It has come to my attention that I misspoke in the voice-over, we had 16 players (not "6") players drop off in room E. It's still a fantastic conversion rate at that 96.1%, especially considering this is only 2 weeks worth of data and players can still come back to the game and complete that room after the date range.
Something super useful imo would be showing how you got the JSON from steam I've been digging but can't seem to find it in the documentation I can manipulate it into a useful format if I could just figure out how to pull it
That Room F that showed the high drop off of players was also the one where it was possible to accidentally skip the checkpoint so a death sent you all the way back to the beginning of the difficult Room E prior. Depending how your stats were hooked up I imagine that would count as completing Room E and quitting on F? Even though its really the thought of having to do Room E again that made people drop off.
Good call out, but no - it's set up so that completing a room permanently sets the variable to complete. So missing the checkpoint wouldn't ruin the stats. However, it could be another point of frustration in F specifically.
Hello Jordan. Thank you so much for mentioning, praising and showing our channel. It was a very nice surprise while we were watching it.😀 We are really happy to know that our video helped you a little bit. By the way, congratulations about all the valuable data you could gather to improve the game. Hope to see your next devlog soon. Regards.
I remember watching a lecture years ago about the game Antichamber (I think it was Antichamber), wherein the developer figured out during playtesting that it's very common that a player would stop playing right after a too-difficult challenge, instead of giving up on the challenge itself. I've had the same experience in games where if I feel that a level isn't fair I'd still insist on finishing it, but would quit right after. This all aligns with what you're seeing with people completing room E but then leaving in room F. I still also agree with the "alternative / extra explanation" of people having a hard time with that wall-jump section in room F, as I myself was slightly buffled in that section when I went through the playtest, as it didn't seem like you intended for people to be able to make that jump. In any case, I did finish the playtest and quite enjoyed it, keep up the good work. :)
@@cryingwatercolours I'm not sure if that's what you're asking, but I couldn't find the original talk (I looked for it, a lot), anyways what I remember is that they just learned to examine the difficulty of the puzzle before the one in which the player quit, instead of the puzzle where the player actually quit.
I'm really glad you emphasized that data should tell a story. That was one of the biggest aha moments for me when I learned about data visualization in college. I think a lot of people fall into the trap of thinking that data should be objective and the reader should make their own conclusions, but the truth is as a data analyst you are responsible for drawing conclusions and helping the reader see how you got there. You can give the reader tools to form an opinion and help them think critically, but you shouldn't just throw numbers on a chart and be as objective as possible, that just leaves them confused. Sorry for the long comment but this is one of my biggest pet peeves in data visualization. You did a great job!! I'm impressed
Wow, this was really in-depth! Thanks for putting this video together and for going into all the data! I loved hearing the story about the wall jump section in room E.
I'm so happy I got to be one of those players in the data. I ended up playing through the demo a few times, just so i could look for any flowers i couldnt find before, and also try to see how quickly I could beat the level! One thing i noticed while doing this was that when i walked into a new room, Mae would enter at the same place every time. I found this slightly disorenting when jumping off one of the buildings and appearing at the bottom of the room instead of still in mid air. Sorry to put this into a youtube comment instead of on discord, but i dont have discord so i hope putting it here is good enough!
I used to think that the best devlog series was made by Game Maker's Toolkit, with Mind Over Magnet. I say "used to" because now I truly think the best devlog series is yours. You guys make the perfect balance, of detailed, interesting, and consistent devlogs, let alone the amazing game your developing. I'm so grateful for you guys, and I can't wait to see more devlogs to come, I kinda don't want you to finish the game because I don't want this series to be over, but at the same time I really want to play the game. This series will be one that I will look back on and say "They were the best!" Also the music in this devlog was really good! Keep up the amazing work!
