Launch HN: Temper (YC W18) – Get more sleep by blocking games at night

33 points by hanrelan ↗ HN
Hi HN! We're Rohan and David, founders of Temper (https://www.temperapp.com).

Temper is a desktop app we built to help gamers get more sleep by blocking games at night. You can set a schedule (eg. no games after 11pm) or set a daily time limit (eg. no more than 2 hours of gaming per day) and once you hit those limits, Temper will block you from playing more games until the next day. We know nobody wants to get blocked in the middle of a game and get an abandon, so unlike most blockers, Temper is game-aware and uses traditional computer vision (specifically, blob detection to identify elements) to ensure the block only activates once a game ends.

We started Temper because we're gamers ourselves (over 5k hours of Dota 2 between us!) and believe games are positive experiences that can help you de-stress and keep in touch with friends. But we also realized that we regularly stayed up late at night playing just "one more game" instead of getting a healthy amount of sleep. After surveying other gamers and looking at the research, we discovered this was a really common phenomenon, with international studies reporting that 9% of gamers overuse video games. So we decided to build Temper to help people (and ourselves) maintain a healthier relationship with gaming.

If you're a gamer who struggles with the same issue, please check us out and tell us what you think! It's Windows only for now and we support League of Legends, Dota 2, Heroes of the Storm, Overwatch, PUBG and Hearthstone. We're adding new games every week so if you have any requests let us know.

56 comments

[ 4.8 ms ] story [ 119 ms ] thread
Do you need computer vision to tell a have is running? Do you need an app to prevent you paying games? Did you start a for-profit tech company just to "help" people?
The computer vision is not to detect if the game application is running, it's to detect if the game application is running and the user is in the middle of a "match". In other words, if the application can be shut down without negatively impacting the player's stats or team members.

I don't think that whether or not someone "needs" an app for self-control is the question. It's whether or not outsourcing that self-control is useful. I've seen many comments from people who get benefit from blocking HN, reddit, and FB for specified periods of time.

We don't need blob detection to tell us the game is running, but one of the core features of Temper is that it doesn't block you while you're in the middle of the game. To do that, we need to know when the game ends which we do by capturing the game screen and detecting when UI elements corresponding to the end of the game show up.

Definitely not all gamers needs an app like Temper, but we have noticed a fair number of people (including ourselves) benefit from the additional nudge to not play that extra game or two and go to bed instead.

To your last question - we do think that we can both be for-profit and build an app to help people. One similar model that we often think of is apps like MyFitnessPal - a tech company built to help people lose weight.

All good questions, thanks!

> "Do you need an app to prevent you p[l]aying games?"

This is like asking if people need drugs/therapy/whatever to prevent them from smoking cigarettes.

Games are addictive when designed innocently, and many modern games are designed to create addiction. Loot boxes are possibly the most visible example right now.

> "Did you start a for-profit tech company just to 'help' people?"

Helping and charging people are not mutually exclusive. I'm certainly not a hard-core capitalist, but I do know many people whose passion is to help others and the rate that they charge is to allow them to have good quality of life at the same time.

How are you going to block a CIV 6 game? ;)
Premium. A bunch of thugs break down your door and reads Tolstoy to you until you fall asleep.
We have some ideas for dealing with single player games like putting an overlay over the screen or slowly grayscaling it, but we've decided to focus on multiplayer games for now. I also kind of like this Tolstoy idea...
Just start making the AI take forever on their turns until you give up.
Do you really need computer vision to detect when someone is in a game? There are certain static elements on the screen in every game that you could probably just take a current screenshot and match against known images?
We tried matching against known elements (template matching), but unfortunately that didn't hold up well given the large number of resolutions games support.

One problem is that the elements themselves don't necessarily scale proportionally. For example, if you have a button with some text in it and you reduce the resolution, often the game will keep the text larger but reduce the padding around the text. This makes template matching tough.

We found blob detection works pretty reliably and uses minimally CPU, but we're not CV experts so if anyone has other ideas we'd love to hear them!

