The short answer: Comparing two items at a time is cognitively easier, produces more consistent results, and often reveals priorities you wouldn't have spotted when looking at your whole list. The math backs this up, and so does how human decision-making actually works.
Let me walk you through why this is true and what it means for how you organize your work.
The Drag-and-Drop Problem
Almost every task app works the same way: you look at your list, then drag items up or down to put them in order. It seems intuitive. It's also asking you to do something your brain isn't particularly good at.
Here's the problem. When you drag a task from position 8 to position 3, you're implicitly making a claim about how it compares to everything in between—and everything above and below. You're not just saying "this task is third most important." You're saying it's more important than tasks 4 through 8, less important than tasks 1 and 2, and you've accounted for all the relationships between items you've already placed.
For a list of 10 items, there are 45 different pairs you'd need to consider to fully rank everything. For 20 items, that's 190 pairs. For 50 items—a not-unreasonable backlog—it's 1,225 pairs.
When you drag and drop, you're not consciously evaluating all those relationships. You're making a rough guess, usually based on how important the task feels in the moment rather than how it actually compares to the others. The result is a ranking that looks orderly but is often inconsistent: task A above B, B above C, but if you actually compared A to C directly, you might put C higher.
The Alternative: Compare Two at a Time
Pairwise comparison takes a different approach. Instead of asking "where does this task belong in the order?", it asks "which of these two tasks is more important?"
This question is dramatically easier to answer. You're not juggling your entire list in your head—you're looking at exactly two items and making a single binary choice. The cognitive load drops by an order of magnitude.
But here's the interesting part: by making enough of these simple comparisons, you can reconstruct a full ranking. And that ranking will often be more accurate than what you'd get from trying to order everything directly.
How the Ranking Gets Built
The math behind this uses the same algorithms that power competitive game rankings, like the systems used in chess or online multiplayer games. Each comparison is treated like a "match" between two tasks. Win, and the task moves up in the ranking. Lose, and it moves down. The system tracks not just wins and losses but also the strength of the opponents—beating a high-ranked task moves you up more than beating a low-ranked one.
After enough comparisons, the rankings stabilize. Tasks that consistently win against other tasks end up at the top. Tasks that consistently lose end up at the bottom. The middle sorts itself out based on indirect evidence: if A beats B, and B beats C, the system infers that A probably beats C too, even if you've never compared them directly.
You don't need to compare every possible pair. Somewhere around N to 2N comparisons (where N is the number of items) is usually enough to get a reliable ranking. For a list of 20 tasks, that's 20-40 comparisons—way fewer than the 190 pairs that exist, but still enough to capture the important relationships.
Why Humans Are Good at Binary Comparisons
There's something almost instinctive about comparing two things. We do it naturally: "Would you rather have X or Y?" "Is this better or worse than that?" The comparison format maps cleanly onto how we actually experience preferences.
Researchers in psychology and behavioral economics have studied this extensively. When people are asked to assign numerical ratings to items independently (rate this task from 1-10), their responses are noisy and inconsistent. But when asked to compare items in pairs, the responses become much more stable and predictive of actual behavior.
Part of this is about attention. When you're rating a single task in isolation, you don't have a reference point. A 7 could mean almost anything. But when you're comparing two tasks directly, each one serves as context for the other. The comparison grounds your judgment in something concrete.
Part of it is also about forcing trade-offs. When you rate items independently, you can give them all high scores and avoid hard choices. When you compare pairs, ties aren't allowed. You have to decide, which means you have to think about what actually matters.
The Consistency Problem
One of the biggest issues with manual ranking is inconsistency. You put something at position 4 on Monday, but if you ranked the same list on Tuesday without looking at Monday's order, you might put it at position 7. Both placements felt "right" at the time, but they can't both be accurate.
Pairwise comparison handles this differently. Because you're making many individual comparisons rather than one big ordering decision, the random variation tends to cancel out. If you compare A and B on Monday and say A wins, but compare them again on Wednesday and say B wins, the system accounts for both data points. Over time, the consistent signal emerges from the noise.
This matters especially for tasks that are close in importance. The difference between your 5th and 6th priority might be tiny, and any single judgment could go either way. But after comparing each of them against other items—and against each other—the aggregate picture becomes clearer than any single snapshot could be.
What You Learn Along the Way
There's a side benefit to pairwise comparison that isn't immediately obvious: the comparisons themselves are informative, independent of the ranking they produce.
Every time you look at two tasks side-by-side and decide which matters more, you're articulating something about your values. You're noticing which tasks you always pick first and which ones you consistently deprioritize. Patterns emerge that you might not have seen if you were just dragging boxes around.
