Making Mental Arithmetic Fun and Competitive With MathHeads

Hi! Can you tell us who you are and what you do?

Aman: Greetings everyone. I’m Aman Rawal and I’m joined by one of our co-creators, Ben Hollier. We’re the ones who created MathHeads—a game dedicated to making mental math fun and engaging.

With MathHeads, you can choose to play independently, with friends or challenge a stranger! The game has adjustable difficulty settings, so you can find the perfect challenge and mental exercise that suits your needs!

What are your favorite games to play?

Aman: For me, it’s Sudoku. I'm playing a lot of Sudoku on my commute to work. Just say there's time for a break and then I just come back to it. I think it's a lot of fun to play.

Ben: as for me, I have a couple of games. We have, like, daily stand-ups where we'll just kind of do a different online game. And it's generally just the New York Times games. So Connections, Wordle and Mini Crosswords are staples, and I really enjoy those.

For me, Connections is really a mixed bag. Sometimes I'll play it and I'll just be like, “Yeah, got it. Easy peasy.” And then other times, it's like, “I have no idea what this would even be. What is the connecting thread?” And then I finish it, and I feel like, “Yeah, I get it. How did I not get that?”

Why call your game “MathHeads”?

Aman: Well, I think for the name, a few of our friends were just sitting together and then we were just telling them about the game and stuff. We were just brainstorming the names and then somebody said “meth heads”. And then we're like, “What about MathHeads instead? We love math, and our heads, and we just want to focus on math.”

How did MathHeads start?

Aman: So basically, it was me and Khalid. We have another backend engineer who couldn't join us today. All of us are software engineers. We just wanted to develop something to improve our software engineering skills. And Khalid came to me one day about a game. Its premise was you’d tell the person what date it is in any century, and the other person would tell you what day of the week it would be in, like, less than 10 seconds.

And there was this memorization technique that you just memorize a few numbers and then you can do it. I liked it quite a lot, so I started learning about different techniques of doing it. I can't do that right now because I've forgotten it altogether. But that's how we started.

We liked that game quite a lot. There was this website where you could get a high score, every 10 seconds you have to answer based on what day it is. So we thought, “Why don't we make a multiplayer version of it?” It sounds fun. It's a niche, but, you know, people might like being able to play against each other, whoever solves it faster.

We started doing that and then what happened was we realized that it's actually too niche. And Khalid, who's not here, studied math in university and he used to play this mental math game called Zetamac quite a lot. It doesn't have a good interface, but you can do some configurations. Basically, all you have to do is solve as many mental math problems as you can.

We thought, “Why don't we make a version of that for multiplayer? That might be more fun.” So that's how it started. Both of us started the game back in March last year, and then Ben joined us in June. And then since then, yeah, we've just been working on the game.

What’s got people hooked on your game?

Aman: So I think probably two things. One is that how it works on mobile, it's pretty fast to load. You can save it as a mobile app on your phone, even though it's not an app. And the UI user interface, people have liked that quite a lot. It's easy to use, there's no lag. 

And the other thing that people liked a lot was the fact that the difficulties. The problems don't start very difficult right from the start. They just get harder. So the progression of difficulty makes people enjoy it quite a lot, at least, that's what we got from the feedback.

Ben: We also have a bit of a USP, which is like multiplayer, you know, head-to-head. Two different people can solve questions against each other, which I don't really think has been done much outside of our app.

Was it challenging to create that multiplayer feature?

Ben: It's definitely had some challenges recently. We had to completely redo the multiplayer backend to facilitate some of the features that we wanted to have. So we've added elo rankings. So if you play two different people who will have different elo rankings, we're calculating how much elo you gain or lose depending on that. So yeah, it's been a bit of a challenge, but it's been interesting as well.

What is the most played game mode in your game right now?

Aman: In terms of our statistics, the “play with friends” mode is the least used, so we haven't focused a lot on it. I think that was the first thing we released when we released the game, single player, classic mode, and play with friends. But I feel like what we have seen is that we don't really get a lot of new users who play with friends. 

And we've had teachers reach out to us. Right now you can only play with three other friends, but they have been using it in their classes and they want to be able to support more people, have all their students just join the link and then play. So that's something we are going to look to do in the future, but at the moment, we aren’t tackling that.

