Take small wins seriously. When I first started, I was only focused on the big picture — how accurate the final model was, how close we came to the actual election results. I didn’t always appreciate the small wins along the way, like developing a new feature or solving a difficult coding challenge. But those small victories are what keep momentum going. One moment that really stands out is when we first got our model to run without bugs — it wasn’t perfect, but it was a huge step forward. Learning to celebrate those small milestones kept me motivated for the long haul.
As part of my series about young people who are making an important social impact”, I had the pleasure of interviewing Asher Labovich.
Asher Labovich is the founder of 24cast.org, a radically transparent machine learning-driven election prediction model and first-ever campaign finance simulator. A student at Brown University with a double major in Applied Mathematics and International and Public Affairs, Asher discovered his love of politics as a Fellow in his Congressman’s Democracy Summer program, Asher’s aim is to empower informed political engagement by educating the public on the power and usefulness of prediction technology.
Thank you so much for joining us in this interview series! Before we dig in, our readers would like to get to know you a bit. Can you tell us a bit about how you grew up?
It’s great to be a part of this series! I grew up in Bethesda, MD, right outside of DC. Being right outside the center of political life definitely impacted my life and made me want to go into the political process. I’m looking forward to talking more about my focus on the political process throughout this interview!
Is there a particular book or organization that made a significant impact on you growing up? Can you share a story or explain why it resonated with you so much?
I participated in Congressman Raskin’s Democracy Summer program when I was younger, during COVID. It really cemented for me the power that a small group of passionate people have when they put their mind to changing a political reality. It’s a weird story, but I remember calling to remind constituents to fill out the Census and ending up on a call for half an hour helping this lady download Disney+ so she could watch Hamilton. It was an amazing program that has very rightfully grown far larger now than it was when I participated in it.
How do you define “Making A Difference”? Can you explain what you mean or give an example?
For me, “Making A Difference” means using data and technology to create real-world change that positively impacts people’s lives. It’s about going beyond just providing information — it’s about giving people tools they can actually use to shape outcomes in a meaningful way. We are living in an era with unprecedented potential, with technology we can barely dream of just barely beyond our current reach. It’s up to us to get there quickly and use our knowledge and skills to apply the technology as best we can. That’s what we have tried to do in 24cast.org!
Ok super. Let’s now jump to the main part of our interview. You are currently leading an organization that aims to make a social impact. Can you tell us a bit about what you and your organization are trying to change in our world today?
Absolutely! With 24cast.org, our main focus is on transforming the way elections are predicted and, in doing so, making a meaningful impact on how voters, donors, and political operatives engage with the electoral process. Our goal is to provide the most accurate and transparent election predictions possible, but beyond that, we want to empower people to act on that information.
Our campaign finance simulator is a perfect illustration of this. We’re not just telling you who’s ahead in a race — we’re showing you how every dollar raised or spent could affect that race. By giving people insights into how campaign finance influences outcomes, we’re allowing individuals to make smarter decisions about where they put their money or their time.
We also want to increase political engagement by showing voters that their contributions can actually make a difference. When we show that a small boost in donations could flip a key Senate race, we’re helping voters see the tangible impact of their actions. In this way, we’re aiming to change how people view their role in elections — not just as passive observers but as active participants who can influence outcomes in a real, data-driven way.
So, what we’re trying to change with 24cast.org is how people interact with elections. We want to demystify the prediction process, make it more accessible, and ultimately use data to drive positive social impact by encouraging informed and empowered participation in democracy.
Can you tell us the backstory about what inspired you to originally feel passionate about this cause?
My passion for this cause really goes back to the aftermath of the 2016 presidential election. I remember following the polls and predictions pretty closely, just like millions of others, and when the results came in, I was stunned by how far off many of the forecasts were. It really got me thinking about the reliability of the models that so many people were trusting. That surprise led me down a rabbit hole of research into how these predictions are made and, more importantly, how they can be improved.
As I dug deeper, I realized that these models were largely being built using techniques that had been around for a while, but there was this reluctance to embrace newer, more advanced machine learning methods. That didn’t exactly sit right with me, especially since I knew from other fields just how powerful and accurate these methods could be. I wanted to do something about it, so that’s when the idea of building a better election prediction model started to take shape.
But more than that, I became passionate about this because I saw an opportunity to make a real difference. Elections matter — they shape our laws, our policies, our society. If we could build a model that helped voters understand the dynamics of a race more clearly, and even gave them the tools to influence those dynamics, then we weren’t just predicting outcomes — we were giving people a way to get involved in the process. That’s the kind of impact I wanted to make.
