Woogles.io: The Wholesome Whereabouts of the Word-Witted
Published Jan 9, 2024
Hi! Can you tell us who you are and about your game?
Hey everyone! We’re César and Jesse and together, with several more passionate co-creators, we founded woogles.io—a place where you can enjoy playing online word games like Scrabble and the like with your friends and people all over the world. Both of us are really into Scrabble and word games in general, and we’re both proud to say that we still compete at a world-class level even today!
Currently, I’m the CTO of CodeComet and Jesse’s working as a Technical Product Manager for Noom Inc., but we’re still both active board members of woogles.io!
Woogles.io is a site that has almost everything you need to enjoy a word game. OMGWords is basically our version of Scrabble, and we have other word games you can play, such as Anagrams and License to Spell. We also have a Board Editor for people who’d like to play with a different ruleset or dictionary.
If you want to study and learn more to improve your vocabulary, spelling and overall word game skills, we have Aerolith, Random Racer and Word Tree for you to try out!
Why did you create your game?
César: There was a major inspiration for me at that time, around 2013 or 2014. I was browsing through Reddit or Hacker News (or Ycombinator) and somebody was talking about different sites for playing chess. I clicked the thread and in there, somebody else mentioned that lichess.org is a beautiful site to play in, and that it had everything a chess player needed and it was all made by one person. That piqued my interest!
I went to the site and it really did have everything you’d need in a chess site, like you have different games, analysis features, tournaments and all that! I was really shocked all the way through it, and it really was made by just one person, and it’s open source! I was so impressed and inspired that I immediately fired an email telling them that the site is one of the most impressive things I’ve seen from a solo developer.
I also told them that I was inspired heavily by their work, and wanted to create something similar for Scrabble, but of course, there’s the copyright issue to deal with and all that. And they responded just like a nice “thank you” and wished me well with my goal. And the goal just stayed in my head for quite a while until the pandemic hit.
During the pandemic, me, Jesse and a third guy named Conrad got together and decided to make something similar to what I had in mind. Yes, we had a lot of inspiration from lichess.org, but we also have our own secret sauce in woogles.io.
Did you have anything going on before the creation of woogles.io?
Jesse: Yeah, before the pandemic started, César and I had already been on a project to develop an AI. There were even times that we worked at the same job and we’d stay behind after work and like, sketch out things, but nothing too concrete.
Frankly, I was mostly preoccupied with the kind of selfish pursuit of becoming as good of a Scrabble player as I could be, and I actually made it to the World Championship twice. The first time I stepped onto the world stage, I was two games away from winning, and the second time, I was one game from winning!
Then the pandemic happened. That meant no face-to-face tournaments and competitions for a while. And as much as I wanted to compete, the available websites to play my game of choice out there were the same ones that were created 15 years ago when I’d started out as a teenager.
The UI hadn’t changed, and there were lots of vulnerabilities and other security and technical issues… they were all just not built for high-level word game competition. We were just tossing ideas during the pandemic and then it started to get an amount of seriousness because our friends got interested. And that's when we decided to actually raise money and try to turn this into a reality.
César: And it was right around the pandemic as well that we considered the AI project we were making actually had already taken shape and was pretty good. We wanted to show the world that our AI is about as good as the current existing one on the market, but we were at a crossroads.
Do we actually want to go forward with the AI project or create a way for everyone, including both of us, to play the games we both love?
And the word games won, so we decided to create woogles.io instead and put the AI on ice for a bit. But lately, I’ve been working a bit more on the AI and Jesse already came up with some really good values for different parameters that are superior to that of Quackle—which is the existing AI for Scrabble right now.
On the topic of AI, how hard is it to build one for a word game like Scrabble?
Jesse: It's extreme. There's no other way to put it just because of the number of permutations. When I draw tiles out of a tile bag, there are 100 tiles, always the same tiles every game. But there are 100 factorial different ways or orders that I can draw tiles out of out of the tile bag. So to solve it deterministically is a massive challenge beyond any other game, but it is not frequently approached that way. A lot of people think Scrabble is solved simply because learning the words is such a challenge that very few humans do so at the level of a computer because a computer can just find all the plays immediately.