The challenges we did face were mostly for not just the play with friends mode, but multiplayer in general. iOS devices work differently than Android devices, and we found that there were a few challenges with iOS devices where if you're playing the game and suddenly switch to another app, by the time you come back, you’re already disconnected. That's not what happens in the browser. So we did have to deal with some of those challenges.

Have you ever considered an asynchronous mode for your game?

Aman: We have thought of that. We basically have a list of many things we want to do. Right now, we want to improve our multiplayer service, like improving matchmaking first, before we introduce something like this. We would want users to be matched with somebody who is quite equal in skill. So once we have that done, we’ll start working on that.

Because we have seen in terms of our multiplayer, multiplayer is our fastest-growing game mode. Like we have in terms of the number of unique users, it's our 2nd last game. But in terms of the number of games, we have a lot more games being played on multiplayer. So we have fewer users in the multiplayer mode, but we have a lot more games being played there. 

So right now we want to grow this out first before we do async. We did think of async as one of the solutions for the disconnections as well. So if it is async, then we don't have to worry about a user disconnecting and not coming back again. It's something we've thought of, but right now it’s currently in the backlog.

What are your current stats for the game?

Aman: So right now we have, on a monthly basis, around 35,000 monthly active users. In terms of unique users, the single-player mode is the most famous, which has around 10,000 users. And then we have training mode, which is fully on the front end just for users to practice. There's no ranking involved, and that’s around 7500 users. 

There's a survival mode, and it’s a game mode that is not time-based, instead, you have lives and you have to solve them before you lose all three lives. That mode has around 5000 users. And multiplayer at the moment has around 4500 users. Play with friends, again, is dead last with around 2600 users for the total number of games played.

Multiplayer, with just 4500 users, has around 37,000 games being played. And then we have single-player mode which is 36,000 games being played, then training 20,000, survival 13,000, and then play with friends around 10,000 games.

Can you tell us more about the user experience aspect of your game?

Aman: So I think Zetamac worked in an old-school way; you’d enter your answer then you’d press enter. We had to think beyond that. We want to focus more on users being able to get as high of a score as they want. So in order to do that, you shouldn't have to press try the answer and then press enter, because then you're losing a second.

And if you see our high scores, we have Mental Math World Champions. So all of them are aiming for the top ranks in hard mode. They've got a score of like 110 or 115 or something. With no more pressing enter every time you input, you are answering a problem in less than a second.

Ben: I think on Discord, we've had people be like, “How do I increase the time it takes me to type on my keyboard?” Because that is an actual limiting factor. Speaking of our pro players, we have had cases where we blocked them because we thought that they were going too fast, and we were like, “Oh, wow, people could really go fast!” So we've had to increase our limits.

Protection from cheaters and scripters

Aman: But, yeah, we did have these limits because we thought it was impossible for a human. But we have been proven wrong by a few guys on our Discord server, so we have slowly changed the limits of it. 

I remember the first time back when we didn't have the logic to detect bots. There was some guy who was using the name of a top scorer to just play a game, and at that time, you couldn't sign it, so you could use any username. And then he just took over a leaderboard, because there was a public leaderboard when we started. Anybody could just take it over because there was no protection.

So in a day, we had to go back to our database and remove all the entries because it was the same guy with so many scores, like the top one was 250 or something, which is impossible. So what ended up happening was that I spent the whole night clearing the database, and then when I woke up the next morning, the guy was there again. We had to spend a lot of time to be able to add some board protection by seeing what's humanly possible, tracking the IP addresses and then just automatically blocking them from officially ranking. 

They can still play the game, but their scores won't be saved in our database. They won't be part of the leaderboard. And then we found out there was this one guy in Tamil Nadu, India, who was pretending to be this guy, top scorer, and then he was using some sort of script to do it. So once we put his IP address in the block list, nobody was doing it anymore.

What can you tell us about community engagement?

Aman: I can't remember when we started the Discord group, but it was pretty early on after we launched the game, and a friend of ours who's not officially working on MathHeads was kind enough to just handle the Discord server. And I think we nudged our users to join Discord to get the latest updates and stuff. At the top, we put a banner saying that we just created a new community. Any new user that comes would see that banner and a button to join Discord.

So initially, we started to get some members, and then we tried to make sure to keep posting more, any new features, any, even the smallest changes that we are making in there. We made sure that there’s a way to monitor feedback, I think our friend has a channel for people to report bugs or to get some support or any help that they need. 

