EDIT: I'm noticing a lot of people in the comments blaming the testers or just commenting "skill issue" and I think I am partly to blame for this. The testers simply provided honest, valuable feedback and I believe the vast majority of it was useful. They are not stupid and were all selected through an application process. As a software developer, my job is to write software that will provide value to the user and I can't just write off all negative feedback as "skill issue" and "haters hating". The feedback I received from this process was extremely useful and has made the product better. The feedback was integrated into the game without sacrificing depth and the gameplay was not changed significantly. It just communicates to the player more clearly and is more polished. ORIGINAL COMMENT: If I survive the creation of this DLC, I promise I'll make a train DLC. For legal reasons this is not a real promise. Thanks for watching everyone! Without this channel's sponsors, none of these projects would be possible. To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/AngeTheGreat . You’ll also get 20% off an annual premium subscription.
I gotta be honest, I'm 99% here for Engine Sim but seeing how much you care about quality in the DLC gives me a lot of confidence that you'll do Engine Sim right.
@@DUKE_of_RAMBLE -- YES! The transparency is a massive factor in community trust and support. It's one of several things that separate the "WUBE software"s from the "Intercept Games"s.
you make what's pleasing to you and find the niche of players that share the same values you can't please them all, but those who understand your views can help you develop the game into something that you'll enjoy even more in the first place
2:26 as someone who worked on nuclear power plants I kinda love this. It is incredibly hard work and we follow extremely strict SOPs set out by physicists for a reason. Heh while you didn’t set out to build a rage game. I think the difficulty is inherent to the subject matter and I like that your initial builds did not shy away from it.
That cable snapping event needs a lot of sparks lying all over the screen, with the cables continuing to spark as long as they are broken to really sell that something is not right.
13:10 When accounting for power generation, the higher amounts of power demands need to be weighted heavier. Under-producing power isn't an option, so the price has to lean towards the maximum, not the mean.
Honestly, at 4:11, I think you might be misenvisioning what it is that a good chunk of people look forward to hearing about from you. I'd reckon a quarter or so of the people who've wishlisted the SES DLC on Steam originated from RU-vid, stopping by on Steam primarily to support you and only secondarily because they want the DLC. And let's not forget that Patreon is a paid service. As is, there are undoubtedly plenty of people who would join the Patreon (and therefore have desires more aligned with the "Patreon" category on the diagram) if they could more comfortably afford it, but remain only within the "RU-vid" and/or "Steam" category in practice. These are just two of the forces at play and there are plenty of other forces and counterforces I could list off, but the point is that you might want to reconsider what it really is that the different groups of people are really looking forward to. Especially your following on RU-vid; it's common knowledge that just making videos doesn't cut it on this platform. There may even still be viewers whose only interest was 2D engine simulator and are holding out hope that the 3D version will be a cut-and-paste replacement for it! You do seem to have a good grip on the way you handle the different groups in practice, though. This video does a good job addressing a lot of points people might want to bring up on many different fronts, and paints a positive image of the progress made on the DLC. I suspect there'll be a lot of ES fans that won't watch it and several that'll click away halfway through, but for those who watch it in its entirety, I think they'll agree that your work on the DLC has been entirely justified and be satisfied with the video too. Also, at 13:00: in general, \frac{ \int g(f(t)) dt }{ \int 1 dt } eq g( \frac{ \int f(t) dt }{ \int 1 dt } ) for continuous real-valued functions f, g defined on certain real intervals (where the integrals are taken over specific time intervals), right?