It would be nice to see a metric for performance impact of this blob detection so that gamers with less beefy computers don't have to compromise their gameplay settings to accommodate for this app
We don't have official benchmarks, but you shouldn't see much of an impact at all since we only do it once a second and on a small portion of the screen. If you do notice a difference please let me know.
Can you read from memory? Speedrunners use programs that auto-split (detect when the runner has reached a certain stage of the game), I think by reading from memory.
We thought about doing that but didn't explore it much to be honest. One issue we have to be careful of is not triggering the anti-cheat features of these games. I'm not sure if looking at memory without modifying it does that.

The blob detection does work surprisingly well though, doesn't take us very long to add a new game using it. Most of the time is spent making sure we understand the flow of games we don't play ourselves.

I’m assuming you can’t just look at the active process because the game might be in the foreground but the user is in a lobby. I would look at the bandwidth profile of the game over 30 second windows. I bet it looks way different while playing a game vs. sitting in the lobby, and you can watch the traffic without modifying the binary of the game (at least, you can watch the relative volume of traffic, and the IP addresses it interacts with).
couldn't you just look at windows processes and keep track of how long its running based on when a user starts the timer. Really wouldn't an annoying timer do the job just as well ?
Do you do the computer vision part on the end users machine or your server ?
End-user's machine. The images aren't saved and never leave your machine
You could just detect key presses. For most games 30 seconds without wasd or 1234 means you're in a menu.
We do use that for idle detection (so we don't count it for your total played time). But for the end-game detection we need to be more accurate because we don't want to accidentally trigger in the middle of a game (maybe you're waiting for a respawn and stepped away from your comp while dead) and we want to immediately trigger when the game ends so you can't queue for the next one.
I'm a target user and installed Leechblock extension for that purpose to block op.gg

Really smooth download and config. Kudos. I like you don't make me register right away.

I was surprised UAC came up twice though.

Tangent: Recently I found that "3 games" a night really changed my approach to league and gave me more quality and less stress to start another game etc. etc.

Thanks! David put a lot of effort into making the config really intuitive so he'll be pleased to hear that. The UAC is once for the installer and once for the app since it requires admin privileges to apply the block.
I don't know spefically about your internals, but you could try to install with user rights only (= no UAC) and then only do UAC when you need the block?
If I remember right the reason the installer needs admin rights it to be able to kill and overwrite a previous installation of Temper
How hard is it uninstall and reinstall?
It's straightforward using standard Windows add/remove. However if you uninstall while blocked, you stay blocked until you reinstall, so you should only uninstall while not blocked if you want to get rid of the app. If you do happen to uninstall while blocked and want to remove the block contact me and I can tell you how to do that.
I'm interested in what your thoughts, if any, have been to support MMO's. I think it'd be easy enough for folks entering and exiting instanced content to detect, but for folks doing something like crafting or gathering, which are generally repetitive tasks that don't change the field of gameplay much to demarcate a "done" step.
This is a good question and something we've been thinking about. We don't have a good answer for non-episodic games, but it might be something along the lines of slowly graying out the screen after you pass your cut-off.
Neat idea! Is there a mac version in the works?
Not at the moment - we're focused on Windows and increasing the number of games for now. But we'll definitely keep it in mind
I wrote a program which would take an initial seed password and run a number of sha1 iterations to get a digest back. I used it to lock myself out of websites such that I couldn't get back in without dedicating N hours of serial cpu time. In hindsight, I had invented my own proof of work system.

Never had that sort of addiction with video games, but I can understand it from that experience. I'd only fear that I'd quickly learn how to hack this tool and disable it whenever I wanted.

Haha that's a really clever scheme
i'm at a loss here as to why people bother installing these kinds of apps (the focus time apps overall i mean) to prevent them from doing something? if i have this on, can't i just simply kill the process through the task manager and be done with it? genuine question here...

bearing that in mind, i'm curious to understand what was the rationale from YC to have them into the W18 batch from a business/revenue standpoint?