After that, people kept joining, and now we have around a thousand members. And anytime something goes wrong with the site, that's where we check as our users are always there providing feedback. They’re not complaining, but letting us know that, you know, something went wrong.

Connecting with their playerbase

Aman: I think a lot of people have helped us out. If we want to release something, we would just reach out to them to check our staging site before we release it to the public. Our suggestions and feedback channel is filled with stuff. So if there's something small that we can quickly release, like statistics, we deliver as fast as we can. 

I think statistics was something that took quite a while before being released, and it was because we only released it when we started seeing some feedback. People in the channel said they’d like to see some statistics about how they are doing. Then we were like, “It's actually not that difficult to do. So how about we just release it?” Yeah, I think it's just a feedback and QA and testing a lot of things, and Discord has helped us out a lot.

How did you market your game?

Aman: So we launched this in March 2023. Reddit was the reason why we thought this is something we should be focusing on, because, as I said, it started as a project for us to improve our skills, to learn a bit more about sockets, which is the technology for real-time communication that would implement multiplayer mode or play with friends. 

When we started posting on Reddit, some forums, trying to reach out to some Facebook groups and basically using all social media, Reddit was the most responsive. We did some posts on some channels. This is where we use ChatGPT to tell us about the channels where we should be posting about this sort of game. So it came up with a few suggestions and we posted them there.

We got some really good feedback over there, especially about the design of the website, the user experience and the progressive difficulty. Then we had analytics implemented on our website right from the start so that we could see how many users were coming and how many were playing. We saw that jump quite a lot in April, as I think March was around 1000 users and April was around 15,000.

At that time, we weren't spending money on ads. We were just posting on Reddit, either on a mathematics subreddit or a game subreddit. And it worked out really well. And then we thought, okay, now that we’ve posted on Reddit, and because all of us have day jobs, we can't just keep doing this all the time. Anytime you were releasing something new, we would post something. But then we were like, we need a way to be able to promote our website. That's when we tried Google Ads and Reddit ads.

We found out that Reddit ads were not useful at all. It's all bots. We spend like 100 pounds, I think because they said you spend 100 pounds, you get like 100 pounds worth of credit or something. So we spent that and then we saw that it's actually bots. Our budget got fully used, and we haven't really seen any users.

Then we tried Google Ads because they also had an offer, like, if you spend this much money, you get the same amount back. And Google Ads surprisingly worked out really well. For us, I think we were paying around, I can't remember how much we were doing, but it was a lot cheaper for us to use Google Ads versus Reddit. And then we actually were sticking with Google Ads till I think last week when we decided that we were going to stop Google Ads now and just get organic traffic from that point on and look for other sources.

Have you monetized the site?

Aman: Yeah, we did try to monetize it last year around November or December. We tried to get approval for Google AdSense to be able to show some ads, but they didn't approve our site and then they closed our account. We didn't get enough information from them as to why it happened. But then we thought, let's focus more on growing our user base, releasing more features, and then we're going to try again.

So I think we just then stopped thinking about monetization and prioritized releasing more features. And then maybe, you know, once we have more users, we'll have many sources of making money from the website.

We didn't have any infrastructure costs to run this website, though we were lucky enough to apply for AWS credits, which is where the site is managed, and it just expired this month. That's why we are thinking of how we can monetize the website now because we have to pay our infrastructure costs. 

We didn't have any infrastructure costs, we only had marketing costs. So we thought, why not just spend some money and then see if the users are increasing per month or not?

Have you had any interesting opportunities come your way after building your website?

Aman: We’ve had a couple. One was with this guy who, so this was, there's this website, a world mental calculation league or something. I can't remember the exact name, but basically they were building a simulator just like ours for world champions to compete in. They then would use our platform basically to either train or qualify and then compete and then basically one of them gets an award.

That's how our site got used by, I believe, some of the world champions in this mental math space. We decided that we are going to think about it, but then we realized that what they wanted was a lot different than what all of our users wanted. They wanted it to be more tailored for one operation, like having ten different operators solving one problem. And then the game would go until like 30 minutes or something.

And it had a lot of challenges if you were to adapt and build something like that in our current state. Our app is used quite a lot on mobile and if you try to fit a problem with ten operators on mobile, it's not going to work. You have to think of a clever way of doing it and it just wasn't worth it. So we decided to, we'd let them know how we build the real-time feature, but we decided to say we won't be able to build that tool for them or adapt methods to that specific use case.