I consider Steam to be a separate audience because there is an unexpectedly large lack of overlap between my audience on Steam and my audience on RU-vid. A lot of the people who play Steam Engine Simulator actually don't know this channel exists and don't know who I am. It's surprising but true and I've analyzed this data extensively. Only about 2000 downloads of the base game originated from my RU-vid channel initially. When the game went viral on Steam and was featured on the Steam home page, the game reached a significantly larger audience than even my RU-vid channel could reach (over 100,000 players organically with almost no downloads originating from promotion on my channel). Again, it seems counterintuitive but it's true. Furthermore, while this is only anecdotal, most people I interact with on Steam and Discord about Steam Engine Simulator, clearly don't know who I am and found me purely through the Steam algorithm. In fact, almost everyone that arrives on Discord via Steam doesn't even know that Engine Simulator exists. It's for this reason that the priorities of this group are surprisingly different from the priorities of my RU-vid and Patreon audiences. There is a large (but not complete) overlap between RU-vid and Patreon audiences but the priorities are clearly different. A lot of my viewers are game developers or software engineers and don't actually care to interact with my products. Many of my viewers who join my channel Discord server have never even downloaded Engine Simulator, so they're primarily interested in devlogs for the programming content, and not necessarily to get development updates. Anyway, the point is that the nature of the three audiences don't line up with what you'd think intuitively. Most would probably think that my RU-vid audience contains essentially everyone with Patreon and Steam being subsets, however, this is definitely not true. Maybe I should have explained that more clearly in the video. Your mathematical answer is correct, those things are not equal in general. Thanks for watching!
@@AngeTheGreat -- I will say that referral links to Steam probably aren't the most accurate. I found out about SES from this channel alone, yet I obtained it by searching it up on Steam. That might just be me being highly suspicious of links in general though, I have a habit of avoiding them if I can find the destination myself.
@@AngeTheGreat A question: I for example downloaded the game because of your videos and your channel. However, I did so by searching for your game on steam and not clicking any links. Would that count as a Steam or a RU-vid audience. Because such a situation might be skewing your numbers.
I don't think you should worry about making a steam train simulator. To all the people who really want that, check out the in development game Century of Steam, which is a 3d narrow gauge railroading game that simulates realistic locomotive firing, operation, and railroad management. Give Ange some breathing room, let him finish this up and get to engine simulator.
If they hated your game, they would be outraged at the turbofan test cell simulator I built years ago while working for Lockheed and Rolls-Royce at Stennis Space Center. I'm a former turbofan test engineer. Back in 2010 I designed and built a 6 screen stimulation, including a custom throttle, that simulated the testing of large jet engines. Specifically the Trent 9000, 1000, and XWB. The simulator had failures to test the crews reaction to problems as well as a playback of a real jet on our test stand that was synced to the simulated jet engine run condition. It was used to allow training of test crews without risk of damaging a real jet.
@@scottwatrous, it was interesting! I had a blast building and programming it for over six months. The test crews and other facilities also enjoyed using it. Better than destroying a 20+ million dollar R&D engine!
Yeah, when you're training to test 20 million dollar jet engines, and it's literally your job, the threshold for acceptable complexity and finesse is admittedly a whole lot higher than somebody who wants to have fun playing a little steam engine game for 20 minutes.
@@lucchesi87 well the Patreons just gave him honest feedback on the game which he asked for. Ange wasn't frustrated, he was just not sure what to do since he has a heavy workload right now and has to choose what to prioritize. but that has NOTHING to do with the Patreons, they just did what he asked them to.
At 13:00: The average-value operator does not (in general) commute with a nonlinear function. This could be "fixed" by having price-by-demand simply be a linear function, but that would probably be boring. Instead, an effective but approximate solution would be to sample demand at several points during the day, take the price at those points, and average those samples. As long as demand isn't too spiky and the price-by-demand curve is smooth, that should be close enough to the true value.
That's basically it. The precise solution would be using calculus, though. You'll have to redefine average as taking an intergral and then dividing it by the period over which it was taken. Then you'll just take the average of the price as a function of demand, with demand being a predefined curve in this case.
You need to integrate two different things. First, you integrate demand * fair price vs demand to get a total price for the amount of power needed that day, then you divide by the total demand for the day (integration of demand).
yer AI gizmos might be handy later if you ever implement automatic transmissions or EFI in engine simulator, those are both effectively black boxes trying to optimize something by flipping levers based on some engine condition parameters
I see that you included titles in the letters from NPCs. Are you planning to include an option to change title (Mr. / Ms. / Mrs. / Mx.) and such in game?
The thing about the throttle is that your AI system is basically exactly how real life steam engines controlled their throttles and were able to maintain extremely consistent RPMs. Real life steam engines in industrial settings used a centrifugal govenor, where the output wheel was linked to a vertical rotating shaft with two balls. As the wheel sped up the balls would swing further out and close the throttle and when the wheel slowed down it the balls would fall inwards and open up the throttle valve, you can find schematics for it in one of James Watt's historical patents. Basically a mechanical feedback system that was remarkably effective, in fact so effective that steam engines make for some of the most accurate mechanical clocks. Perhaps you should just add the centrifugal govenor as either a difficulty setting or an upgrade, you could include it along with other historical upgrades to the steam engine like a double acting piston, firetube boilers and compound expansion. Also a consequence of the centrifugal govenor is that historical steam engines were able to maintain a consistent RPM by varying their wattage.
