I Scraped the Entire Steam Catalog, Here’s the Data

[@NewbieIndieGameDev]

(Links to the dataset and charts can be found in the video description) There were so many fascinating insights I couldn’t fit into the video, so I wanted to share a few that just didn't make the cut. 💰 Publisher Revenue: I was curious about which publishers dominate Steam in terms of revenue, so I calculated each publisher’s revenue by summing up the individual revenues of their games. While this doesn’t reflect what the publishers actually take home, it serves as an approximate measure of their commercial impact. Estimating revenue is complex, and interpreting these results requires caution, but I took into account factors like VAT, returns, regional price differences, currency conversions, discounts, and the platform's cut. That said, my approach has several limitations, so these results should be viewed as a rough indicator rather than a precise ranking. The most commercially successful publishers, according to my analysis, are: Amazon Games, Electronic Arts, Ubisoft, Xbox Game Studios, Bethesda Softworks, Game Science, CAPCOM, PlayStation, FromSoftware, CD PROJEKT RED, Valve, Bandai Namco Entertainment, Pocketpair, Paradox Interactive, Larian Studios, SEGA, Rockstar Games, 2K, and Activision. Notably, indie publishers are absent from the top rankings, which makes sense since indie developers typically don’t rely on traditional publishers and often have only a few games in their catalogs. What we see here are companies that not only have one successful title but maintain a steady stream of high-performing games. Overall, these names are important to know, as they represent the most commercially significant players in the Steam ecosystem. 🎮 Indie vs. AAA Genres: I found some interesting contrasts when comparing genre popularity between indie and AAA games. As I mentioned in the video, both groups feature genres like roguelikes and platformers, but these are notably more common in indie games (~9% and ~15%, respectively) compared to AAA titles (~5% and ~7%). What I didn’t get to include is that genres like turn-based strategy appear more frequently in AAA games, making up ~7.5% of their catalog versus only ~5% in indie titles. Looking at less popular genres, we find that real-time strategy (RTS) games are more frequent in AAA (~4.5%) than in indie games (~2.5%), while genres like walking simulators show up more often in indie games (~5.5%) than in AAA (~3%). MMORPGs appear in ~2% of AAA games but only ~0.5% in indie titles. Similarly, MOBAs make up ~1% of AAA catalogs, while only ~0.5% of indie games belong to this genre. These differences generally reflect the relative resource demands of each genre. Genres requiring more assets, larger teams, or extensive balancing tend to be less common in indie games, which often have limited budgets and staff. 📈 Genre Evolution Over Time: I also created a chart showing the evolution of major genres from the launch of Steam Greenlight, which allowed independent developers to publish games more widely, resulting in a rapid expansion of the catalog. Surprisingly, there were no significant shifts in genre popularity; the distribution has remained largely stable over time. 💵 Reviews vs. Price: I was wondering if there is a relationship between a game’s review score and its price. My hypothesis was that higher-rated games could potentially charge more, as we often assume that quality justifies a higher price. I plotted the median price for games across each review score category. Ignoring an unexpected drop in the “Positive” category, it seems that higher-rated games tend to be more expensive on average. This drop in the “Positive” range is interesting though. I found another study showing a similar trend, where games rated in the “Positive” range generated less revenue on average. Do these games earn less because they’re priced lower, or does the “Positive” rating lead developers to set a lower price? At the same time, it’s curious that “Overwhelmingly Negative” games are often priced relatively high compared to their scores. This suggests that higher prices might have contributed to those low reviews, with players feeling that the cost didn’t match the quality. 🐈 Cat-Themed Games: While looking at the best-rated games in the catalog, I tried out different ways to compare them. At one point, I used the positive review ratio (instead of Steam's rating) and included games with very few reviews. I found a quirky trend: for some reason, the rankings were filled with cat-themed hidden-object games (games about searching for hidden kittens). This led me to compile a separate top 20 list of "cat-themed" games, which includes some beautifully aesthetic titles, making this niche genre a fun part of the rankings. Here are those games: Cats Hidden in Jingle Jam, A Castle Full of Cats, A Tower Full of Cats, A Building Full of Cats, Purrgatory, Cats Hidden in Paris, Sudocats, An Arcade Full of Cats, Epic Battle Fantasy 5, Milo and the Magpies, CATO: Buttered Cat, Cats and Seek: Osaka, Cattails: Wildwood Story, Minami Lane, 100 Ninja Cats, Cats Hidden in Maple Hollow 🍂, 100 Funny Cats, Cats Organized Neatly, Cats Hidden in Italy, and 100 March Cats.

[@mahadevovnl]

You could make a video about each of these subjects and I'd watch every second of it.

[@joshuaneiswinter253]

In your "reviews vs price" bit, at the end where you are saying the review score is affected by the price.... one thing that is going on as well is a scam. Selling a crappy half-baked game at stupid price, buying it yourself and leaving a good review, then including it in a bundle that has some kind of discount and it makes it look like a really good deal. Then the user who bought this crappy game leaves the negative review. I'm explaining this very poorly. .. anyway, loved the data gathering and pointing out weird correlations.