Great video! One note, you’re not considering whether the player is dying because of the level, or because they’re trying to get a blossom. For me, in room E, I didn’t find the room itself challenging, I actually just died like 30 times trying to get the blossom. Same with another room that was way before room E. It’s valuable to consider this data (even though I don’t know how you would track it) because you don’t wanna accidentally make a room worse because you didn’t know the actual reason why the player was dying so much.
This is so insightful, impressive how much data you can get by just connecting to Steam. Feels like steam isn't just a simple "buy my game", but a much more in-depth platform for you game.
I had a room that I took a TON of very quick deaths on that I was sure other people would struggle on, but when you finally showed room E in the video it was a totally different room than I expected! I wonder if i was just doing the other room the wrong way... I'm a moderately experienced Celeste player (300ish hours, lots of modded maps up to Expert difficulty) and the biggest challenge for me was the spirit movement. It felt a bit more floaty than I liked, I would prefer less acceleration and deceleration on keyboard for snappier movement and better response times to key presses.
Good grief! Finding your channel has been such a boon to my own game development. It’s been so hard to find quality devlogs that use Game Maker. You’ve brought to light so many features of game maker that no one ever mentions! Oh, and TetherGeist looks amazing, both mechanically and aesthetically
i think one big problem in the world of developing any product for a customer, is getting the "why". knowing why and contrasting it with what you are trying to provide is a crucial aspect of product development and engaging customer feedback. i liked the use of data analysis in this video, and it was an interesting watch seeing its use on your game as i am in the process of looking for a job as a beginner data analyst (aiming for a data engineer)
Great information on how to use steam to the developer’s advantage! Though be aware that data can easily paint the wrong picture if key information is missing.
Something that I've done with great success is to save recorded playback of players that drop off early. Going through those recordings is incredibly enlightening as far as where people have problems.
For me, the timed spikes level where so many players dropped off was only bad in the way that I cheesed it, and I killed myself so that I could try to figure out the intended solution. When I did figure it out, it was more difficult than another cheese Id found in the room, so instead of ever completing the level the intended way, I finished in two cheesed ways, and it took a long time to figure out the intention. I dont know if that is related to why others dropped off there, but that was my experience.
Oooo pretty numbers For me personally, the hardest part about that “hard room” was landing on the middle platform/pillar without sliding off. That’s been my main skill issue/pain point in general with the platforming: landing on narrow platforms without overshooting and falling off.
Super interesting, first I thought that the high death rate areas must be where people are quitting, but not necessarily as you showed. The high death, high completion area seems to me to be a good area, players are challenged, but are still engaged and persists through it.
i commented about room F and that exact issue on Steam, even though i persevered and got past the room Im glad to see I wasnt alone and that it was fixed :D
They have to right?? Too much money at stake in a AAA studio to not run analytics... that said, I'll bet they err too much on the side of optimizing sales rather than experience.
Just played it. Overall nice experience. I played with mouse and keyboard (i'm not sure if its meant to be played with the controller) but when i got into the spirit mode something was off about the movement. I am not sure if its too fast or whats going on but with the keyboard it seemed off. Also, it was weird that I had to press J to go into spirit mode instead of using the mouse. But I really liked the overall difficulty of the game and the art was really nice!! Great job and I can't wait for you to release the full version!
Thank you for the feedback! Rebinding is also an option in the pause menu if you prefer different controls. That said, we need to add mouse button support!
How dare you farm my gaming data to make your product better! Just kidding - nice in-depth analysis. Now we all need to speed run + 0-death the demo so Jordan will make it harder. 😉
I would love a Steam-GameMaker tutorial. I managed to get achievements working for my game but I never figured out how to get stats working. It would be a a very good tutorial for the GameMaker community to have. I was astonished by the lack of tutorials for this particular topic when I was trying to get it to work.
Awesome, yeah we've had about 5 comments saying they'd like it so I think it's worth making. If anything just to give back to the community that has helped us so much
I love this, I'm a big fan of analytics and is definitely incorporate something like this in games I make as well. One thing though, I can't information on this it seems, but are you releasing the game on any consoles? Your steams stats will be strictly for steam (right?), did you take that into consideration when selecting steam stats, or are you planning to use something else (if you are releasing on other platforms)? Loving your videos so far, thanks!