So, in a way, this project is about combining my love of math and data with my belief that technology can help people better engage with the world around them — particularly in something as crucial as deciding who represents us in government.
Many of us have ideas, dreams, and passions, but never manifest them. We don’t always get up and just do it. But you did. Was there an “Aha Moment” that made you decide that you were actually going to step up and do it? What was that final trigger?
In 2022, I was in a dining hall at Brown and I saw a group of people writing down a list of past elections on a whiteboard in an attempt to create some sort of pseudo-election model. I joined them immediately and quickly realized that we’d have to move to a spreadsheet and code rather than just a whiteboard if we wanted to make any model worthy of publication. I can’t say exactly when my brain switched from “this is a cool way to spend a lunch” to “I’m going to spend the next month teaching myself to code so we can publish this”, but somewhere along the line I did just that. We put the model out a week or so before the election. The group at this time was pretty small–it was just me and two others–and this initial model ultimately became a test run for a much bigger version we wanted to make for 2024.
After the 2022 election, I wrote down a list of our model’s shortcomings and how I wanted to improve it. I then spent a considerable amount of time during the spring and summer semesters working on a preliminary model. It was just me at this time, slowly figuring out what I thought the election prediction field needed in order to build the most accurate model possible. We really ramped things up in the fall of 2023, and the team grew to its current size over the course of 2–3 months. I didn’t expect the team to grow so large, but people saw the vision and wanted to join! The last few weeks before launch were the most difficult — I worked 8–10 hour days on the model, all while studying for finals, as did many others on my team. We released it on May 4, 2024, and it’s been gaining traction ever since!
Many young people don’t know the steps to take to start a new organization. But you did. What are some of the things or steps you took to get your project started?
Honestly, I wish I could say that I had some eureka moment for how to make 24cast.org, but it was quite a gradual process. I built this model because I had fun working with modern algorithms in a way I felt no others had done, and kept at it because I found a team of people who were just as passionate as I was. So, I guess the best step I took was to get together a group of people who probably should have been working on homework, but instead saw a vision for a new type of election prediction and decided that was a better idea. I kept going because I absolutely loved standing at a whiteboard with two or three other math nerds, writing and revising formulas for hours. That definitely sounds pretty nerdy (and it was!) but I don’t think anybody can create an organization without really buying into the vision and the methods for reaching it.
Can you share the most interesting story that happened to you since you began leading your company or organization?
So, I’m from MD, and I go to Brown in Rhode Island. But my cousin lives in Texas, and he was randomly talking about 24cast.org with a few friends, and one of them apparently did a double take and started to tell my cousin how much he loved my website and how he’d been looking at it for weeks. I’d never talked to that person, and my cousin had never mentioned it before — he’d just come across it sometime and enjoyed our analysis and visualizations. That’s when I first realized we had really succeeded, when people I had never met were looking at 24cast.org.
Can you share a story about the funniest mistake you made when you were first starting? Can you tell us what lesson or take away you learned from that?
Due to some mistaken code, I accidentally set the margin of all races to be 0 (a tie). I then ran my model, and it ran before telling me that it had reached a 100% accuracy. I pretty much immediately dropped my food, thinking I had created a model that somehow perfectly predicted every election this century. I remember barely breathing until I looked through my code and found the error, quickly disappointed. I guess that taught me not to get too excited too fast, and to do plenty of tests before running with new code. I’ve kept that lesson with me as we update our code after major political incidents, like Biden dropping from the race.
None of us can be successful without some help along the way. Did you have mentors or cheerleaders who helped you to succeed? Can you tell us a story about their influence?
Absolutely! My mentor is definitely my middle school English teacher. I was a pretty nerdy kid in middle school as well, so me and my friends would sometimes hang out in her room during lunch to talk about school and the like. I remember she once gave us a problem she had to solve in her college physics class, and that kept us stumped for a little while (we got it eventually, though!)
She’s one of the nicest people I’ve ever met, and we kept close after I left for high school and then for college, meeting a few times a year to catch up. I dealt with a lot of imposter syndrome in high school, so it was really nice to have a mentor who had seen me work and knew that I had the potential I hoped I had.
Without saying specific names, can you tell us a story about a particular individual who was impacted or helped by your cause?
Absolutely! There are actually quite a few people who were impacted in very similar ways by 24cast.org, and especially our campaign finance simulator. Since we released that, I’ve got a lot of texts of screenshots of people donating to campaigns, saying they noticed it was close and that just a few donations like theirs might swap the result. I thought that was really cool — people were putting time and donations into something they might not have otherwise, and it was due to a project I worked on!