Yeah. Responding to changing power demand or other conditions should be a challenge, but in reality keeping a power plant stable under constant conditions was a solved mechanical problem so the player shouldn't need to babysit it. Though, a breakdown of your governor forcing you to BRIEFLY manually control it could be an interesting disaster.
As someone who has developed software for years myself, I find one of the harhest parts to be how the majority of people simply do not have the prerequisite knowledge necessary to understand how much time and effort it takes to actually develop high-quality software. This is, of course, not their fault -- I can't expect everyone to know the details of software development. But it is one of those pain points every dev has to deal with, particularly solo/indie devs. Kudos to you for working through that and making a product up to your standards, while also taking harsh criticisms to heart and making big changes!
Honestly, I got into watching your vids by the Gen1 Engine Simulator videos I found while randomly hitting youtube. Now? Well, I'm here for the science lessons, and I'm waiting for the Steam Engine DLC. I know that it's going to be an excellent piece of work and I will definitely enjoy it. Keep up the good work and ignore the people who can't criticise constructively!
It's honestly incredible how much you accomplish in such a short timeframe. You don't really see people interacting with communities of this scale while single-handedly writing a massive project, editing videos about said project, and doing the same thing concurrently with another even larger project. How do you find time in your day to sleep lol.
16:30 I study computer svience and I got second hand anxiety from watching this video. I can't even imagine how stressful it is to undertake such a massive project ❤
I want the steep learning curve and no spoon-feeding. KSP frustrated the shit out of me but it was soooo rewarding when I conquered things, because it was just me vs a game mechanic. I feel like story modes in general detract from that, so if it's appropriate to add a sandbox mode with zero help I would definitely try that option first.
I know this is a selfish request, but I'd like you to recompile it for linux, so that I and the other players that have linux-based systems can enjoy it without too much virtualization
10:10 is that control theory sneaking in? Imagine a DLC for the DLC where you need to discover the plant transfer function, and tune a PID controller for the engine, maybe extend it to a MIMO problem :)
6:56 These points are amazing considering there are literal books trying to help people overcome this mistake. Profit First comes to mind which the tests might want to read lol
I love that the term AI has finally reached the point where a control loop is finally an AI 😂 Edit: yes, I do know that NPC behaviour usually referred to as AI too
I acknowledge that you edited your comment but I just want reiterate, LLMs and the "AI" of today has very much hijacked the term "AI". So you have the progression a bit backwards. Until very recently, the term AI for games had a very specific meaning and most of these AIs were very simple state machines and nothing more. My usage of the term is not incorrect, it's the fact that the term has been monopolized by LLMs and modern machine learning that is the problem.
@@AngeTheGreat In addition, the term is hopelessly overused these days. To the extent that it is about to or has already lost any meaning. See all the random claims of “A.I.” PC parts manufacturers. It is ironic that legitimate uses like Ange’s are now called out in what must be a sort of attempt to over correct this trend. The hype really must be nearing its peak.
Please find a balance to avoid burn-out. Negative feedback can be crushing. I wish you all the best with whatever you will decide. I admire the effort you put into this and your approach. You really are Great!
At the end of the day, ultimately what you are putting your effort into is honestly of little concern to me. The reason I continue to come back to your projects and channel has less to do with what specifically you're doing and more to do with how you're doing it. Your meticulous attention to detail and consistently impressive ability to design solutions to issues you encounter is really captivating, and it's clear that you have a great passion for the things you work on and want to do them right. This project didn't gain a large following because what you're designing was something people really wanted, but rather because people enjoy seeing skilled, passionate individuals perform their craft. Just keep doing things the way you enjoy doing them, and I've no doubt that the end products will turn out fine, and your audience will continue to enjoy watching you work your magic.
A lesson about how things can look way more intuitive than they actually are when you've built them. Also I find it fascinating that this game about operating a steam engine is getting a storyline. Well, always fun to explore unconventional storytelling lenses!