Yeah, it was definitely a constraint we are aware of - but the fact that Steam Stats is free and we're thinking trends should be relatively similar across platforms, we liked the solution. We'll definitely be releasing on consoles in the future!
Honestly kinda surprised the blossom flowers are only 4% average, i collected a ton of them Will say i'm glad the stats basically told that the game is fun, because it is, dying doesn't feel unfair, instead it feels like it's your fault and could've done better, you're either learning or getting better, and it's pretty cool
So, after playing the game and leaving my feedback, going back to feedback of the little girl and her dad playing. I would first ask them what spirit controls they were using because the default spirit controls you have for the game is BAD....it's not your fault, it's just that choice of controls. I am EXTREMELY happy that you added holding the button down to ghost because man, the first time I played, that was one of my BIGGEST issues. I would naturally want to hold the button down to ghost, and it wasn't until I finished the game that I looked at the feedback on the discord and even found out that those controls existed (oh thank goodness). I'd personally make the spirit hold controls the default controls. It's just so much more smoother and natural feeling. My guess is that child was playing with the default controls. I would recommend them to go back and play the game with the hold controls and see if that jump is still difficult or not.
Very helpful. Thank you! I'm curious, did your playtest version communicate to the player that their gameplay was being monitored to help make the game better? Ethically speaking, is that something players need to know? Should there be a way to opt out of data collection? I'm not sure of the answers. I'm genuinely curious what you think about the privacy angle, both for full release versions and for playtest versions. I can imagine reasonable arguments both ways.
Great call out. We actually looked into this, and since every Steam player signs an agreement with Steam directly, they've already consented to being tracked within Steam. That and we're tracking aggregated data so there's no way to know individual data without having their player ID.
My friend that doesn't play videogames saw me playing it incclass the other day and wanted to try. She struggled with the wall jump too. It can be a bit difficult with inexperienced players so I think that is a good change :)
I think a tye game should be more difficult, celeste for example seems more difficult, i think that if the game is more challenging is more rewarding when you overcome the challenge. And platformers usually have similar gameplay so it needs to be difficult to be fun and rememberable
have you tested putting a collectable near where a wall jump should start for the first while in t he game so the player builds the expectation of being able to jump at those angles
Nice analysis. Although the video title is a bit misleading, since the stats you managed to gather were only the total counts and you only have about 35 numbers. While I'd say that's alright for an indie game, it is far from "seeing everything". If you really wanted to "see everything" you would have had to log data per player. That would have given you more info like where did the less skilled players strugle the most.
I'm a game dev and I'm totaly stealing your data using Steam Stats. It's funny, too, because Steam presents it as a sort of afterthought system that is attached to Achievements (somehow), but it's NOTHING like that. It's just a basic (but powerful) telemetry system which you can use without a cookie disclaimer. Convenient!
Yeah, I'll bet its skewing things a bit. But I bet it's not too bad. We now have nearly 700 players, and a few outliers are to be expected. That, and most speedrunners start as any other player, dying several times before they get the hang of things. After that, you're not dying all that much in the rooms, but your existing data from when you first played the game is still there. I guess if you intentionally keep dying in the same spot over and over and over that would start to skew things. But still, it's just one data point of hundreds so things still smooth out to the almighty bell-curve.
Haha funny story about dying in the same spot over and over, I may have taken 20+ deaths in that one pit right after night town just to get the achievement
I think your math is wrong for room E. If 411 people completed room D and only 395 completed room E, then 16 people quit not 6. That's almost 4% quitting in a single room, which makes sense given the completion rate is ~96.1, I think. I think you may need to realize that steam playtest requesters are probably not the best sample of a general population either. They are already interested in your game and will persist where the average player will quit much earlier. TL;DR - check your math for room E. Also, I'd suggest refining room E a bit as well for the average and below average skilled players.
