Why We Failed to Save the Ridglan Dogs

Our plan to save thousands of dogs from Ridglan Farms exploded with a can of tear gas. On April 18, 1000+ people converged on Ridglan Farms with a seemingly simple plan. Rescue the dogs from criminal abuse using every nonviolent means possible. But the police and industry responded with brutal violence, including attacking a disabled grandmother. Few of our rescuers even got near a building. Not a single dog was freed. What went wrong? There were at least four major factors in the plan’s failure: overconfidence, complexity, poor communication, and rigidity. But before I can unpack those failures, let me describe in more detail the actual plan. Our strategy was to use the “flexibility, scale, and intelligence of 1,000+ nonviolent rescuers to overcome any barriers to the mission,” i.e., to rely on teams to act independently on site and find creative means to rescue the dogs. Our plan, in short, was to have no plan—but rely on nonviolence, problem-solving, and the sheer number of the rescuers instead. The most important factors in giving our rescuers this freedom of action was to clear the site of obstacles, including both physical barriers and security or police. We had special teams who volunteered to arrive early to clear the physical obstacles within minutes—fencing and doorways, specifically—but clearing the site of security and police would require an element of surprise. For that reason, our team leads were instructed that, while the date of the action was set for Sunday, April 19, circumstances such as weather might require a change. The final briefing could happen at any time, and at any place, and team members were asked to be prepared to adapt. That is precisely what happened in the weeks before the action. Locals told us that the Dane County Sheriff’s office, which up to this point had been civil and respectful of our rights, was planning to shut down the entirety of Blue Mounds Road on April 19. Checkpoints would be set up at roads, and only those with an ID showing a local address would be allowed into the area. That would make a demonstration, much less a rescue, impossible. “We have to move before they’ve set up the checkpoints,” I told our leadership team in a meeting a week prior to the action. “That probably means going on Friday, right when people arrive.” We had envisioned this possibility from the start. I first mentioned the possibility of moving on Friday in late March. But there were two problems. The first was that moving on Friday would leave people untrained—and perhaps unprepared—for the action. While we had tried to implement online trainings in the weeks prior to the action, nothing is as good as an in-person walkthrough. And many teams would not have even met in person yet. “We are just going to have to trust our team leads,” I said. “Because it’s a choice between doing a rescue with less training, or not doing a rescue at all.” The second was that the weather was not cooperating. What was initially forecast as rain became thunderstorms and tornados in the days leading up to Friday, April 17. “I don’t think we can safely rescue puppies in this weather,” one member of the leadership team shared. On Thursday night, the team was evenly split on whether we should move the next day. “Let’s keep open the possibility of doing Friday,” I said. “But unless the weather improves, we are delaying.” On Friday, our recon teams were telling us that no police were on site, much less setting up perimeter checkpoints. Locals told us, surprisingly, that there was not even private security visible on site. One volunteer even told us he used an infrared camera to confirm there was only one human being at Ridglan Farms. Many members of our team felt that we should move forward. “This is our best opportunity to catch them off guard,” one told me. “They are going to be prepared over the weekend.” But the weather was getting worse. A flash flood warning was issued in Dane County, and a local reporter told me that his station was not even allowing him to cover our briefing due to tornado risk. I called off the Friday mission. It’s too dangerous for the dogs, I thought. We would wait until Saturday morning to reassess. The rest, of course, is history. By Saturday morning, police cruisers were present and following any cars that drove by Ridglan Farms. One pulled over at the intersection of Highway E and Blue Mounds and appeared to be preparing a checkpoint. There was even an armored personnel carrier on site. We had to move immediately if there was going to be any chance at rescuing the dogs. And even if we moved immediately, we had to adapt. As I drove towards Ridglan, I sent a message to our team leads chat. “We have to pivot. Wait for everyone to mass up and move in one large group,” I said. We instructed breach teams to hold back and arrive at the same time as the larger mass of Red/Yellow activists. But before I could organize anyone to make this pivot, I was arrested on a public road. Within minutes, tear gas canisters were flying through the air. The police refused to arrest or detain people — their main approach to stopping us on March 15 — and assaulted people instead, preventing anyone from getting anywhere near Ridglan Farms. So what were our mistakes? The first mistake was overconfidence. We assumed, based on the prior action, that the police would use the same tactics they used on March 15: detention and arrest. On April 18, they used brutality instead. I played an enormous personal role in this mistake, as I spoke to many people about how I doubted the likelihood of police violence.1 “This is a progressive county where the Sheriff has pronouns in his bio,” I said. “And there’s never been police violence at any open rescue in animal rights history.” Indeed, the main threat of violence we perceived was from agent provocateurs within our own movement. That was a mistake. And while we can’t be sure there was anything that could have been done to overcome police violence, we could have engaged in better scenario planning —consideration of different assumptions about the state of the world we were in—to reflect that possibility. When evidence started coming in that the police might shift to violence—e.g., when Sheriff Kalvin Barrett made a “crazy” video accusing us of violence in the days before the protest—we should have recognized more quickly that our assumptions about police were wrong. The second mistake was complexity. Large and uncertain challenges require simple plans, especially when they involve large numbers of people. And our plan, beneath its surface simplicity, was far too complex. Nearly a dozen independent teams, from recon to breach, were positioned all over the facility and developing very specific and detailed operational plans. When the chaos and uncertainty inevitably took over, those plans fell apart, and some of our most trusted teams were quickly neutralized. It would have been better for us to keep plans simple—and vet them with contingencies, such as an escalation in police violence. In short, we spent too much time on details that didn’t matter (e.g., how to move a haybale) rather than the high-level simple factors that were more important to mission success (e.g., how to handle a large police presence). We should change this. Simplicity should always be our mantra, especially when organizing at scale. A third mistake was poor communication. While the primary reason for moving away from Friday was the weather, a secondary factor was the surprising number of people who did not get the memo that they needed to be ready for action starting on Friday at 6 pm. There is an old mantra in communications research—that you have to repeat a new idea at least 7 times before it sinks in—and we did not do this enough. Worse yet, some people got conflicting messages, partly because we didn’t establish a single source of truth until relatively late in the action. Even our Beagle-Bot—an AI chat agent trained on the documents we…