But there's this whole deep strategic aspect to it that is really only the surface is being scratched. Our best player, Nigel Richards, is this otherworldly kind of savant who can read different dictionaries and then play in that lexicon at a world-class level. He does stuff that none of us can understand, really!
César: Yeah, we’ve even tried to replicate that skill of his. I have a video where I replicated something that he did that was based on the move that his opponent made. We tried to apply inferences. It would be like, based on the move Nigel's opponent made, can our bot figure out what the best move that Nigel should do in return? Because Nigel, at the time, made a move that we all thought was ridiculous, but it turns out that our bot could replicate that specific move, but not the other moves that are in the same vein.
It's interesting to see and learn what he's doing differently from the rest of us high-level players. But hopefully, with the advent of machine learning, we can replicate more moves like that in the future. But at the moment, just like Jesse said, there are so many possibilities that we do a probabilistic approach where for any given position, you just do something called the Monte Carlo sampling.
Monte Carlo sampling is where you play out different random choices for your rack, for their rack, and your rack again. After that, you play the best move from those, and you use a very simple heuristic to find the best move because you have to do so many of these and it has to be fast. But that heuristic is still very underdeveloped, and it's the same heuristic that came up in the 80s, and the very first AI that came for this is called MAVEN. While it may be very underdeveloped, even 30-40 years later, we still don't have anything better that can top that old heuristic.
Actually, we do have an active Discord channel where people are talking about ideas and trying different things. I’m excited to see what comes out of the constructive exchanges there! Even if we just come up with a better evaluation heuristic, we could probably improve that Monte Carlo significantly. But as it is, it’s still really good. Like, even with that simple heuristic, it’ll beat almost any human player… well, except Nigel, I think.
When we get to a point where we’re confident it can take on the best, we’ll have it matched against Nigel, if he agrees. And that would be so awesome.
Woogles.io is a non-profit organization, why is that?
Jesse: It was for survival. Unlike other popular games that are public domain, Scrabble is copyrighted all over the world. Here in the U.S., Hasbro is the copyright holder but for the rest of the world, Mattel outbid them and owns most of the copyrights outside the U.S. While there are expirations on copyrights, they have the choice to reassert it. We’ve already dedicated a large portion of our lives to playing Scrabble already, and we want to help others get into it, learn it and uplift the whole competitive scene entirely.
But we had to do so without treading on copyright grounds. So to state our intentions, which is just to promote and elevate the game and community as a whole, we went with a non-profit structure for woogles.io. We expected the big companies to be very defensive of their copyrights when we started off, but they’ve actually been receptive to us. We’ve already met with Mattel, and we’ve had contact with Hasbro since then.
And the reason behind their receptiveness is that we have a great mission in mind. We’re not aiming to profit off of it, we have a more scholastic focus behind us. A lot of our players are schoolchildren from countries like Malaysia, Indonesia, India, and Sri Lanka… and I think they like what we’re doing.
We’re definitely in this for the long haul and we have these missions that are covered under the 501c3 category.
What do you think are the differences between creating a non-profit and a start-up?
César: I actually think there are a lot of similarities between the two. Like I've done a start-up before, and I'm doing one right now. Back in 2011, I started with YCombinator, and we were able to get it off the ground, almost ran out of money and stuff, and we were actually able to have a nice exit in 2014. We got acquired by a big company in San Francisco and since then, I've been itching to start a new one! It was a fun thing. Lots of work, stress and uncertainty, but it was rewarding and fun!
After the pandemic hit, we found another chance to create another startup and apply the lessons we learned from our experience. Then woogles.io came to life. We created a Kickstarter and raised money with mock-ups that were not like what the existing site has. It's a little bit like how companies do that where they create a mock-up and then they see there is actual interest in this idea. And right now, we're considering doing the same thing for my current startup.
Jesse: I think it’s super interesting that when you work on a non-profit org like woogle.io, there are very tangible things between us versus a profit motive. When you’re the CEO of a profit-driven start-up, you have to do pretty radical and often brutal things just to keep things afloat in terms of operating costs. But everyone is aligned to one thing and that’s to drive profit.