[@jabo_o]

Poor paradox to be outprofited by palworld. Also damm 0.5% for indi mmorpgs is surprisingly high

[@opitir]

I would theorize that the relationship between high score and high price is the other way around: high price leads to more hesitation by buyers when considering the purchase, and more motivation to play the game fully. People who buy a more expensive game can be expected to enjoy it, simply based on the fact that they expect to have fun.

[@dudenarima2528]

I really want to see this project on github if there's a possibility : 3

[@lephtovermeet]

I personally feel the 1st person tennis cat simulator genre is severely under-represented.

[@gvcvbbhvbbccxcvn]

a game like this would be fire

[@Gl_3tch]

Seems like the market could be cornered with one great game! lol

[@Newsdee]

kzbin.info/www/bejne/gmrTaZSmrs5-h6M there was a Tennis Cat at least - maybe now it needs a 3D version in 1st person....

[@pmnt_]

I'd try the opposite: Adventure Game Simulator: Space Edition. With so many popular words, what could possibly go wrong?

[@osvaldoprado9906]

You made me spit my drink! lol

[@davidalty1018]

I feel like this data is an indie game dev team's dream. Very useful for understanding the market and how it has changed

[@NewbieIndieGameDev]

Hey, thanks for the comment! Honestly, I think these insights are more useful for sparking ideas than for measuring commercial potential. We definitely need to tread carefully when interpreting these results.

[@deletedTestimony]

Yes, it's less predictive than it is a snapshot of reality. Which is honestly more useful.

[@TaigiTWeseDiplomatFormosan]

０．０

[@bobshagit]

steam is not the place to find this data since its corrupted AF look at the "reviews" how many of them have like 30 minutes on the game then leave a great review.... then look at how many times they have done this..... they are paying people to give reviews!!! then they hide and bury the real reviews

[@Qefx]

Data driven game development! Just as usefull as data driven series and movies! Netflix *cough*

[@AdrianHereToHelp]

4:22 "Even the least-reviewed of these has over 500 reviews" On Steam, you need at least 500 reviews to be able to qualify for "Overwhelmingly Positive" (if you're under you stay at "Very Positive" regardless of percentage). I expect the same is the case for "Overwhelmingly Negative".

[@NewbieIndieGameDev]

Interesting. And these rules seem to be quite volatile, some of the games I mention in the video have new ratings on Steam already. The number of reviews shown in the UI vary quite a bit as well.

[@TaigiTWeseDiplomatFormosan]

0.0

[@TheMagicDragon-mm5dr]

Yeah, that's why there are only 11...

[@bobshagit]

steam pays for reviews

[@BVSSIC]

@@bobshagit no they dont lmao

[@nepdisc3722]

Space Game Simulator: Adventure Edition

[@MAGNETO-i1i]

😂

[@arthurbarros5189]

DLC: cat & cats

[@Sir_Steven]

This sounds like a crappy mobile game title

[@thedragonflytapes5727]

Shut up and take my money

[@01766F]

That's quite literally Spore: Galactic Adventures

[@LazarusBell]

My favorite thing about the tags is when users tag games with heavy amounts of microtransactions with psychological horror.

[@AAAAHHHHHHHHHH]

my favorite game with a funny tag is 5d chess with multiverse time travel having the survival horror tag

[@xeno-eh4pq]

This is such a well made video, was really surprised when I saw the view count.

[@NewbieIndieGameDev]

Thank you! 🙂

[@SolTomLive]

What the hell You made me look and now I'm shocked too

[@bluesusername]

Woah

[@bluesusername]

Get this man more views!!

[@pamplemoo]

also this is a surprisingly niche video, there's less than twice as many views as there are subscribers

[@bosmervor]

While the analysis is superb, I want to point out how great the visualisation and animation is. Everything is very clear and the explanation is precise and concise. As a data scientist myself, hats off to you, I'll definetely explore the data more. I think the last analysis is the most interesting: influence of certain games on the amount of the games developed after in the same genre or combination of genres. I think games like Stardew Valley, Hollow Knight, Terraria or Isaac probably sparked some interest in respective genres as well.

[@NewbieIndieGameDev]

Thank you very much for the kind words :)

[@amelavnyt]

I'm so happy that youtube finally showed me a well made video from a random channel with low views. Also, your voice sounds very similar to two minute papers and I love it

[@NewbieIndieGameDev]

Haha that's a compliment for sure, thank you!

[@Aurirang]

Same. I was a bit confused why this random video wa sin my feed (mostly consisting of relaxing music and christmas stuff). Usually i'd just dismiss such videos, because they're usually not in my line of interests for the time being. But i'm glad i did click on it.

[@nathanduncan6919]

Except he doesn’t sound like theres a full stop after every word

[@coob9678]

this is the type of video that gets half a million views once the algorithm kicks in

[@NewbieIndieGameDev]

Thanks for the optimism, he.

[@monafea]

@@NewbieIndieGameDev omg he was right

[@isaacwakefield5915]

Look like it's heading in that direction man​@@NewbieIndieGameDev

[@terrsus]

Yeah the material presented is pretty interesting

[@th3_ph4ntomreborn31]

​@@NewbieIndieGameDev already halfway to the 500.000 lol

[@ADVscout]

Tags for smaller games, especially free games, are very unreliable. Often there are games tagged with dating sim, psychological horror, simulation, etc. despite not having any of these elements.