Oh whoops, lol 16 is right. Still, that's within the range we want to see. Especially considering this is just 2 weeks worth of data. Players can still come back and finish the room later, and that isn't shown in this dataset yet. And yeah, it's true this is a more exclusive group of hard-core gamers. But it's also our target audience. The type of folks who love precision platformers like Celeste.
True. But one thing I haven't seen mentioned in your devlogs yet is optional accessibility modes. Since you mentioned celeste, there is an assist mode in it where the player can toggle invincibility, slow down the game up to 50%, enable infinite jumps, etc. Probably a good idea to allow players the freedom to decide how difficult they want the experience.
I think it would probably be best to do a video about Steam integration with GameMarker, especially if like you said there isn't much information about doing so for GameMarker and it would be very helpful for GameMaker users who want to do something like this. Plus it could be a great piece of content for your UA-cam channel and getting possible new views/viewers.
It has come to my attention that I misspoke in the voice-over, we had 16 players (not "6") players drop off in room E. It's still a fantastic conversion rate at that 96.1%, especially considering this is only 2 weeks worth of data and players can still come back to the game and complete that room after the date range.
Something super useful imo would be showing how you got the JSON from steam I've been digging but can't seem to find it in the documentation I can manipulate it into a useful format if I could just figure out how to pull it
That Room F that showed the high drop off of players was also the one where it was possible to accidentally skip the checkpoint so a death sent you all the way back to the beginning of the difficult Room E prior.
Depending how your stats were hooked up I imagine that would count as completing Room E and quitting on F? Even though its really the thought of having to do Room E again that made people drop off.
Good call out, but no - it's set up so that completing a room permanently sets the variable to complete. So missing the checkpoint wouldn't ruin the stats. However, it could be another point of frustration in F specifically.
Hello Jordan. Thank you so much for mentioning, praising and showing our channel. It was a very nice surprise while we were watching it.😀 We are really happy to know that our video helped you a little bit. By the way, congratulations about all the valuable data you could gather to improve the game. Hope to see your next devlog soon. Regards.
Thank you and your daughter for the awesome playthrough!
I remember watching a lecture years ago about the game Antichamber (I think it was Antichamber), wherein the developer figured out during playtesting that it's very common that a player would stop playing right after a too-difficult challenge, instead of giving up on the challenge itself. I've had the same experience in games where if I feel that a level isn't fair I'd still insist on finishing it, but would quit right after. This all aligns with what you're seeing with people completing room E but then leaving in room F.
I still also agree with the "alternative / extra explanation" of people having a hard time with that wall-jump section in room F, as I myself was slightly buffled in that section when I went through the playtest, as it didn't seem like you intended for people to be able to make that jump. In any case, I did finish the playtest and quite enjoyed it, keep up the good work. :)
Excellent insights here. I'm glad you enjoyed the playtest!
do you remember their conclusion or solution to that point? i’m intrigued but not invested in the game lmao
@@cryingwatercolours I'm not sure if that's what you're asking, but I couldn't find the original talk (I looked for it, a lot), anyways what I remember is that they just learned to examine the difficulty of the puzzle before the one in which the player quit, instead of the puzzle where the player actually quit.
I'm really glad you emphasized that data should tell a story. That was one of the biggest aha moments for me when I learned about data visualization in college. I think a lot of people fall into the trap of thinking that data should be objective and the reader should make their own conclusions, but the truth is as a data analyst you are responsible for drawing conclusions and helping the reader see how you got there. You can give the reader tools to form an opinion and help them think critically, but you shouldn't just throw numbers on a chart and be as objective as possible, that just leaves them confused. Sorry for the long comment but this is one of my biggest pet peeves in data visualization. You did a great job!! I'm impressed
Exactly! It's always great running into a fellow analyst who just gets it :)
Wow, this was really in-depth! Thanks for putting this video together and for going into all the data! I loved hearing the story about the wall jump section in room E.