Even though I'm mostly here for engine simulator, I'm always just excited when another one of your video's come out because of the educational aspect of them all! Thanks for making such great content.
Does this confirm that you're doing engine sound design with beamng? All of your simulators are incredible for sound, this would be a really incredible collaboration to see
"Maximizing profit without progressing" ah yes, the inevitable end of capitalism. If it makes one person one more dollar at the expense of everyone and everything else, they're going to do it. What a wonderful way to run a society.
12:55 The reason taking an average doesnt work is because you need to root mean square. The average of a sine wave is its DC offset. RMS gives you the effective power delivered
You keep going. It’s insane to me the level of commitment and passion (though I suspect you might not like working on this as much as you did previously lol) you have for this. Nobody is perfect, but you’re doing pretty dang great
You didn't mention it, so I don't know if it's something you thought about but couldn't figure out the feasibility of, but when you don't want to do a tutorial yet want to fix the learning curve, one of the easiest and most intuitive methods is to adjust when players have access to features (similar to AI but less overwhelming). One example: to teach players to deal with expenses, a simple way would've been simply to disable most of the economy at first, only keeping expenses. Anyway, I know it's unlikely to be something you'd want to revisit, so good luck working on Engine Simulator!
You are a rare find as a game dev, thank you for taking the time to make things right and taking the time to listen to the community, that means a lot! Even if the community can sometimes be harsh and less than helpful! I appreciate all of your hard work!
12:54 is it that the demand -> price function is nonlinear? if it was a linear function, the (function of the average of demand should) be the (average of the function of demand) and linearity is a common assumption
I think my comment got autoremoved due to having a URL in it, so I'll try again. As a kid, I used to watch TV with closed captions, and it dramatically improved my reading comprehension. I'm also a big fan of Tom Scott's commitment to a11y. With those two points in mind, please (to steal a phrase from Tom) "buy some damn subtitles". RU-vid's auto-captioning just isn't the same, it fails miserably with technical content like yours. I suggest Amara, which is what Kurzgesagt uses. Amara also supports languages other than English. Hell, I'm a Grease Monkey patron and I'd be willing to help transcribe your videos for free.
@@hobbified That's why I said I'd be willing to help for free, to offload some of the work. I could have a deep learning transcribing tool ingest videos, and then manually touch up the subtitles. Or even start with YT's autosubs.
Ain't no way, I just randomly though "Lets check if Ange has posted a new video, it's been a month now" and here I am, just 1 hour after you posted it.
I loved the game, but I liked it as it was so simple. If it ended up the DLC was simply a train track with hills, brakes with some kind of engine wear, I would have picked it up.
I understand. This DLC was supposed to be simpler than it is now but it just didn't appeal to testers despite the fact that many people asked for it. One surprising thing that I learned as a developer is that high demand for something doesn't always translate into something that people actually enjoy. I want people to feel like they're getting decent value for their money. In any case, a train DLC may still happen at some point when I have the time. Thanks for watching
Hello, first timer here, this video has been recommended to me. It looks like a good project, and a great presentation too. Just one thing I've found is that it's really weird from what you've said: "One of the main objectives of the game is to increase the town's population", but I haven't seen a single number or graph that directly measures it, other than achievements. It's not that odd that the players keep maximizing profits when the big numbers on the center of the screen are about money, not the town's healthiness.
Kinda funny, I just came from watching LGR's "SimHealth" game review, which is essentially the original SimCity + healthcare system economics. SimHealth allows you to do some incredibly fascinating stuff with it, but it just wasn't fun unless you were already trained on the subject. I think you're doing a lot better than SimHealth in the "fun" department, and showing it's still possible to make a complicated, model-driven sim fun. Btw as a frontend dev, I must compliment the consistent feel you've built with the interface.