[@saltysalt844]

Yeah most shitpost hentai game had world war 2 and psychological horror tags combination despite doesn't have them

[@johnjohnson3681]

​@saltysalt844 This has to be my favorite meme on Steam

[@minari99]

I'm releasing a free game on Steam myself that IS psychological horror

[@semiruu]

I believe someone commented to say that the games with tags for psychological horror are by users to show heavy use of microtransactions 😂

[@aqua4089]

@johnjohnson3681It is quite funny but unfortunately it also makes statistical analysis like this video’s and game browsing a lot harder

[@CodeMonkeyUnity]

Such an excellent video, awesome data analysis! Quite useful to explore all the graphs and come up with some unique ideas

[@NewbieIndieGameDev]

Thank you! It means a lot ❤️

[@EmptyHouseGuy]

What are the odds of this. Just randomly watch this guy plotting unrealized genre-mashups. 3rd one mentioned is "FarmingSim + Shooter" @7:15. I spit out my drink. I'm one year into my Farming Sim Shooter in space. The Game Dev Gods must be with me.

[@NewbieIndieGameDev]

🤯

[@TheMagicDragon-mm5dr]

@@EmptyHouseGuy Me, who's beaten year 10 Robusta in Atomicrops: AMATEURS!

[@MrTurbo_]

Well starting on a game is 1 thing, now you still gotta finish and release it

[@Sairof]

This format is amazing and I would love to see some more videos of this style. Hope this blows up so more people can see it

[@NewbieIndieGameDev]

Thanks!

[@CLOCKHEEDMARI]

Thank you so much for making your work public, you're genuinely making the world a better place with it. My dream is to gather and analyze data like this, it has to be one of the biggest interests in my life. Thank you over and over again.

[@mizohican]

2:57 "cube" is actually quite a popular and common name in the puzzle genre. This is such an awesome and useful analysis. Good work!

[@NewbieIndieGameDev]

Is it? Interesting...

[@tai-bere3500]

Constructive feedback for the narration: try a pop filter or maybe running a high pass filter over the audio track to reduce the sibilances and smooth out those sharp S noises :) Great video, loved the insights!.

[@AdrianHereToHelp]

5:34 I immediately knew two rhythm-based shooters off the top of my head lol. BPM: BULLETS PER MINUTE (a rhythm-based FPS roguelike), and Pistol Whip (a VR title)

[@NewbieIndieGameDev]

A previous version of the video showcased BPM: BULLETS PER MINUTE haha, I ended up taking it out, good that people know about it 😉

[@tuhan4780]

@@NewbieIndieGameDev Metal Hellsinger? :)

[@fernando47180]

​@@tuhan4780 Metal Hellsinger was the first one to jump to mind for me. Such a good game.

[@molopower2354]

I own both! And they’re absolutely amazing games !

[@obsidianflight8065]

Hellsinger and blasterbeat were the first things in my head. Interesting how many there are lol

[@aungthuhein007]

These data analyses where you're interested in the domain yourself are usually much, much better than the average "AI data analysis" video out there. Awesome work!

[@pepper3677]

this is the type of high quality content I want to see on youtube. Thanks for this, subscribed.

[@NewbieIndieGameDev]

Thank you!

[@zzmarx]

I would now like to play a First Person Match-3 Game or ... rather a text-based platformer: "You successfully jumped over the block. You see a large turtle approaching you. What do you want to do?"

[@NewbieIndieGameDev]

I would play it.

[@Ssimpleperson]

​@@NewbieIndieGameDev I would enjoy a text based Dungeons and Dragons platformer

[@Verchiel_]

i think it's criminal that the best match 3 game out there, with no micro transactions or online garbage like that, with actual depth to the gameplay Is a NSFW anime game. I would genuinely pay a fair bit to see a match 3 game with actual mechanical depth and a genuine story.

[@ultimaxkom8728]

**interesting pitch & trailer** I would play that... For 3 minutes.

[@anusaukko6792]

@@Verchiel_ Nah it's awesome, cause Huniepop is a genuinely good game. Even if you took out the NSFW aspects of it, the game would still stand well on its own.

[@memkakes1514]

That is one of the best video descriptions I've ever seen on this platform, I can see why you keep trying to send viewers to it throughout the video.

[@Trupen]

This data is amazing!

[@aer2648]

Super cool seeing you here! Love the content!

[@NewbieIndieGameDev]

Thank you so much!

[@p3tzel]

Amasing data and two amazing youtubers! Hi Trupen! 😄

[@XetXetable]

I'm not sure why you would expect the "female protagonist" tag to become more popular over time. 10% seems about right; which is to say "90% of games either have a male protagonist, a protagonist with no stated gender (e.g. factario), no protagonist (e.g. most puzzle games), character customization (e.g. most RPGs), or multiple playable characters (e.g. most fighting games)" seems right.

[@saveoursquirrels4241]

Video games marketing towards women or multiple genders has increased over time, I can see why it would be expected that female protagonists in games would increase along with that. I would like to see percentages compared to exlusive male protagonist games, games without a gendered protagonist, or games with both options. It's possible that games with multiple playable characters or gender choice also use the female protagonist tag.