Thanks for watching! Steam Analytics seriously feels like a pretty under-utilized tool
I'm so happy I got to be one of those players in the data. I ended up playing through the demo a few times, just so i could look for any flowers i couldnt find before, and also try to see how quickly I could beat the level! One thing i noticed while doing this was that when i walked into a new room, Mae would enter at the same place every time. I found this slightly disorenting when jumping off one of the buildings and appearing at the bottom of the room instead of still in mid air. Sorry to put this into a youtube comment instead of on discord, but i dont have discord so i hope putting it here is good enough!
I'm so glad you're one of the players in our data too! Good call out!
I used to think that the best devlog series was made by Game Maker's Toolkit, with Mind Over Magnet. I say "used to" because now I truly think the best devlog series is yours. You guys make the perfect balance, of detailed, interesting, and consistent devlogs, let alone the amazing game your developing. I'm so grateful for you guys, and I can't wait to see more devlogs to come, I kinda don't want you to finish the game because I don't want this series to be over, but at the same time I really want to play the game. This series will be one that I will look back on and say "They were the best!"
Also the music in this devlog was really good! Keep up the amazing work!
Holy high praise, to be compared to GMTK, wow. You are far too kind! We're so glad we have you along on our journey!
By the way your channel is totally crushing it. Congratulations on your devlog 0 doing so well! It's truly awesome to see :)
@@OandCoGames thank you so much! I wouldn't be there without you!
Great video and great data. I also struggled in the same part that the daughter of Dual Gen Studios failed. Can't wait to see the game finished.
2:38 OMG I couldn't figure out how to get that flower! Thanks for the unintentional tip!
hehe, I think you can still get it returning straight up again, but it is significantly more difficult that way.
I was sure you had to get it from the right and then go straight up. Which i did but it was super hard!
Great video! One note, you’re not considering whether the player is dying because of the level, or because they’re trying to get a blossom. For me, in room E, I didn’t find the room itself challenging, I actually just died like 30 times trying to get the blossom. Same with another room that was way before room E. It’s valuable to consider this data (even though I don’t know how you would track it) because you don’t wanna accidentally make a room worse because you didn’t know the actual reason why the player was dying so much.
This is so insightful, impressive how much data you can get by just connecting to Steam. Feels like steam isn't just a simple "buy my game", but a much more in-depth platform for you game.
I had a room that I took a TON of very quick deaths on that I was sure other people would struggle on, but when you finally showed room E in the video it was a totally different room than I expected! I wonder if i was just doing the other room the wrong way...
I'm a moderately experienced Celeste player (300ish hours, lots of modded maps up to Expert difficulty) and the biggest challenge for me was the spirit movement. It felt a bit more floaty than I liked, I would prefer less acceleration and deceleration on keyboard for snappier movement and better response times to key presses.
Yes, please make a GM/Steam tutorial 🙏🏼
Its truly inspiring how much work you're willing to put in order to perfect your game, good job!
Thank you!
Another great devlog! Appreciate your showing so many different parts of your dev process (design, storytelling, music, analytics, etc.)!
Thank you thank you!
As one of the 546 I have never played Celeste but I still was fond of the amazing controls and art style
Good grief! Finding your channel has been such a boon to my own game development. It’s been so hard to find quality devlogs that use Game Maker. You’ve brought to light so many features of game maker that no one ever mentions!
Oh, and TetherGeist looks amazing, both mechanically and aesthetically
Hey that is so good to hear! Always excellent meeting a fellow GMS2 dev :)
This is really smart. I ve been using steamworks and achievements for a while and never thought about using for this purpose. Sweet!
Sweet, glad to give a new idea!
Wow. This is one of the most usefull game dev videos around there. Thanks for sahring this.
