Hitting Up Words and Friends With Contexto

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

Hello, I’m Nildo Junior, the creator of Contexto—a word-guessing game so hard and frustrating you’ll end up loving it. I work as a software engineer. I’m originally from Brazil, but I moved to Oslo, Norway to work and have been here for over five years. I mostly work on corporate development and assist dev teams, but I also used to write on the side, with many of my articles on Medium.

Aside from my corporate jobs, I also created Daydash—a platform where you can play and learn at the same time through games that are simple, fun yet challenging! If you’re looking for a harder word-guessing game, try Contexto. I also have other games there, so check them out too!

What’s your favorite game to play?

My favorite game to play is League of Legends. I've been playing it for much more time than I wish, actually. But I do play it fairly consistently, and it is something that bonds my family together, my whole family likes the game. So I feel like I'm a little bit addicted to it and I don't like that feeling, but, yeah, it does bring me some joy and social interactions, and that's what I mainly play.

And I have a certain feeling about being addicted to League compared to something like chess. I feel like that chess, there is a little bit of a social positivity to it. Like, no one really judges you for being addicted to chess because it's intellectual.

People being ashamed to share their addiction to a game

I feel like playing games is not something that you hide from people. You know, I even have a suggestion because I've been listening to the podcast before. And I know that in Hey, Good Game, there’s a spiel where you guys say, “We talk about the games that you secretly play in the cracks of your day.” And I have a bit of an issue with the word ‘secretly’ because people usually don't play games secretly.

From my experience, what I see from my player base in Contexto, they're not secretive at all about playing the game. Actually, they play it because they have someone to talk to about it. So they want people to know that they have been playing every day, even.

How did you end up working in Norway?

I moved here for a job five years ago. Before, I was trying to build a startup with my friends back in Brazil. After I finished my master's, our startup was trying to get the teaching of computer science and programming to schools for many reasons that did not work on a business level. So after we decided to take a break. I started looking for opportunities where I would get different experiences elsewhere.

So I wanted to move abroad, and I did a bunch of interviews with different companies, and I ended up coming for a job here for a company in Oslo. And then I stayed ever since. But it was basically me wanting some different experience, wanting to move abroad, and looking for a job that I could do and get some more experience somewhere else. But I really like the country. I think it's a very good place to be and a very good place to live.

Can you tell us more about Daydash?

I created Contexto as a side project. It was during the COVID-19 times. We were all having a lot of free time at home, I guess, and Wordle was a big thing. Then I remember discovering Wordle because everyone was playing it. So people were sharing their little Wordle squares all over Twitter and even at work. People would share in Slack and stuff, and everyone was playing it.

And with it came all the new games that I now call the Wordle genre of games. Like, it created a whole bunch of Wordle-likes, and one of them was Semantle. But you already interviewed the creator of Semantle, David Turner there. I found Semantle on Twitter through someone I followed there. I was navigating, I saw someone sharing their results for the day. And I was curious, so I clicked it and started playing it.

And I was fascinated by it because it was so difficult but so engaging at the same time. It took me so much thinking to get the correct answer that day. And I knew then, that what I felt was true—this was a very, very good game. I wish I could play it in Portuguese because in Portuguese things are easier for me. And then I looked it up and, yeah, there was no version of this in Portuguese.

Being inspired by Semantle

At that time, the code for Semantle was open source. So I clicked the link there and you’d see all the implementations for Semantle there. The only thing that you needed was to translate the interface and find a database of word vectors for your target language. Then, you could make a version of Semantle in your own language. A lot of people did that.

So there are Semantle versions in Spanish and French, I think. Actually, if you go to Semantle now, I think on the Frequently Asked Questions, you can get links for all the versions of it that exist. And then I decided that I wanted this game in Portuguese because I want people in Brazil to be able to play it because it is fun. 

We have our version of Wordle, which is called Termo. Termo just means “Word” in Portuguese. And everyone playing that was also in the wave of Wordle players. And I thought that we should have Semantle in Portuguese so we can play too. So I'm gonna do that. But as soon as I started thinking about what was I going to do. I also started thinking about all the problems that Semantle had because I was playing the game.

I thought it was very good, like the playstyle of it, but it was also very hard. And thinking about my target audience, I thought that people would not get this, and it would not appeal to people. Like, it's too much to think about, it's too much numbers, it's too much information. And especially my biggest problem with the game is the main point of information that I used to guide myself while trying to find the right word— the last column.

In Semantle, the last column was a number from one to a thousand, meaning almost the same as I did in Contexto. However, the number scaling in Semantle was very confusing, the thousandth word was what you were looking for. I kept guessing words and I got 997 and I was like, “Okay, but to a thousand, that's seven. So I'm seven away…” or if I got 500 something, okay, how far away am I still going? And that was a mental overload for me during the game that really demotivated me.