[@Pmurder3]

@@saveoursquirrels4241 Female protagonists are over represented in story games or "artsy" games.

[@migzo9532]

I WAS just trying to get the steam Api so I can work with the tags few days ago to find unique game combinations. That saves me so much work. you are a legend, thank you!!!

[@A_Random_W33b]

As an indie dev, this type of data is really valuable. Would you consider doing a video about your methodology and process?

[@genthefrog18]

amazing video, the data is super interesting and the way you presented it is really easy to understand. really enjoyed it, hoping to see more

[@NewbieIndieGameDev]

Glad you liked it!

[@moomooheadme8843]

Not all games have protagonists so I think that "games with female protagonists" would be better measured in comparison to "games with male protagonists".

[@PatPatych]

It's stupid to tag any of this anyway. Like, who cares, which of the two genders your protagonist is

[@mayolover.]

theres no "male protagonist" tag

[@realwolfingen8896]

definitly, calling 10% "Suprinsingly low" without any godamn references is so lazy for a video like this, you need 10 seconds to check, even Overwatch2 has no female protagonists tag.

[@PatPatych]

@realwolfingen8896 The guy seems to be woke, sadly. Why evrn include this in the video

[@mayolover.]

@@PatPatych oooo i kiss men oooo spookyy

[@aggapuffin]

8:50 Thinking about it, the reason that games with female protagonists are surprisingly low might just be because there are a lot more games with either a customizable protagonist that can be male or female or a protagonist that isn't gendered at all, be they non-binary or not known to the player. There are also quite a few games with no real protagonist, but playable women and men instead, with Enter the Gungeon coming to mind.

[@FantasKanal]

7:00 pretty sure the most famous card game shooter is Neon White.

[@NewbieIndieGameDev]

Hey, I didn't know about this one! It seems it doesn't have any of the tags related to card games on Steam :(

[@AbbreviatedReviews]

Interestingly, Neon White doesn't have a "card" tag included. I'm not sure why given that it's a central theme of the game, but maybe it's because they're not really used in the traditional sense.

[@ultimaxkom8728]

There's so many games with "cards" without actual card mechanics that it's questionable if it's good for the game industry.

[@kormannn1]

Tags on steam are not really reliable source of information

[@TheMagicDragon-mm5dr]

Honestly, I feel like 'card game' is kind of meaningless. Deckbuilding is the real core of the genre, otherwise you get cards being used as placeholders for power-ups.

[@robotics_and_stuff]

I know how hard and time consuming is to put together a good video for YT, but you made it and it's only 1 month since your previous video. Good job! Thank you for sharing.

[@NewbieIndieGameDev]

Thanks!!!!

[@polarrobin]

It's super funny, I am developing the 4th member of the "Rhythm + Tower Defense" Category in Virtual Reality right now. :D This video is super awesome!! Thanks for sharing!!!

[@NewbieIndieGameDev]

That's awesome

[@BlueSR]

Really interesting and well presented video! Cheers👍

[@NewbieIndieGameDev]

Many thanks!

[@netking767]

I don't even know what other videos you have, but I'm subbing!!

[@NewbieIndieGameDev]

Hahaha awesome! Glad you enjoyed it.

[@JuniperDev]

Wow, you could do this every year like a game data "wrapped" with games released on Steam that year :)

[@NewbieIndieGameDev]

Hey, thanks a lot!

[@JusFinnZalar]

As an Indie game dev, this is incredible and very useful. The names and unexplored genres data is a goldmine, great work!

[@Garthaxius]

thank you so much for providing a very academic and thorough analysis of current trends in videogames, it's one thing to collect the data, it's another to actually study it and draw conclusions I'm sure this information will be useful to anyone thinking about releasing games on steam for the next couple of years, you should consider applying this same methodology to other platforms

[@dugetwin]

This might be one of the coolest videos i've watched in a while the amount of effort in it is insane good work dude!

[@mralexs]

5:07 C&C 4 wasn't a spinoff, it was the FINALE of one of the 3 series within C&C and was the last C&C game to be released until the C&C Remastered Collection. Its hated because they stripped out all the mechanics that people loved about C&C, turned it into a MOBA, and it had a really, really bad story. To make things worse, in order to unlock the full unit roster for SINGLEPLAYER you had to grind in Multiplayer. The game is really emblematic of EA as a whole.

[@Crafterz]

that’s definitely interesting data. really nice to see so many indie games in top 50 best rated, especially with both of top 2 being indie. it’s also interesting that out of so many games, only 12 of them have overwhelmingly negative reviews. data on combinations of genres is likely very useful. ...and visuals in the video are nice too.

[@КостяКиндалюк]

4:38 Buying a game that doesn't have an english page on steam and then writing a negative review about it not having english language is so smart

[@okipoki818]

smartest americans

[@calvinlittle7]

This was a very well researched and produced video! I wish more channels did high quality data analysis like this on KZbin. Subscribed

[@zeroyuki92]

If you're planning to make more videos with this data, I would suggest a deep dive about popularity vs quantity in genre. It is commonly known that for various reasons there are an overwhelming number of 2D plattformer games, but it is also a genre with very few success. We can divide them to four quadrants ranging from overcrowded unwelcomed genre to underserved well received genre. Thank you for making some of your (?) data public, I will have fun exploring it myself later. Subbing as well to see what other dishes you can cook with the data.