This is SO DOPE! Game has been wishlisted (long ago) and can't wait to follow your success.
ive never felt so entertained and excited to see data
i think one big problem in the world of developing any product for a customer, is getting the "why".
knowing why and contrasting it with what you are trying to provide is a crucial aspect of product development and engaging customer feedback.
i liked the use of data analysis in this video, and it was an interesting watch seeing its use on your game as i am in the process of looking for a job as a beginner data analyst (aiming for a data engineer)
Great information on how to use steam to the developer’s advantage!
Though be aware that data can easily paint the wrong picture if key information is missing.
I love this report you made :D
Something that I've done with great success is to save recorded playback of players that drop off early. Going through those recordings is incredibly enlightening as far as where people have problems.
Ooo yes, recorded playbacks are the king of insights!
Thanks for sharing such an in-depth view and analysis of your statistics! This will be immensely useful (once I finally finish my game lol)
For me, the timed spikes level where so many players dropped off was only bad in the way that I cheesed it, and I killed myself so that I could try to figure out the intended solution. When I did figure it out, it was more difficult than another cheese Id found in the room, so instead of ever completing the level the intended way, I finished in two cheesed ways, and it took a long time to figure out the intention. I dont know if that is related to why others dropped off there, but that was my experience.
Thanks for the additional insight!
Oooo pretty numbers
For me personally, the hardest part about that “hard room” was landing on the middle platform/pillar without sliding off. That’s been my main skill issue/pain point in general with the platforming: landing on narrow platforms without overshooting and falling off.
Oooo yeah, I should have mentioned. Our next build has more friction on the ground when the player lands to help prevent the slippage.
@@OandCoGames that's huge thanks!!
Super interesting, first I thought that the high death rate areas must be where people are quitting, but not necessarily as you showed.
The high death, high completion area seems to me to be a good area, players are challenged, but are still engaged and persists through it.
i commented about room F and that exact issue on Steam, even though i persevered and got past the room Im glad to see I wasnt alone and that it was fixed :D
Yes! I read that comment too! I should have included it in the video. You are not alone :)
Thank you for all these infos !
Very insightful video, great job!
i really look up to this team
I wonder if AAA games work like this, if not i would be extremely disappointed. Awesome content!
They have to right?? Too much money at stake in a AAA studio to not run analytics... that said, I'll bet they err too much on the side of optimizing sales rather than experience.
Didnt know you could do this with games, that sounds really handy
Maan, that was great!!! 👏👏👏❤️
that is really valuable, thank you
Just played it. Overall nice experience. I played with mouse and keyboard (i'm not sure if its meant to be played with the controller) but when i got into the spirit mode something was off about the movement. I am not sure if its too fast or whats going on but with the keyboard it seemed off. Also, it was weird that I had to press J to go into spirit mode instead of using the mouse. But I really liked the overall difficulty of the game and the art was really nice!! Great job and I can't wait for you to release the full version!
Thank you for the feedback! Rebinding is also an option in the pause menu if you prefer different controls. That said, we need to add mouse button support!
This is amazing.
How dare you farm my gaming data to make your product better! Just kidding - nice in-depth analysis. Now we all need to speed run + 0-death the demo so Jordan will make it harder. 😉
hehehehe don't tempt me.
I would love a Steam-GameMaker tutorial. I managed to get achievements working for my game but I never figured out how to get stats working. It would be a a very good tutorial for the GameMaker community to have. I was astonished by the lack of tutorials for this particular topic when I was trying to get it to work.
Awesome, yeah we've had about 5 comments saying they'd like it so I think it's worth making. If anything just to give back to the community that has helped us so much
I can’t wait to play this game! Just gotta wait for the Mac version to come out
I love this, I'm a big fan of analytics and is definitely incorporate something like this in games I make as well. One thing though, I can't information on this it seems, but are you releasing the game on any consoles? Your steams stats will be strictly for steam (right?), did you take that into consideration when selecting steam stats, or are you planning to use something else (if you are releasing on other platforms)?