Oftentimes, having just one step in the way of something is enough to break the habit. Like, move the cookies off the counter and into the cupboard above the fridge even works.

This is just another application of that.

But then why not just make the shortcut to the game inconvenient to access (hide it in a folder in a folder in a folder) or just plain uninstall the game (adds the extra step of reinstalling to play).
1) Windows Search puts all my applications just a few keystrokes away, regardless of shortcuts.

2) It's not about abstaining from the game, its about playing a healthy amount. Reinstalling the game daily isn't going to teach you not to play all night, it's going teach you to not uninstall the game every night after your session.

and killing a process in the task manager is easy enough for anyone who games enough to see this as a problem...

I can edit my hosts file to point facebook.com to 127.0.0.1 in general, but it's even more trivial to get rid of it when i do want to check facebook (not that i crave it that much anyway)....

this works well in the real world and even more so when parents do it to teach restraints to children, but anyone who has been in this position of wanting to reduce/prevent/limit their usage of a specific software (like focus apps) is usually (and i mean in most cases) savvy enough with their use of a computer to get rid of it in the first instance... I think most would agree there, and so it only leaves the ones that do not know how to do it in the first place.

I can believe there are a few for whom this is valuable and useful. but not enough of them that i can warrant a paid for service.

I created a Chrome plugin to block all Hacker News comments and just provide a list of the links because I found myself wasting too much time reading Hacker News comments but I still wanted to be up to date on the actual news part.

All it did was make me waste more time because I got into the unconscious habit of opening Hacker News in incognito mode to avoid the plugin.

I'm not a huge gamer really, but I've also found it difficult to turn off games once I get involved. I've spent many many hours / early mornings playing Hearthstone, which is engineered to be addictive.

I eventually cut free by destroying my collection in game and locking myself out by resetting my credentials to a random email and password. The cost of starting completely from scratch has kept me from returning.

Some people have willpower. Some people need to consciously avoid things that they find unhealthy for themselves.

If you have willpower to tone down your usage, you don't need something like focus apps. And if you need to avoid things unhealthy for yourself, then a mere barrier will not cut it.
I think there may be varying degrees of willpower such that a barrier can help and avoidance is itself a barrier.
Agreed. Waste of money. Ridiculous.
It depends on the app. I use SelfControl.app for Mac, which blocks a list of websites for a duration of time. It resolves the IP addresses of all blocked domains, then generates a set of ipfw rules and applies them to all interfaces.

You can disable the block, but it’s a huge PITA especially if you just want to disable one site. It’s a minimum five minute effort to unblock a site; for me that’s plenty sufficient motivation to get back to work.

Any insight on how this will scale into a sustainable business? I'm an indie app developer and find that most of my ideas for new products die in ideation phase because they seem financially unsustainable. Maybe I'm thinking about it wrong?

Temper sounds like a feature in a larger app, not a business. Whats the next step?

We could expand to mobile which includes a large number of gamers. We also think there's more to be done around the area of healthy gaming in general (eg. avoiding toxic games).

Another thing is we've had some people ask us for tools around Facebook/Netflix/YouTube, but we're focused on gaming for now.

I'm starting to see the vision a bit. A holistic solution to creating a healthy balance of things like Facebook and video games could become a sustainable business. Headspace might be a good company to learn some ideas from. They seem to be doing really well selling "peace of mind".

Good luck!

How in the world is this a potential business? Has YC gone down the drain?

Dear lord. This is a clock for christ's sake.

An app that only sent the word "yo" was once valued at $10 million and hyped as a revolutionary new paradigm in communication.

I can't even be surprised anymore at the things people will throw money at.

I was stoked for a sec. I'm a gamer who overuses video games, for sure, and especially late at night.

But it'd need to cope with my (evolving) Steam library. So far in 2018 I've spent meaningful late-night play time on half a dozen different games. Mostly single-player. Another year might see another pattern.

I look forward to hearing about your progress. Good luck!