[@moxiousch]

Well this blew my socks off! Very impressive and funnily enough exemplifies perfectly how my data analysis intro course described the process and outcome of DA. Incredibly well done video on top, with attractive animations and presentation; your script was clean, flowed well and honestly you have a highly pleasing presenter voice and accent. This is super motivating to me to continue diving into data analysis, so glad to find this video!

[@megacarls9894]

Wow!! This project is incredib... RAW DATASET?! The're a bunch of marketing teams at videogames companies that must be mentioning you right now.

[@maksimpimin5594]

Actually very impressive work. As a data engineer myself I can imagine amount of effort to collect, clean and concise the data. And insights are quite interesting as well . Well done!

[@james-s-smith]

4:20 Dauntless released on Steam December 5th, and it's now the 4th-lowest rated game on this chart with 86% negative reviews from 1,117 reviewers.

[@anatoliansettler9922]

You've made a wonderful job there. I see a great quality of data there. Also, making a similar analysis of "games to stream" would have substantial value for newcomer streamers to explore new positions

[@kiriyalwasp]

9:50 i frankly have a hard time believing that only 8.8% of indie games are roguelikes

[@Sir_Steven]

The indie game market is gargantuan but most of the games go under the radar. You hear more about roguelikes because they tend to be more popular, but for every roguelike that releases in a year, there are like 12 other indie titles that you'll just never hear about.

[@ruaangrobler3035]

This is incredible on all counts - your process, the presentation, the insights, the sharing. Thank you so much for your good work. Respect!

[@Gigusx]

Interesting analysis! The only thing I'd add is some representation of different game engines and how they correlate with different genres, specs, maybe interesting patterns in reviews etc. The rise of published 3D games is also pretty interesting,

[@obsidianflight8065]

If only we could easily find out that information, it would be very interesting to see.

[@nayutadere]

Super high quality video from a small channel, what a find! Best of luck with your future endeavors!

[@pokemnfan1]

Cool analysis, but I reject the assumption that you can just combine *any* two popular genres together and make a successful niche game. Auto-battler + platformer is an obvious example, platformers are at odds with any gameplay that doesn't give you control of your character(s). I like pizza, I like ice cream, I wouldn't want to eat both in one dish.

[@ultimaxkom8728]

Ice cream pizza? Y-Yeah! Who would do that, right? Haha I never did!

[@obsidianflight8065]

My friends told me I was crazy for thinking I could put cheese and pepperoni with icecream and it would taste good... Also yeah an auto battler mixed with platformer? I genuinely can't imagine them at the same time, maybe a game that has both elements in separate places? Like the discrepancy between day and night stages in sonic unleashed.

[@ryuwaizu9087]

You totally could make an auto-battler platformer. It might be a totally new genre but you totally could.

[@TheMagicDragon-mm5dr]

You can twist the auto-battler genre to make it work, it just takes a lot of creativity. How about an auto-battler where you set up your guards to defend you from some kind of enemy while you run away via platforming? You could even have their battling interfere with you and the environment with some risk/reward stuff where your guys can fight more effectively at the cost of destroying terrain you need to platform with. See? It works. Try again. I mean that politely, I'm actually kind of curious.

[@NewbieIndieGameDev]

I agree. I wasn’t trying to say that the number of games in a genre directly points to commercial success. Do you think the video made it sound that way? I tried to be careful with my wording, but squeezing such a big analysis into a 10-minute video for a general audience is tricky. That’s why I framed these as opportunities for “innovation”, which doesn’t always mean they’ll be commercially successful. The charts are really meant to spark ideas. I picked a few pairs that I personally thought were fun or interesting, but I know a lot of combinations won’t appeal to everyone or might not inspire developers at all. I appreciate your thoughts! Curious to hear how others see these conclusions too.

[@haydenmarshall44]

This was a fantastic, interesting and educative video. The data was presented beautifully and the script was written in a way which made everything easily digestible and entertaining! Incredible stuff.

[@hermanatator]