Loving your videos so far, thanks!
Yeah, it was definitely a constraint we are aware of - but the fact that Steam Stats is free and we're thinking trends should be relatively similar across platforms, we liked the solution. We'll definitely be releasing on consoles in the future!
@@OandCoGames thanks for the reply! And I agree, the trends are probably similar :D
Would definitely interested in a steamworks tutorial
Good data
Sure, I'll try it.
Great video
6:45 I think more people like me just went to the first level you could die in to get the 100 deaths achievement
Very interesting!
I feel honored to have the achievement of dying over 100 times 😤😤😤
It is a true honor indeed :) For what it's worth... I've died 1145 times lol.
@@OandCoGames wow!!!
Amazing vid
How do you get the JSON you described at 4:25? I can set stats in the game, but can't find any docs on how to get any dump across all users.
Hey did you figure this out? This seems awesome and I've set up all the stats, but I can't find any report or json
Honestly kinda surprised the blossom flowers are only 4% average, i collected a ton of them
Will say i'm glad the stats basically told that the game is fun, because it is, dying doesn't feel unfair, instead it feels like it's your fault and could've done better, you're either learning or getting better, and it's pretty cool
Excellent to hear! That's the exact experience we're aiming for :)
7:30 i swear i died there like 300 times.......
So, after playing the game and leaving my feedback, going back to feedback of the little girl and her dad playing. I would first ask them what spirit controls they were using because the default spirit controls you have for the game is BAD....it's not your fault, it's just that choice of controls. I am EXTREMELY happy that you added holding the button down to ghost because man, the first time I played, that was one of my BIGGEST issues. I would naturally want to hold the button down to ghost, and it wasn't until I finished the game that I looked at the feedback on the discord and even found out that those controls existed (oh thank goodness). I'd personally make the spirit hold controls the default controls. It's just so much more smoother and natural feeling.
My guess is that child was playing with the default controls. I would recommend them to go back and play the game with the hold controls and see if that jump is still difficult or not.
Each player has their preference 👍 it seems that the hold/release is generally more popular so we’ll likely make it the default.
@@OandCoGames Oh, I'm so happy to hear that. It's nice to hear, when it's not only you.
Data is so cool.
Agreed
Very helpful. Thank you! I'm curious, did your playtest version communicate to the player that their gameplay was being monitored to help make the game better? Ethically speaking, is that something players need to know? Should there be a way to opt out of data collection? I'm not sure of the answers. I'm genuinely curious what you think about the privacy angle, both for full release versions and for playtest versions. I can imagine reasonable arguments both ways.
Great call out. We actually looked into this, and since every Steam player signs an agreement with Steam directly, they've already consented to being tracked within Steam. That and we're tracking aggregated data so there's no way to know individual data without having their player ID.
@@OandCoGames Thank you for the answer!
My friend that doesn't play videogames saw me playing it incclass the other day and wanted to try. She struggled with the wall jump too. It can be a bit difficult with inexperienced players so I think that is a good change :)
Ha! Which class? I have fond memories playing World of Goo in my economics 101 class lol.
@@OandCoGames Audiovisual history class hahahha sometimes it gets boring so I play the demo through the zip file
Thanks is very insightful, I would love a how to deploy this steam analytics with Game Maker.
I think a tye game should be more difficult, celeste for example seems more difficult, i think that if the game is more challenging is more rewarding when you overcome the challenge. And platformers usually have similar gameplay so it needs to be difficult to be fun and rememberable
Agreed. We're just getting started.
have you tested putting a collectable near where a wall jump should start for the first while in t he game so the player builds the expectation of being able to jump at those angles
Hey, that's a good idea
YEEEEEEES!
please, do the tutorial :)
Nice analysis. Although the video title is a bit misleading, since the stats you managed to gather were only the total counts and you only have about 35 numbers. While I'd say that's alright for an indie game, it is far from "seeing everything". If you really wanted to "see everything" you would have had to log data per player. That would have given you more info like where did the less skilled players strugle the most.