Of course, woogles has its own set of problems it faces in terms of operating costs, but everyone in there who’s contributing or volunteering is freely giving away their time and effort for their own reasons. Sure, everyone is up for the plan, but all of them have different goals and motivations for being there. We have the common ground of enjoying word games, but for example, there’s one who’s actually in here to create something and include it in their portfolio, or there’s one here who just wants to play with more and more people…
In a nutshell, what I think in a startup it’s like it’s transactional—you get a goal or task and you get a salary. While in a non-profit, everybody’s bonded together by something they share in common, and work together towards something, but are never super aligned and have their own distinct motivation for being there.
When did you start loving word games and words in general?
César: I was originally from Venezuela, but I moved to the U.S. when I was nine. Back in 1993, I had to learn English to adapt here, and my older brother, who’s older than me by ten years or so, already had learned English beforehand in school. So when we arrived here, he was teaching me English before, but I was still very much a neophyte in the language itself.
Then suddenly, my school started having spelling bees, and for some reason, I was just good at them. I reached the school finals, but I remember losing on the word “facial”, the one that pertains to a face. I just couldn’t figure out what the speaker was saying! I’ve never seen the word in print and I think he also mispronounced it now that I think about it.
So I ended up spelling something like “fashion”, and I lost. That was a very memorable time for me. After that, when I was around 16 or so, I found Yahoo! Literati online. And it was like, a light bulb just turned on inside of me! Sure, I’ve played the other Yahoo! games before like Hearts and Graffiti and all those games back in the early 2000s.
But there was just something about Literati that just got me hopelessly hooked. I can’t explain it. It just fills me with delight putting long words together and scoring lots of points. And that’s when I knew that I loved word games.
Jesse: I have a similar background to César, and I actually grew up out of the country around ages 5 to 13. I was five when we moved to Italy, and eight when we moved to France. And I think doing all that kind of helped me develop language learning skills and it’s still something I’m very passionate and fortunate about.
I discovered Scrabble when I was around 16 or 17 years old, and it was like learning a foreign language again. Because frankly, when you look at the words and the definitions, there’s some arcane stuff in there, I swear! It’s just a giant convolution of years of language evolution, and there’s all sorts of wild stuff to learn and see.
The thing is, in Scrabble, there are two perfectly reasonable lexicons. On one hand, we have the U.S. version, and the other is the International, which has 30,000 more words, and I’ve gravitated to the latter. There are just so many other words in there like Scottishisms, Hinglish, Singlish, which is Singaporean English, and even Mauri! And for me, it’s all like a treat. For other people, it’s a pain in the butt but for me, the International is just my dream.
What advice would you give a new player of word games like Scrabble?
César: I’d really advise to have new players learn the two-letter words. There are only around a hundred or more of them, depending on the lexicon in use, but they’re really useful in getting good scores or getting out of tricky situations.
There are a lot of weird two-letter words, like za meaning “pizza” or qi which means “Chi” (life force) in Chinese, but it’s all officially part of the English language. Even if you learn just the two-letter words extensively and the rest of the ones you know are common words, you’ll pretty much beat an average player because no one really knows much of the two-letter words aside from, well, serious players like us. With just two letters, you can make a lot of parallel plays or hit big scores by hitting the right areas.
Jesse: There are a lot of ways to become good at playing word games. Even if it deals with letters, there’s a strong math component. There’s also a spatial reasoning component. I have friends who are experts, but they really just focus on finding a nice place that locks the board creatively and in their favor.
But aside from that, I really think the best advice I can give is to just play. Come to woogles.io, play with someone, learn and have fun! Win or lose, you can look back on the moves you and your opponent made through woogles. Practice, study, have fun, everything in there, you gain experience every time. And best of all, it’s all free! No paywalls, all fun and learning, and it will never be gated by any kind of monetization.
Can you tell us how you got into development and games in general?
César: Well, back when I was a kid in Venezuela, my dad brought home an NES, the Nintendo Entertainment System, and when I saw it I was like “Whoa, what is this?!” And I immediately got hooked when I played like the Mario games on there, and that’s how I got into gaming. And then, while enjoying these games, there was a part of me that was very interested in how people made these video games. How did they make this happen?
And in the past, my older brother had a Casio calculator. He was already in high school back then and was studying for college. So I fiddled around with that Casio calculator and tried to see what I could do, and all I made were little graphics… nothing special. Then I realized that the calculator had programming, and there were instructions in there I could learn!