I love the work you made on gathering all the data, and the style of the video is very polished. At the same time, I feel that the data analysis is very basic and could even be misleading. I will try to explain: - First point: In the "most common words" and "opportunities for innovation" charts you give information about the published games in Steam. That data focuses not on "what works", or "what users like", but on what developers have tended to prefer to create games about. Maybe the word "cube" happens to be a consequence of fresh developers doing their first game and wanting to do a very simplistic art, but in percentage terms most of the "cube" worded games tend to have really bad reception (it's an example). It would be much more insightful to get the most common words for published games and compare it to the "most common words on successful games" (filter by some 'success' metric you could define, like a % of games buyed, or rating, or a combination). That would give a picture of what "developers develop" vs "what users like". For example, maybe "cube" is a common word in published games but disappears on successful ones, and instead for some reason the "retro" word gains more relevance in the successful games. There, the message is clear, there is an oversupply of cube-themed games and under-supply of retros. Same with genre combinations: which combination is most loved by developers vs which one the players actually like? The current common words and combined tags (opportunities for innovation) charts answer the question "what is already there and what could be added", as if the games in Steam were a canvas of colors where all colors had to be included. To me, it's not. I would not develop a romance fighting game just because few like those have been developed, I would develop it if a had a really good idea about it (an artist wanting to express their art) or if I saw that there was a demand about it (the art being demanded by the people). - Second point: Top best / worst game tiers are a type of chart that tends to work well, because people love leaderboards, but in terms of data it's not really insightful. that's because there is too little data of the very best and very worst for it to be significant to learn something from it that can be generalised to all games. But specifically best and worst elements of any group are outliers, meaning that they are a consequence of many weird stuff happening at the same time, and require a more in-depth analysis that explained what was going on. For example, you point out that the worst rated games had many comments about being money-grabbers and full of bugs, but I am pretty sure that there must be many other games with the same scammy vibe and more bugs and crashes with much better reviews. Maybe one of these games was very aggressively promoted, which generated a lot of hype that transformed latter into rage by the users, or maybe it was featured somewhere where it got a lot of unjustified visibility for a completely random reason. Seeing that the worst games got comments like "it's a scam" kind of makes sense to my intuition, but doesn't really help me understand what's going on. To analyse what makes games work and what makes them fail I would get a group of about 100 games of a specific type that are very similar between each other and compare their ratings / "success metric" to categories of praise and complain in the comments, to see what things ("loved the story", "amazing art", "hated the main character") are shifting the success in one direction or another. Maybe I notice that games with "strong storytelling" comments get 5% better rating in average, and "amazing art" comments give 10% better rating. That would mean that focusing on good art will tend to have a stronger impact than storytelling. Same can be done on all games, although it will be less insightful because genres may differ a lot on what they prefer (character development in fighting games may not be such a good driver as in roleplaying ones). However, maybe the % amount of crash reports can give some interesting insight. I noticed some other things but none worth to mention. That said, I loved the chart about the top rated roguelike deck builders. I assume you focused on that combination as it's the type of game you like, but showing more "top 20"s from the most demanded genre combinations would be very interesting. I would love to see another video with more insightful charts. Again, I think you made an incredible work gathering and cleaning all the data.

[@sidhawker]

This is so great! Data analysis and indie games are two of my hobbies, so I love seeing all this data (and having the chance to dig around in it)

[@someonerandom704]

You should compile data from systemrequirementslab to see trends in how hardware intensive new games are in relation to their genre. Like are most of the super intensive games AAA, etc.

[@MG-mh8xp]

oh my god as someone who used to have a potato PC I WOULD LOVE TO SEE THAT

[@anonymousontheinternet1309]

Big kudos not only for acquiring all the data but particularly for the appealing presentation. Cheers

[@Will_Forge]

Steam has so many games in the lost because they give their games away for free as a loss leader to help newcomers to steam fill out their library. Everyone has Half-life 1 & 2 by now.

[@descai10]

This data is also great for seeing which genres are likely to be hit-or-miss and vice versa

[@thenicesven5328]

Great Video and Interesting Insights, insane that you only have 5k subs

[@NewbieIndieGameDev]

Thank you for the nice message!

[@gborgesw]

My man, this entire thing is a piece of art in itself. Great presentation, very interesting use case to apply data engineering and analytics. So cool!

[@IanCannonPiano]

This is a really awesome video- Thanks for your hard work! The only thing is I had to watch with just subtitles and no audio because the "S" sounds are really harsh and triggered my sound sensitivity. You deserve way more attention though, keep it up!

[@noart4you]

Your visual representation is insane man, great work. Keep it up !

[@FlippedClips]

Hey boss. Consider looking into some noise and echo removal for your audio. Bring that quality up to match the rest of the incredible presentation. OBS has some built in audio filters or something like NVIDIA broadcast has some one click stuff too. Whatever you're editing in probably has some built in effect chains as well. Even some pillows around the mic during recording. All free stuff, you usually don't need new gear. This is crazy good stuff, but you sound like you're coming from an abandoned building right now lol just a suggestion brother, you rock!

[@kingszeno]

Wow, I was searching for exactly such a database to filter my preferences. Great work!

[@arthurbarros5189]

3:50 it saddens me that Dwarf Fortress is not there... but I am soothed, because Deep Rock Galatic is. . I don't want to remember Doki Doki Literature Club exists.

[@fieryelf]

I think you did a really good job with the data analysis part. It's very well structured and easy to understand.

[@NewbieIndieGameDev]

Thank you!

[@srkem]

8:56 I mean... There's no way to compete against protagonists who either a bean or just human arms...

[@MrTuchat]

Your production values, pedagogy, and engineering rigor are through the roof. Honestly you'd do really well making a bunch of thematically different videos such as how to make such beautiful presentations, how to figure out the BI/data analyst stuff to obtain and process data, and how to design infra solutions for efficient (and rate-limited, surely) scraping. You have a rare combination of talents, and this video was great. Also your dataset and analysis could be genuinely sold out to game companies, although I surmise this is already done at a premium by 3rd party consultants.

