that everyone has their thing to show and people to talk to. 1:50 The idea sounds great on the surface, but that would require the entire achievement system to be reworked. And knowing how game statistics for each account are stored in your save file, people could easily modify them, rendering the whole system useless as it wouldn't necessarily indicate player's skill. 2:27 It's a neat little addition, however trying to communicate that you can click the demon icon might be a little difficult. It should also be an account option, so you can choose to show it to everyone, to your friends or just to you. 2:46 Separating like rates from dislike ones was suggested by a lot of people, and I too am a part of them. The trouble is, this statistic was never recorded separately, because like rating is calculated by taking away dislike count from like count. 25 likes and 3 dislikes will translate to a total rating of 22 likes. Pre-2.2 levels will never get this rating accurately. 2:53 It's totally fine, really. You can't tell if people have beaten the level legitimately by looking at the leaderboards, too. We just see a player who got 100% and some coins. We will never know how many attempts it took them, and how much time they spent on this level. 4:00 Modifications for a collaboration system are already being used by players, so it's not impossible to see this coming in future updates. Having the game's servers handle it would be an issue, as server blackouts could lead to loss of building progress. 5:36 The hardest part of translating the game text would be to implement fixes for graphical interfaces that weren't supposed to handle different glyphs and paddings. So, for example we have a text string in English that has a certain width. By translating it accurately to another language the width of the string can vary a lot, causing some text to appear in places they weren't supposed to be or go off-screen entirely. Preserving context in it's original form in another language is fine, but doing the same by preserving the same width can be impossible. This whole crazy operation would also quadruple the amount of gamesheets and separate textures, doubling the file size of the game and lowering performance of some devices. 8:35 YUCK. Absolutely not. I grew up playing this game because it was the first time I've seen the whole world united by passion and creativity. There was no word of nationality and culture of the outside world. Everyone talks one universal language and connects to anyone. By implementing this system, it will obliterate the whole feeling and bring in political drama and personal hate. What used to unite us should not try to separate us. At the very least, if this abhorrence ends up being considered, make it an optional feature people can turn on and off. Or make it so each account can choose whether or not they want to display their country of origin. 9:55 This should be a small submenu under credits that can tell you the game's version, and other meta information like the amount of account log-ins.
5:58 is just straight up insanity because let's be honest here, NO ONE IS GOING TO BEAT ALL THE TOP DEMONS JUST TO GET AN ENDING, BECAUSE BAD PLAYERS WILL NOT LIKE IT EITHER.
I have a suggestion that could be used for making the main level menu more cool. So, with the new online level loading screen, (y’know the one where the play button on the online level has a green loading circle around it) But with the main levels, you could have a green line border the rectangle with the name on it when its loading. Or, you could make a green loading line surround the difficulty marker. I just thought of this so let me know if it sounds good or, if you didnt understand what I meant, feel free to reply with what didnt make sense.