Are there three things the community/society/politicians can do to help you address the root of the problem you are trying to solve?
Yes, I’d say there is.
First, and mainly, communities should be heavily pressuring existing election prediction models to release their code and as much of their data as legally possible. It’s impossible to critique or analyze models without looking inside, and that’s our number one goal for the future of election prediction. That’s why we’ve made our model open-source, and why we’re got a detailed methodology page that explains all the major details/math of our model.
Similarly, election prediction websites must move beyond more simple statistical methods (as great as they are) and phase in machine learning models. There is a trove of data with regards to elections ready for use, and any model that refuses to utilize every bit of it will very quickly become obsolete. I think, however, that this is not something “society” can do by itself — it is merely a long-term result of an increasing amount of data and constantly improving ML models.
Moving away from election prediction, I think political parties should really be emphasizing the work of data scientists in their campaigns. Door-to-door volunteers are an incredibly important resource, but only so much as they are knocking on the right doors. At this point, any campaign with available money that chooses not to employ a data scientist is missing out on votes and donations. With local campaigns, it’s obviously harder — data scientists can be expensive, and your average state house campaign doesn’t have hundreds of thousands of dollars, especially not for just one person. But every state party should be employing as many data scientists as they can, without exception. I’m really hoping our model, and especially our campaign finance simulator, can show these political operatives the importance of leveraging data in every realm of politics.
Fantastic. Here is the main question of the interview. What are your “5 things I wish someone told me when I first started” and why? (Please share a story or example for each).
- Never work alone. Too many times while working on the original 2022 model, I stayed up until 2 or 3 am in the morning working on the code and the math because I was simply the only person who knew how to do it. I fixed that problem for 24cast.org, bringing together a team with expertise across the technology spectrum, giving me constant opportunities to bounce ideas, code, and math off of whenever it got too much to handle in my head.
- Be open to failure, but learn from it. I was really afraid that the 2022 model would perform poorly, and we’d be unable to work on another one for this upcoming election. And, to be clear, it wasn’t a great model. But instead of throwing my hands up and walking away, we analyzed our mistakes and went to work making a model that wouldn’t make the same ones. This time around, we backtested on previous elections to analyze how we would’ve done — and had we run this model on the 2022 elections, we’d have been the best election prediction model on the market. We hope to continue that trend for this election.
- You don’t need to be an expert from “day one.” I was told by multiple professors not to attempt to work on an election predictor, as I was simply so far behind existing models that it would be a waste of time and energy to bring myself up to that level. I learned not to take that advice when I just started to get to work and realized that expertise comes from doing, not just studying. I learned just as much during the process of making the building than I did in any of my math classes (though, the latter were certainly important!)
- Don’t wait for perfection — just launch. I was absolutely terrified to launch our model on May 4. We introduced a “Minimum Viable Product” — something that was missing a few small elements, but nothing strictly necessary. Our campaign finance simulator, for example, came later, after a month or so of work. If we had waited to release until our model was “perfect”, we’d still be waiting to do so. Perfection is an illusion, but progress is real.
- Take small wins seriously. When I first started, I was only focused on the big picture — how accurate the final model was, how close we came to the actual election results. I didn’t always appreciate the small wins along the way, like developing a new feature or solving a difficult coding challenge. But those small victories are what keep momentum going. One moment that really stands out is when we first got our model to run without bugs — it wasn’t perfect, but it was a huge step forward. Learning to celebrate those small milestones kept me motivated for the long haul.
If you could tell other young people one thing about why they should consider making a positive impact on our environment or society, like you, what would you tell them?
I suppose I’d say that we’re the ones living in it! Our world is only great because of the generations before us who have dedicated their lives to making ours better, so what better way to spend our time than to continue the tradition!
Is there a person in the world, or in the US with whom you would like to have a private breakfast or lunch with, and why? He or she might just see this, especially if we tag them. 🙂
That’s a great question. I think I’d choose to have lunch with Nate Cohn from the NYT. I’ve been reading his analysis for a while, and data journalism is one of my dream careers. He designs the NYT needle, which we’re working on a somewhat similar version of, and I’d love to compare the math and data behind each of our analyses. I think real-time prediction is another great place to add ML models, so I’d love to work alongside him to discover what something like that might look like!
How can our readers follow you online?
Our website updates every day: http://24cast.org
We also post twice-weekly election data insights on our Instagram account. Please follow us! Instagram: @24castbrown
This was very meaningful, thank you so much. We wish you only continued success on your great work!
Thank you so much for having me!
Young Change Makers: Why and How Asher Labovich of 24cast.org Is Helping To Change Our World was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.