[@NewbieIndieGameDev]

Much appreciated!

[@marularch]

As with all kinds of data, this should be taken with a grain of salt. Games that don't have "female protagonist" tag might be just games without any specific genders or protagonists at all. They might also be "choose your own gender" type of game. This statistic would only make sense if there was a "male protagonist" tag and we compared the two.

[@NewbieIndieGameDev]

Yes, absolutely! It's very hard to arrive to any conclusions with the data we have publicly available. Thanks for commenting!

[@devent7098]

the quality of the editing and effort behind the video is great. liked and subbed, goodjob

[@fcurdi]

Awesome video. Interesting findings and clrealy explained 👌

[@NewbieIndieGameDev]

Glad you enjoyed it, thanks ❤️

[@FenTheWolf]

The amount of effort and work that must have gone into this video is astonishing. And the editing is amazing aswell. Very nice work!

[@Fafr]

By the way, a general note to people: Steam tags that represent relatively niche genres are often misunderstood by the general community, I see that happening a lot with rhythm games and bullet hells. Well over half of the games with the bullet hell tag are misrepresented by them, like Vampire Survivors clones for example aren't bullet hells, and shmup doesn't necessarily mean bullet hell either. As for rhythm games, people tend to use the tag for pretty much anything where music plays a key role, including games like Geometry Dash and Just Shapes & Beats, which is also wrong. So yeah, when dealing with tags on games on Steam, always take them with a grain of salt if the genre in question isn't mainstream like FPS games are

[@leotaku5216]

I'm not sure I agree. This is kind of the prescriptive/descriptive debate in linguistics. If people find it useful to describe Geometry Dash as a rhythm game, who are you to say they are wrong?

[@Verchiel_]

I'm not sure how relevant this is but often times i've seen games tagged with the most non sensical tags, probably most infamously games being tagged with souls-like just because a game might have a dodge roll or a stamina system, or just have a gritty artstyle.

[@Fafr]

@@leotaku5216 Well, the only people who call Geometry Dash are those who don't play Geometry Dash _or_ rhythm games. As someone who's played rhythm games and GD both for a long while, and been active in the communities of both, there are tons of things that put GD apart from rhythm games, main one being that you simply can't properly sync your levels with the song, there are almost no tools for that in the editor, and those that are there barely work. The community doesn't use the music as much of a gameplay thing either, rather, it's mostly used to set the mood/atmosphere of the level, that's especially prominent with extreme demons, which are the most famous part of GD, where most inputs don't really depend on the song. The most widespread example of that is straight flying, where you have to click with a certain specific consistency regardless of the song And, GD isn't the only one in that vein as I said earlier. Just Shapes & Beats doesn't have the player do anything to the song at all. Or how about Super Hexagon? That one's often falsely called a rhythm game too by outsiders, even though absolutely nothing there happens to the song aside of some pulsing visuals. Most people falsely calling games "rhythm games" just because something happens to the song are either not invested in rhythm games at all, or are very casual players

[@Fafr]

@@Verchiel_ Ah, seems like a similar issue of games being falsely tagged as bullet hells, where people often tag games as that in any occasion of there being dodgeable bullets. And it makes it so hard to look for games, too

[@ultimaxkom8728]

Well, as a human on planet Earth, i'd say that if a game use music then it's a rhythm game.

[@laiag4854]

This analysis is amazing in so many levels. Thank you for adding an inclusivity section! New sub

[@PhantomThiefXI]

8:10 do you want a hot dog? or a key to the ship?

[@Flit_tv]

The amount of time and effort to sift through so much data is insane. I'm just baffled and impressed especially for a small channel how insanely good the quality of this video is

[@MiEHaLo]

8:47 HUNT DOWN THE FREEMAN

[@_Harid_]

You did great work ! It is also interesting to see what's going on in the Steam in general. Thank you for publishing all the charts from the video.

[@NewbieIndieGameDev]

Thank you very much!

[@Metacious]

I was looking for copper and found gold Really good stuff!

[@NewbieIndieGameDev]

haha thank you!!!

[@lostGash]

Great video, with some really interesting insights! I'd love to see a follow-up with some more observations. This data seems like it would be so fun to play with!

[@froginabox]

is there any chance you could publicize the database? i would love to explore it further (p.s. in terms of roguelikes, check out slice & dice! i've been playing it for over a year and it recently released on steam; definitely my favorite game in the genre)

[@NewbieIndieGameDev]

I thought about it, but I need more time to figure out how to do it exactly, it really is a lot of data. I don't get into this in the video, but for example I have the full descriptions for every game, that's a lot of text haha. Maybe I can filter relevant datapoints and upload an export to kaggle or something? I could also repeat the process and update the data first, this is a couple months old already... And the game looks amazing, thanks for the recommendation!

[@lumi-lumi-lumi]

@@NewbieIndieGameDev you can split different aspects into different tables / csvs, so that people could only download what they need. To join data back together steam id should be good. And keep everything! I’m pretty sure some freak would finetune some LLM on those descriptions, lol

[@leandromartinveiga9608]

@@NewbieIndieGameDev i beg you please do it, it would be amazing, you could, auto-export all tables to csvs, depending on the db you use.