Tru tru. I wonder if Steam would start flagging us for sending too many data requests..
@@OandCoGames I'm sure your game will make them enough money to make them not care about the data requests)
I also use Game Maker. What resources are best for understanding how to set up Steam stats?
Honestly there’s not a ton out there - but the plug-in has some good documentation.
Please make that video about integrating it into GameMaker, thanks!
Hello, I love your game. I want to give you some advice for the channel, put the name of the game and your Twitter so they have more visibility
Im curious now- do firewalls prevent steam stats from gathering player data and or for devs receiving it?
That's a great question honestly... it's all done through Steam's API, so they might have documentation on it.
Someone share this video with Arrowhead Game Studios 🙏
I'm a game dev and I'm totaly stealing your data using Steam Stats.
It's funny, too, because Steam presents it as a sort of afterthought system that is attached to Achievements (somehow), but it's NOTHING like that. It's just a basic (but powerful) telemetry system which you can use without a cookie disclaimer. Convenient!
am I skewing the results by speedrunning??
also I noticed some new music in the background :)
Yeah that's something I was wondering about. Did he account for outliers like you and other speedrunners?
Yeah, I'll bet its skewing things a bit. But I bet it's not too bad. We now have nearly 700 players, and a few outliers are to be expected. That, and most speedrunners start as any other player, dying several times before they get the hang of things. After that, you're not dying all that much in the rooms, but your existing data from when you first played the game is still there. I guess if you intentionally keep dying in the same spot over and over and over that would start to skew things. But still, it's just one data point of hundreds so things still smooth out to the almighty bell-curve.
Now data points like pausing and skip cutscene actions, those might start getting skewed by the speedrun community lol
Haha funny story about dying in the same spot over and over, I may have taken 20+ deaths in that one pit right after night town just to get the achievement
@@OandCoGamesYeah, that's what I was talking about
please make that gms2 steamworks tutorial! for all us GameMaker users
122 deaths on average!?
I don't think I died more than 30 times xD
Whay you use game maker studio 2? Not godot or unity
Game maker was easy to learn, and we knew we wanted a 2D game :)
please do a tutorial about this!!
8:00 411 minus 395 equals 16, not 6.
Yup. Check the pinned message.
@@OandCoGames i saw you lost the only one who died on lvl 2. ^^
I want a tutorial please :D
I think your math is wrong for room E. If 411 people completed room D and only 395 completed room E, then 16 people quit not 6. That's almost 4% quitting in a single room, which makes sense given the completion rate is ~96.1, I think. I think you may need to realize that steam playtest requesters are probably not the best sample of a general population either. They are already interested in your game and will persist where the average player will quit much earlier.
TL;DR - check your math for room E. Also, I'd suggest refining room E a bit as well for the average and below average skilled players.
Oh whoops, lol 16 is right. Still, that's within the range we want to see. Especially considering this is just 2 weeks worth of data. Players can still come back and finish the room later, and that isn't shown in this dataset yet.
And yeah, it's true this is a more exclusive group of hard-core gamers. But it's also our target audience. The type of folks who love precision platformers like Celeste.
True. But one thing I haven't seen mentioned in your devlogs yet is optional accessibility modes. Since you mentioned celeste, there is an assist mode in it where the player can toggle invincibility, slow down the game up to 50%, enable infinite jumps, etc. Probably a good idea to allow players the freedom to decide how difficult they want the experience.
@@spiregg Absolutely. Accessibility is a huge deal to us that we'll for sure be prioritizing.
nerd lol
bahaha
I think it would probably be best to do a video about Steam integration with GameMarker, especially if like you said there isn't much information about doing so for GameMarker and it would be very helpful for GameMaker users who want to do something like this. Plus it could be a great piece of content for your UA-cam channel and getting possible new views/viewers.