Ranked-Choice Voting Undercuts Our Right to Choose Our Leaders

Bob Zeidman

Two years ago, New York City held its primary election for the Democratic candidate for mayor, but it took two weeks to declare the winner. Last year, in Alaska it took two weeks to declare the winners in its Senate and House elections. Alameda County, Calif., incorrectly counted ballots and declared the wrong winner for district school board director, which took two months to discover.

The reason in all three cases? Ranked-choice voting.

Touted by a growing number of authorities, ranked-choice voting forces voters to choose not just the candidate they want to win, but to rank the candidates by preference. This unbelievably stupid method of voting is a serious threat to American democracy.

The threat is even worse than the potential for fraud and mistakes by other forms of computerized voting because with ranked-choice voting, any candidate can end up winning, and the choice rests with a few computer programmers who maliciously or innocently choose an algorithm that picks a candidate that no one actually wanted to win.

Supporters of ranked-choice voting say the algorithms are easy.

Tabulating the results of a ranked-choice election is not a difficult process for modern computers,” according to the New York Times.

That’s technically true but misleading. The problem is not that the algorithm is difficult to create, it’s that the correct algorithm is impossible to create. Computer scientists can’t agree on the “correct” algorithm for one specific reason: there is no such thing. Let me give you a specific example.

In the table below, there are four candidates running for office: Amy, Bob, Charlie and Debra. There are five voters, each of whom rank the candidates. There are four methods (algorithms) for counting votes to determine a winner.

Method 1 is simple average of the votes. Using that method, Bob is the winner even though he wasn’t the first choice of any voter.

Method 2 awards victory to the candidate who got the most first place rankings and gives the election to Amy and Debra in a tie, requiring a runoff.

Method 3 gives a weighted average where a first place ranking is given more weight than a second place ranking and so on. Using this method, Debra wins the election.

Method 4, whether by intent or accidentally, only counts voter 5 and awards the win to Charlie. The winner is chosen not by the voters but by the computer algorithm.

By using ranked-choice voting, we’re putting our elections into the hands of a few elite experts and computer programmers who create computer algorithms that assign the votes to candidates and determine a winner that, in some cases, no one really wants.

Unlike simple voting that can be hand checked by low-skilled human vote counters, the messes in New York, Alaska and California demonstrated that human counters can’t replicate the computer algorithms to check whether all votes were counted and counted correctly. And even when all votes are counted correctly, it’s the algorithms that determine the winner, not the voters.

You don’t need fraud or incompetence to change the vote, you just need some ill-defined algorithm.

But if fraud in the software changed the vote count, or, more likely, a bug in the code changed the vote count, that would be extremely difficult to know. It would be difficult to differentiate a bug or a manipulation from an algorithm that gave unexpected results.

I know from experience. I’ve designed computer algorithms where the results seemed odd, and it took days or weeks to determine whether the odd results were correct and simply unexpected, or if they were incorrect.

I’ve also probably examined more code than anyone in the world, because of my involvement as a software forensics expert in over 260 lawsuits. Verifying that code is working correctly is one of the most difficult challenges in computer science. And every program has bugs.

That’s why hand recounts of computer voting are often necessary. But if the algorithms are so convoluted that a hand count won’t work, all bets are off.

Not long ago, we went through one very contentious presidential election. I’ve argued that if any fraud or mistakes were made by the voting machine software, we would know about it. It’s what my team and I do for a living.

However, I can also state unequivocally that ranked-choice voting will not only allow fraud and mistakes to go undetected, but even if the algorithms work as designed, it will take control of our elections away from the voter and give it to a small number of people.

We must not allow this to happen if we intend to maintain our Constitutional right to choose our representatives.


Bob Zeidman

Bob Zeidman

Policy Fellow

Bob Zeidman is an inventor, author, entrepreneur, philanthropist, and high-stakes poker player. He created the field of software forensics and founded Software Analysis and Forensic Engineering Corporation to develop and sell software forensics tools. He is the founder of Zeidman Consulting, an engineering consulting company that has worked on over 250 major litigations involving billions of dollars of disputed intellectual property. His cases have included ConnectU v. Facebook, on which the Oscar-winning movie The Social Network is based, and Oracle v. Google that went up to the U.S. Supreme Court. He is the inventor of the famous Silicon Valley Napkin on display at the Computer History Museum. He is also a high-stakes poker player, and his latest tech venture is Good Beat Poker, a new way to play and watch poker online.

Bob writes about politics, society, and business for national magazines. His latest book is the upcoming A Conspiracy of Dunces, the true story of how he challenged his own beliefs about voting machine hacking in the 2020 presidential election and made international news and $5 million.