Making Contexto simpler than Semantle

And I thought this would be just so much better if the numbers were inverted. If the target word was number 1, and the ones that you're guessing were the higher ones because you don't need to do subtraction. So it's one less operation that you need to do in your head to play the game. So that was the first thing. 

But then I kept thinking that because in my work, I always think about usability. I work with great designers, but I also consider myself having a bit of designing background too. And I help with all usability things that need to be done in the projects that I've worked at. And that experience really helped me to shape up, like, okay, I think this game is great, but it really needs to be simplified. 

And by being simplified, I mean that people will not be afraid of it, because that's what I felt like—people were afraid of Semantle because it just felt like it was something they would not understand. They will say, “Oh, we use this complicated algorithm, like word2vec, to calculate the similarity with the cosine differences between all these technical terms.” 

Like, okay, we get it, but that's not important for the game. This number of the algorithm is not important to the player. The only important thing for the game to be more appealing to the player is knowing how much closer or farther away they are from the answer. 

So I'm gonna design the game in such a way that this is going to be its main mechanic. It's going to be simple. It's going to be a word and a number. Done.

Difficulties in teaching people what the game was

So even though the game is now simpler enough, sometimes people just don't get it. There are instructions at the beginning of the game, like a small text explaining things. And that was also very hard to write because I wanted to explain the game without using any technical terms for people. Like, what do people need to understand about the game? Stuff like that.

The funny thing is that I released the game in Portuguese in February of 2022. That was before ChatGPT came on and was released to people, and people started using the word AI too much to the point where we don't know what AI is anymore. We’re now just referring AI to large language models. 

But at the time, I was thinking, how do I explain the game in a simple enough way that people understand that this list of words was set by an algorithm, by a computer? It's not by a human, because that was a big problem in there, and it still is. Like, people will complain about the order of the words because it doesn't make sense. Why is this word closer to that word? Why are those two together?

So yeah, I just wanted to explain to people that this is computer-generated, it's not a human. So if words are a bit weird, it's because the data they were trained on led to this and the algorithm led to this. But a lot of people also saw that as part of the game. It's like, “Oh, can I think like a computer?” or “Can I guess what a computer is guessing?” That's also a part of the game.

But the feedback that I get from people is they really like the game. It is frustrating for many minutes, but once you get the answer, there is a feeling that is very good there, a feeling of accomplishment that really gets people to feel like, “Okay, that was nice. I was suffering for this last minute, like really frying my brain here. But it was nice to finally arrive to the answer.”

How long it takes for brain frying to feel good

So I don't know by statistics. Well, I don't follow the statistics a lot anymore, but it would indicate that the average time on the website was 20 minutes. But that can be people playing multiple games because you can do that on the website too. But yeah, I have no way of really knowing this, but I would guess that 10 to 15 minutes is probably realistic to think that someone would be playing the game for that long.

Can you tell us more about the social aspect of playing games like Contexto?

That is so interesting because sharing and competing is such a common use case for the game. People usually don't play these kinds of games alone. I mean, people play alone, but when they play together, they think they have more fun. So many people have been coming to me saying, “Oh, I played your game at my school.” or “I played your game with my coworkers.” or “I played your game with my grandma.” And I think that's amazing. 

It's like, it's creating a bond factor between people. An interesting note too, is that the game is way more played on the weekdays. And for a long time, I used to think that people were just procrastinating and playing the game. But as soon as we started asking people, that was some years ago… I think it was a year ago or more now… We did a poll with people asking like, “Where do you play the game?” They would say, “I play at school with friends.” or “I play at work with coworkers.” 

And before that, I didn't think about school, actually. I was like, “Oh, okay. Yeah, people go to school on weekdays, so it makes sense that there will be then more players on the weekdays than on the weekends.” I only thought about the work dimension of it, but it's actually played at schools. Like, people who play with their colleagues and have fun together compare how much stress they got to win. And then it's like, my personal friends even tell me that they play it in a bar. My friend was like, “One day I presented your game to people in a bar and everyone started playing.” Oh, that's so fun.

How did you start being interested in tech and mathematics?

You're throwing me back here. Way, way back. So when I was still studying before, like in the fifth grade, I guess I really wasn't that much into math. I was more into language. I liked to write stories. I remember when I was at school, maybe in the second grade, I wrote a text and then the teacher was like, “Oh, let's do a theater play from this.” And then we created a theater play presented to the whole school. And that was very interesting. 