I distinctly remember learning about “if” statements. Nobody taught me, I just stumbled into it and figured it out. There was an arrow in there and I realized it was an if statement. If there’s an if statement then I could pretty much build anything! So with that old calculator, I started building little number-guessing games, I even remember I built something like a dragon that would catch eggs that fell from the top and you’d just have to move it around.
Then when we moved to the U.S., I got a more powerful calculator, like the TI-82 and TI-83. I remember making small games for my friends back then, and I also recall making a little tic-tac-toe game with calculators. But granted, I still didn’t know much about programming and that kind of stuff, I just liked to fiddle around with them. When I made that tic-tac-toe game, I probably just gave it a bunch of conditions, but now, I can definitely create the perfect tic-tac-toe game.
There’s also a memory of this girl who sat behind me during science class, and she’d always ask to borrow my calculator to play tic-tac-toe instead of listening to the lecture. Then one day, she showed me that she had beaten my little game, and I was bewildered and wanted to know how she did that. Then I had to go debug the thing and figure out how she won, and that’s the kind of stuff I just do.
And around when I was 14, I got a computer and finally, then, I was able to start actually learning how to program on a computer. I started off with the old stuff, like QBASIC and C++, but the rest I’d learn on my own through books, then CS classes, then through the internet. But really, I was never formally trained in CS, and I took up electrical engineering in college because I didn’t want to sit in front of a computer all day.
That’s my life. I’m a programmer by day and by night, but electrical engineering is cool too.
Jesse: I just find it hilarious that you can immediately pinpoint how old a person is or what generation they’re from depending on their gaming trajectory. I had a GameBoy when I was three and it had Doctor Mario and Super Mario in it, and also owned a big, doorstop Cube Max where I’d play educational games like Spelunx or Math Blaster.
The games alone would tell you my age range and people outside that age range wouldn’t care at the least about those games. But if you were, you’d definitely remember them. I was also in the pocket of people who didn’t have access to universal fast internet, and during my time was the age of modems that used to dial-up first before connecting to the internet.
Frankly, if I was born several years later, and if I went to college when there were games like League of Legends… I assure you, I would not have graduated or would have some kind of reckoning!
Do you have any other favorite word games aside from Scrabble?
César: I got into Wordle back when it was viral for a while, just like everyone else, but I haven’t touched it in months. Boggle is also another fun game I was into for a while. You may not believe it, but when it comes to the other word games, I’m not very good at them. Just because I’m an expert in Scrabble doesn’t mean every other word game is a piece of cake for me. I do know a ton of words, but no matter what, there will always be people who’ll be like ten times better than me in Boggle or other word games.
Jesse: It’s funny, really. I was on this quest to be like, the best as I could be in Scrabble, then when I played other word games, I felt like I was getting distracted or guilty and I should be studying for Scrabble instead. While being competitive is good, it’s important to mix in some fun in there as well, and that includes trying out other word games too.
In woogles.io, I enjoy playing some of the other word games we have in there, like Clabbers, where you can play any shuffling of the letters of a real word. There’s also Anagrams, as we’d call it, and the boards look insane! We also have bots in there that are so good they’ll crush you and you’d feel inadequate after. So yeah, that’s my guilty pleasure—play hard word games on our site, get clobbered by AI, get humbled and then end up marveling at the magic of computers all over again.
What’s your plan for the future?
César: Well, our future plans definitely involve our AI project. It’s quite solid right now and is on the path of being able to go toe-to-toe with existing AI like Quackle. So in my opinion, we now have the best AI in the world, but we want to do some more experiments and play it against existing AIs. Actually, people on YouTube are already making series like “Best of 100 series” against it and getting their butts kicked, so yeah, I think the current model is the finest I’ve seen so far.
But it’s far from finished, as we’d like to put machine learning to it, and Jesse’s the expert on that. However, on its own, it kicks butt and is already formidable in the state it is in now. Aside from that, there’s also the startup I talked about earlier.
Where can we learn more about you and your game?
César: If you’re itching to play a good word game with someone or just want to try out a word puzzle, check out our website today! For any kind of inquiry, just go to woogles.io, but you can also try reaching out to us by using cesar@woogles.io and jesse@woogles.io.
Have a game to sell?
Let’s find out if we play well together.