[@cynth4941]

​@@NewbieIndieGameDevI also would really appreciate it!

[@mateo-hm]

So interesting and so nice to see a quality process and video editing!! Congrats

[@walterclements]

A theory for the inclusivity tags' unclear trend is that as they become more normalized, users and publishers may feel less of a need to include the tag as a special thing. e.g. none of the portal games have the female protagonist tag plus a large proportion games have a neutral first person (e.g. puzzle games, simulator games), or no clear protagonist (e.g. hero shooters), or a customizable character (e.g. most mmorpgs), so the actual proportion of games with a clear protagonist could be rather small as well

[@Dalleee234]

incredible presentation of your analysis and probs to Flor and Natalia! Also, thank you very much for providing all the data, links, etc. 🙌

[@TheKillamanJoe]

Out of all the things you could have focused on in such an interesting study, your two highlights are… Inclusivity and Roguelike Deckbuilders?

[@pimpcat7772]

Did you only watch the last minute of the video or are you so braindead that you need to nitpick every detail of this man’s video?

[@TheKillamanJoe]

@ it seems like my simple and innocent question somehow offended your totally rational sensibilities. Would you care to elaborate on why, champ?

[@dugetwin]

those weren't the two highlights of the video? That was like the final chunk of it. Most of it was talking about most reviewed games and how certain genres arent as combined as others, did we watch the same video??

[@TheKillamanJoe]

@@dugetwin the majority of the video covers a broad range of games. But out of all the specific tags / categories / genres that could have been covered in this video, only “inclusivity” and “roguelike cardbuilders” have their own dedicated sections. An odd choice, all things considered.

[@dugetwin]

@@TheKillamanJoe Well inclusivity is an interesting metric to see how acceptance has grown in gaming over the decade. On top of that he said he loves deck building rougelikes and as its his video he is fully allowed to highlight a genre of gaming he loves. Finally he says multiple times it could take him hours going over everything so you can look at the data yourself and have fun with it

[@michaelbentley1591]

Could watch hours of insights on this data. Excellent work!

[@silvernull]

the plosives on the mic kinda hurt to listen to lol

[@NewbieIndieGameDev]

Thanks for the honest feedback, very useful.

[@magicaCZ]

i like his mic

[@JJRol.]

@@NewbieIndieGameDev I don't think it's the plosives, it's more the fricatives, the "s" was a little loud. This is a minor nitpick as it was a very, very, good and interesting video.

[@NexGenration99]

cube showing up a lot makes sense because of voxel games. i imagine paper might also refer to games along the same vein as paper mario

[@itj]

Why is 10% female protagonist bad? What is the percentage for male protagonist?

[@BlueKanary]

That's the thing. There isn't a male protagonist tag.

[@itj]

@BlueKanary i see. So 0%

[@itsfela]

I been looking for a video like this for a couple months now. Thanks for the data and the insight of your gathering process. Useful!

[@AdrianHereToHelp]

I love the idea of using the tags to look at inclusivity across Steam. I wouldn't have thought to check that, but it's cool to see and definitely feels like an important thing to consider

[@IndieGamingRoguelikes]

That's an incredible video! Being also a web developer other than gamer (not in data analysis though) I'm really interested in this work and it's outstanding. I would love to see more analysis on this data, as you mentioned in the pinned comment. Visuals help to understand better the content. This is an incredible work, congrats for this! Can't wait to see more of these analysis

[@TheCollectiveHexagon]

2:03 couldn't you have just used...Steamdb, yk, the website specifically made to be a steam database? that already has an API for it?

[@fredecarlslund5407]

He mentioned other services not having the data he wanted and showed a picture of steamdb.

[@quasiportal6138]

Very insightful and thorough analysis, the presentation and polish is astounding, I just finished a Data Analyst course and feel like a noob, I wish to achieve this level of polish someday.

[@nomadshiba]

video ends at 8:40

[@shirothefish9688]

?

[@jupitersky]

No it doesn't? There's so much useful data after that.

[@RichConnerGMN]

wah wah wah

[@Hepponisme]

Thanks

[@monke6075]

I get it!

[@afinelad9366]

This was incredible! Like - from a scientific point of view, this is the kind of presentation I want to see professors give when they talk about game design. It's so well crafted, well put and interesting. Major props!

[@PatPatych]

9:05 Great to see the LGBT tag usage clearly drop in 2024, we don't need any of that garbage in our games.

[@mayolover.]

@@PatPatych idk you kinda look like the weirdo here

[@PatPatych]

@mayolover. I don't push any unnatural and unwanted agenda, unlike *some other group*

[@mayolover.]

@@PatPatych whats ur opinion on vore

[@PatPatych]

@mayolover. and that's exactly the reason people don't like your kind. Should've focused on "love is love" part instead of trying to make your fetishes socially acceptable

[@TheRealGurasic]

This is what happens when you drop out of school, as you are unable to read a graph while this year the tag got les usage the trend only indicates an increese in use, this specificly matters in vieogames were they take soo long to develop mesuring it in years is meh

[@krum1985]

I love this kind of data analysis. So much interesting things to learn. Great job 😊