Sometimes a comparison makes you realize a task doesn't actually belong on your list. You keep losing against everything, and at some point you think, "Wait, why am I even tracking this? It's clearly not important." The comparison process surfaces the clutter.
Sometimes a comparison reveals hidden urgency. A task that had been sitting quietly in your backlog suddenly wins against everything you put it up against. The comparison forces you to reckon with the fact that you've been avoiding something important.
In both cases, the insight comes from the comparison, not from the ranking. The ranking is just a useful byproduct.
When Drag-and-Drop Still Makes Sense
I'm not arguing that pairwise comparison is always better. For very short lists—maybe five items or fewer—the overhead of doing comparisons isn't worth it. Just look at all five things and pick a rough order. Your intuition is probably fine.
Direct ordering also works well when you have clear, external criteria. If you're sorting tasks by deadline, you don't need to compare—just put the earliest deadline first. If you're sorting by estimated duration, same thing. Pairwise comparison shines when the comparison criteria are fuzzy, subjective, or multi-dimensional.
And sometimes you already know what's most important. If the answer is obvious, you don't need a system to discover it. The tool is for the times when the answer isn't obvious—when you're staring at a list of things that all feel equally pressing and you don't know where to start.
How I Use This
I built Darwin around pairwise comparison because I kept running into the same problem: too many tasks, no clear sense of what mattered most, and the act of trying to prioritize manually was making the overwhelm worse instead of better.
The core interaction is simple. I see two tasks. I swipe toward the one that matters more right now. The app handles the ranking math. After a handful of swipes, I have a clear next step. After more swipes, I have a full priority order for my day or week.
What I find most useful isn't the final ranking—it's the clarity that comes from the comparison process itself. Something about seeing two tasks next to each other, forcing myself to pick, cuts through the fog that accumulates when I'm just looking at a list and feeling overwhelmed.
Your mileage may vary. The underlying principle doesn't require an app—you can do it with index cards, or a spreadsheet, or a randomized list you work through two items at a time. The mechanism is less important than the insight: comparing two things is easier than ordering many things, and the results are often better.
The Broader Point
We tend to assume that the most direct approach to a problem is the best one. Want to rank your tasks? Just put them in order. Want to decide what to do? Just pick the most important thing.
But sometimes the indirect approach works better. Ranking through comparison is one example. You never explicitly decide the full order—you just make a series of small, easy decisions, and the order emerges. The final ranking is a consequence, not a choice.
This is how a lot of complex problems get solved: not by attacking them head-on, but by decomposing them into smaller pieces that are easier to handle. Pairwise comparison is just decomposition applied to prioritization. And for a lot of us, it's the key to finally getting our task lists under control.
Darwin uses pairwise comparison to help you rank your tasks. Swipe through pairs and watch your priorities emerge.
Try Darwin for iPhoneFrequently Asked Questions
How is this different from the Bradley-Terry model or Elo ratings?
It's related. Pairwise comparison for ranking can use various underlying models—Elo (from chess), Bradley-Terry (from psychometrics), TrueSkill (from Xbox), and others. They all share the core insight: you can reconstruct a global ranking from a set of pairwise outcomes. The specific math differs, but the principle is the same.
What if I have inconsistent preferences? Like, I prefer A to B, B to C, but C to A?
This happens, and it's called an intransitive preference. It usually means those three items are close enough in importance that your preference depends on context or mood. Good comparison-based systems handle this gracefully—they'll put all three near each other in the ranking and may flag them as needing more comparisons to stabilize. The key insight is that they're probably all approximately the same priority, so it doesn't matter much which one ends up slightly higher.
Doesn't comparing everything take forever?
No, for two reasons. First, you don't need to compare every possible pair—smart selection of which pairs to compare gives you most of the information with a fraction of the comparisons. Second, each comparison is fast. Deciding between two things takes a second or two. Making 20-30 quick decisions is less time than staring at a full list trying to order everything at once.
Can I use this for things other than tasks?
Absolutely. Pairwise comparison works for any ranking problem: prioritizing features for a product, picking between job candidates, deciding what to work on next in a creative project, even choosing what to watch or read. Anywhere you have a set of options and need to put them in order, comparing two at a time is a viable approach.
What if new tasks come in and mess up my ranking?
New items get compared against existing ones the same way. A few comparisons are enough to place a new task in roughly the right spot. The existing ranking doesn't get disrupted—the new item just slots into its appropriate position based on how it compares to what's already been ranked.