But afterward, maybe in fifth or sixth grade, I started being better at math, mainly because my math teacher was better. She was so good at teaching that math became my preferred subject to study at school. And then at the same time, they started creating math Olympics in Brazil that were specifically for public schools because I always studied in public schools. And there I won a prize like, “Oh, you did well in this math Olympiad, so we're gonna give you some money and you can study extra math on the weekends.” And I was like, “Oh, am I getting paid to study extra math? That's amazing!” My mind then, as a teenager, just felt really great about it all. And from there I was really just interested in math. 

And then came the time when I also got a computer at home. I got a computer home fairly late. I was maybe 12 or 13 years old. As soon as I got that, I was very interested and went to like some schools to learn how to use the computer, like how to use the operating system, how to use the text editor, how to use a spreadsheet…

Eventually, in one of those schools, I started being the teacher of some courses. And then I went to a programming technical school, then I went to university for computer science. I did my master's degree, I studied for one year abroad in California, and then I built my own startup and I failed at that. And then I got the job abroad and then I created this game and that's what I do now.

But yeah, so back in the day it was really incentives from outside that got me deeper and deeper. There is one thing that I advocate for, and it is education. I wouldn't be here, like, I wouldn't have created Contexto if I didn't have a whole background in education and if I didn't have my bachelor's in computer science to understand how the word vectors work. 

The main thing responsible for word vectors used in games like Semantle, for instance, is linear algebra. A lot of people in the university even would be like, “Oh, why are we studying this whole math thing? We're never going to use this at work!” Well, you might actually have a use for it, because, you know, I started through linear algebra. With it, I could easily understand something that I, in my computer science career, had never seen. 

I did not study artificial intelligence or this kind of like text processing techniques and algorithms. I was doing network algorithms and those are different. So I didn't have experience with it, but I could understand it fairly easily because I had some background in it, I had enough knowledge to combine with the experience I had with designing at work. I had experience with the technology that I used to build it, like the JavaScript frameworks and everything from work. 

So there is really nothing that you can do that is a waste of time when it comes to acquiring knowledge. It's always good to acquire more knowledge. You never know when you're going to use it. And as my father says, it's something that no one can truly steal from you.

How did you get your first thousand players?

I think that comes a bit from the design aspect of the game, actually. One big feature that I absolutely needed on the website before being released is the ability to share your results because that's what made me discover the game in the first place. That's how I got to know so many of these daily games because I'm a Twitter user. So I was on Twitter and people were sharing these kinds of things and I thought, “Okay, where is this one from?” And I look it up. 

And that was one feature that I needed in the game because I believed that the game would be played by many people if given the chance. I remember when I saw Semantle’s, it was just text, like, oh, okay, I played Semantle and this is what I got in the proximity of in a bunch of numbers there, but no emojis. No emojis in the share result thing.

And I thought that was fine, but Wordle and the other games, they really created this genre. I think it's even almost a characteristic of the genre. What I'm calling the genre of Wordle is to have some emoji representation of your results that you're gonna share in social media or with your family and friends.

So that was one feature that I really needed on the game and that even. That even influenced the design of the game. So the process of my designing Contexto was based on wanting to simplify what I saw in Semantle. So I just wanted people to know if they were closer or farther away from the target world. I thought that that was the only information that is actually useful for this game.

How did Contexto blow up even further?

So the sharing thing was very important because that's what made people discover the game. What I did at the beginning of the game was I created it, put it online and shared it in a group of four. No, with three friends, four includes me. Like, we had a small group in a conversation app and I shared it there like, “Hey, I did this game. Can you test this?” And they picked up, they played it and they're like, “Nildo, this is very cool. Can I share it with someone?” Of course, I was like, “Yeah, of course!” And they shared with people that they know or they work with. 

And that was the only thing that I did, really. This is for the Portuguese version, by the way, the first one. And from there it took maybe a month until very, very famous YouTubers and streamers in Brazil were playing it on their live games and their videos. 

During its peak popularity in Brazil, it had, I think, 300,000 people playing on a single day. So it became very popular basically because people shared the game with others and on social media. We also found out that in our poll with users, where 90% of the users actually said that they discovered the game either through a friend recommendation or through social media. So that's really it.

On Contexto’s traffic stabilizing

We added ads on the website at the end of 2022, but Contexto’s traffic isn’t climbing anymore. That's an interesting thing. Not very familiar with everything, but in our discussions, some things stand out. In Brazil, it was very popular, like one or two months after release, and then it really declined. And now it's not played that much, at least not played daily by many people anymore.

When I released it in English, and it got popular, the behavior was completely different. Actually, it didn't get as popular as it did in Brazil, so it didn't have like a very big peak of users, but it also didn't decline as steeply as it did in Brazil. So it was declining, but it was declining very slowly. And now it has been more than a year, actually, it's almost two years that we have the game in English and we've been observing a seasonal oscillation in the game.