The other opportunity we had was when a school teacher in the U.S. reached out to us. He was building a game for his class, and I think he was building a game for just teachers in the U.S. I think his main goal was a game creator software where you can create games for your students, and then give assignments in the form of a game. 

The plan was to sell it to schools where teachers could track their students’ progress, create assignments and then send the games to them using the software. This is again something that we couldn't do because it was done in Unity. And all of us didn't have any experience in Unity, so it would have meant we would have to learn a lot more. At that time MathHeads was growing so we wanted to focus more on our game, so we decided to not do that.

What’s next for your team?

Aman: One of the challenges that we have had is that we all have day jobs, so it's hard to find time to regularly meet and do it. So right now the main thing is just being able to meet, even if it needs something small, just try to sit, meet each other maybe once a week, and then discuss what we want to do.

We have a lot of things we want to build, though. We definitely want to focus on monetization now. I think what we have done is we have divided the tasks between us. Like, one person's just going to deal with the issue of monetization and how we can achieve that. The other person is going to deal with improving multiplayer. So make your own decisions. Feel free to talk to others about what you're going to do, but you are responsible for it.

So everyone has to be creative and come up with ideas to do it, and one person is going to deal with how we can get users who have created an account and come back and have them play again. So that's mostly sending emails to invite them back or when we release something new. So I think, yeah, we're going to focus on these three verticals while also trying to meet up once a week if possible.

Integrating email sign-ins

Aman: It was the easiest for us to build. We do have integration with sign-in with Facebook, though we just haven't launched it. And there was a requirement for Facebook to be able to give them a business address. They need proof with your name on it for one of the bills. But, yeah, I think a home address doesn't count. So we just added the integration and just stopped it. I think Google is good enough for now.

What advice would you like to give to those who want to follow in your footsteps?

Aman: I think it always helps if you have a friend or somebody to work with. Working by yourself is very hard, at least, it is for me. You don't set proper deadlines, you would just keep delaying it. If you're working with somebody and you're meeting up with them to do something, you know that you have nothing else to do if you have actually decided to go to a place to work. So that always helps to start a game. 

And I think you don't have to just develop everything and then release it. It's okay to release a very basic version, get that out, get some feedback, and then build on top of that, releasing small iterations rather than building old software, because that means there's less time for you to just leave it in between and then do something else. If you have something small to do, you'll do it and then just get it up.

Ben: Yeah, I definitely agree. It's good to have someone else to keep you accountable. So, you know, when you're feeling lazy, you have someone who's just like, “Why haven't you done that yet? Do it.” 

Aman: One other thing, if you do have a game out there, it's very important to collect data on how the users are using it as soon as possible. Like, Discord helps us out. But then Discord has a thousand users, compared to 30,000 users who use the game. So if a couple of people have suggested something, this doesn't mean you should build it, you should track how the users are using the website. You should know what sort of game modes are more popular before deciding what to build next.

One example of this would be that recently, some people on Discord said that it would be good to have a couple of other modes. But then we saw in our tracking that our easy and medium modes are a lot more used than hard. So a lot of people find hard mode… quite hard. There's no point working on something just because, you know, a few users want it. So it's always good to look at the actual data and how people use the website before rather than just taking suggestions.

Ben: Yeah, that's the thing about like, if you have a Discord or something, you have to be aware that these are your “power users”, and they're really invested in the game and they want features, but you have to be aware that they are your power users and they're not necessarily the kind of people that are just using it casually. You have to weigh that up. You know, they really want a very, very, very hard mode. But a guy just off the street or off of Google doesn't necessarily want a very, very hard mode. They just want, you know, to play.

Aman: I think developing to learn is another thing because Ben hates mental math, but he has had to play every time he develops something. So he's definitely improving now.

Ben: Ever since I implemented the elo system in multiplayer, like, every time I play, it's just going down and down and down. I think I'm gonna reach the lowest elo that you can ever get because I suck at it. It's weird developing MathHeads, a mental math game, and not enjoy mental math, right?

Where can we find you to learn more about you and your games?

Aman: I think if anybody has any feedback or something, we are always happy to learn, especially if it's bad feedback, just to know about the things that we can improve on. So I think the link is there on the website itself, and you can also drop into Discord.