@9:33 I think it would be fun to have a "gremlin mode" where the AI maliciously messes with your setting lol. Also, please make a mode with no explanation, I'd like to try it. You also could do a tutorial type thing where the game starts with no explanation, the player accrues debt, and then a helpful person is like "hmm seems like you need a hand." Now the player has to work off that debt lol
I think its important to keep in mind feedback isn't creating problems, its identifying them., for example at 2:45, you mention getting the negative feedback wasn't a great experience. The negative feedback you get in playtesting phases, is doing nothing but IDENTIFYING problems and issues that already existed, you just missed them. it FEELS very overwhelming when you start the first round of playtesting, but these issues would either go unresolved and unnoticed if this wasnt done, and you would end up with a complete failure on launch. Playtesting/QC phases only seem to bring up new issues, in reality, they are issues that were there all along, but unidentified. I found keeping this in mind helped to take the sting from some of the playtesters comments
Bro there is literally no greater hive of scum and villainy than the steam update comments, for no reason whatsoever they are absolutely brutal and unhinged and are never happy, please never look at them
SE sim 1: haha steam engine go brrrr SE sim dlc: so anyways I implemented a multiparamter automated control model, automated testing, multifactorial price testing, NPCs, a story line, natural disasters...
I'm sorry, but if a throttle is stuck open, I am disconnecting the generator, emptying the boiler and removing the fuel source. Yes, this will cause a blackout.. But I am not running defective equipment.
6:58 "Some players completely misinterpreted the game and instead of maximizing population maximized profit by adjusting the electricity prices to always exceed their expenses..." You just described modern economy.
100% gonna buy this next paycheck. a friend of mine plays and loves it, and this cements that not only is it a great game, but its headed by a fantastic dev. great work Edit: holy shit it's free, absolute GOAT dev
I'll bite at the challenge mentioned at 13:02 . My first thought was Jensen's inequality (or at least closely related), but more broadly it looked as though the fair price function was non linear in the first place, so mapping the average post vs pre transform was always going to be different. Solution: apply the fair price function to the moment-to-moment demand function, then calculate your average from there. How'd I go?
You know what? The best way to care about your project is to care about you, so work on whatever the fuck you want to work if you ask me, (do NOT take this as business advice pls). I'm not invested (financially) in your work, i can't support you on patreon, though i do have steam sim on steam and have the dlc wishlisted, and i'm also interested in engine simulator... however what i take from your videos is: you'll do a great job, and the only way to get you to where you want to be, is to let you do the great job that you do, and to allow that to happen, just work on what you want to work. hopefully things'll go better over here and i'll eventually will be able to support you, i'd love to! Cheers
Thanks Ange, love a well detailed update that speaks to us like we're competent humans. Cannot wait for the DLC to be polished and released. Super hype for Engine Simulator to be finished as well.
The goal being population growth? I had to pause the video and study the screen, zooming into each section, before I found the thin lettered “population “ in the black area under the city picture. The money stood out like dogs balls. If you put the “population “ counter at the TOP, in bold and have some visual feedback “that more population is better” (there is some shiny reward for increasing it), everything else you hope for will solve itself.
You will never make everyone happy. And the loudest screaming group is not always the biggest one. You are the creator and I am excited to see, how _your_ vision of this DLC turns out! And I figure, there are lots of people who prefer your judgement on the development of features over the critics'. You really care about this project, and for me that's what makes it most interesting
Holy crap how can people be so stupid while being into something as complex as internal combustion engines? Every single thing you thought was obvious, IS obvious to anyone who gives a shit about power generation, utility supply economy and steam engines. Their responses to the learning curve and difficulty is also spooky similar and realistically like the nepotism hires in the real power industry, maximizing profits without properly managing the plant instead of maintenance and upgrades to keep up with the demand and treating the customers like human beings with a critical need for what you provide. If this was a very delayed and convoluted aprils fools I would have had a laugh, but i know it is not. Dont let them get you down Angie!
Month late comment, but; I don't care about the Steam release at all, however, I would so much much rather see you make something great without the disappointment of scrapping something you worked so hard on in the back of your mind. That would make for an uninterested and depressed Ange that nobody wants to see. I think your decision was smart and well thought out.
i think not having a tutorial is good, and there's something really really cool about playing around with the physics of the engine to find out what works and what doesn't and how/why. there's probably a really good balance between finding the physics well and not making it impossible to learn that you could figure out
not a game dev, but I don't think it'd be a bad idea to encourage players to develop skills by charging money for each aspect of the engine the AI is controlling. On top of being encouraged to learn to have the experience of running the engine, the extra incentive from profit could encourage learning from the AI worker. of course, with my limited experience making games (it literally only extends to semi-decent map making) I'm not in a position to properly say how to balance it. your game, your rules.