Even so, people were playing less in recent months, and now they're starting to play more again. And we hypothesize about these kinds of things, like, okay, what is it? Is it summer vacation ending? Yeah, it might be because people are going back to school, people are going back to work, people are going back to the social environments where they are reminded of the game that they used to play with co-workers and other students.

And also, we've gotten some feedback from teachers too, on our feedback link. So some of them use the game in the classroom, and I'm very proud of that because as I said before, the one thing that I value is education. And if something I've done is being used to educate people, that makes me very proud.

Is Contexto your full-time occupation?

Yes, it is my full-time occupation. I had a big dilemma about it because I'm a very safe person, and I don't take risks with things. So I had a good job, a steady one, and then I started monetizing the game for extra income. So I was like, yeah, this is valuable to some people, so why not get some value back? But then it started being too much work for me, working a full-time job and working on projects that I really cared about, which are Contexto and some other games. I got really exhausted at some point, and I had to do some therapy until I was able to quit my job because I was afraid to let people down. 

But at some point, the game was what I wanted to do, and it was more financially rewarding than the job that I had. So it wasn't really a difficult choice, but as a person with people-pleasing tendencies, it was a big issue for me. But now it's my full-time occupation. I’m all for developing Contexto, developing new games and bringing nice experiences for people to have fun and learn new things while doing it.

Can you tell us more about the two other games you have in Daydash?

Yes, Conexo and Letroso, and I added them over the course of the last year. Conexo is a bit older and is a version of Connections in Portuguese, or at least in the beginning, it was. It was the same thing that I did with Contexto, and I’m just following that with the other games. So Connections came out, and Connections itself is not too new, because before that, there was that BBC game. That show was “Only Connect” and I used to watch that too, but it wasn’t available in Portuguese. And people in the community of the daily games, they're like, “Ah, it's too bad that we cannot play this in Portuguese.” 

So I saw that as an opportunity to do something. But I don't like it. I like to do things in a way that I also think I am improving it. For instance, for Conexo, you can have unlimited guesses in your attempt. Because for me, as the designer of a game that I call a puzzle game, it's a puzzle, you should have all the chances you need to solve it. I think it's a bit weird that a puzzle has a lose condition, like, that you can lose in playing a puzzle. I don't think that should be a thing. 

A puzzle should have two states. It's either unsolved or solved. I don't think it makes sense that you lose at a puzzle. And for me, Conexo is a puzzle that you're trying to solve. So you shouldn't lose, you might not be able to solve it, but you shouldn't lose at it. And it's the same thing for Letroso. 

But, yeah, as long as I'm making games and I'm calling them puzzles, I will never do it in a way that people will lose the game because you don't lose a puzzle, you only solve it.

The purpose of continuing that approach to game design

Maybe it has a little bit to do with the feedback that I've been getting from Contexto. The nature of Contexto is it brings feelings to people. People get frustrated and people get mad when they cannot solve the game. And I understand. Sometimes I play it and I get these feelings and I understand people that do that. 

So in the beginning, when Contexto was super popular in Brazil, I used to have my Twitter handle in the credits of the website, and people would go follow me and send me a message. Positive messages, like, I love your game or so. Then someone said to me, “I play your game with my grandma. Every day it is something that we do together now.” And that's so heartwarming. It's so nice. It's so good. 

And I was really enjoying myself getting this positive stuff through Twitter. But two months later, no one else would give me positive feedback, and I would get a lot of hate. So when people would be mad at the game, they would look up the credits, look up my Twitter, and send swear words on my Twitter. And then I started being like I don't need this anymore. I removed my credits, and that's when the company I was working for recommended that I get therapy.

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

As a game creator, you will get a lot of feedback. And actually, you should actively look for getting feedback on your game. But the feedback that comes naturally will usually not be good. It will usually be, “Hey, this is wrong with your game.” or “This I didn't like. I wish it was different.” And then you need to know how to handle these things. But you need to know that most people who are playing your game, are actually enjoying it, but you don't get the positive feedback most of the time. 

I learned this from I think, the most important book that I have read so far, which was Donald Norman’s The Design of Everyday Things. And what I really learned from there that is very important is that good design is invisible, but bad design stands out in your eye. And I think that's also the same for the feedback that you get on your game. It's like while your game is good, you will hear nothing, but if someone thinks it's bad, they will tell you, they will seek you to express either their frustration or their wishes on a different thing. So I think it's important.

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

Go to daydash.com! You're going to find all the other games there and emails to contact us. So that's where you should go. Feedback is very important!