How Your Rating Changes After an Arrow Race

You finish a race, win or lose, and a number ticks up or down. That number is your rating, and while it can feel a little mysterious, why did that win earn so little, why did that loss sting so much?, it follows a logic that, once you understand it, makes the whole competitive side of the game far less stressful. A rating is not a judgement handed down at random; it is an estimate, constantly refining itself.

What a rating represents

Think of your rating as the game's best current guess at how good you are, expressed as a number. It is not a score you earn like points in a level, it is an estimate that the system updates every time it gets new evidence, and each race is a piece of evidence.

Because it is an estimate, the goal of the system is accuracy: it wants your rating to settle at the number that best predicts your results. That framing explains almost everything about how it behaves. A rating is not rewarding or punishing you morally; it is adjusting its guess based on what your latest race revealed about your true level. Hold that idea and the ups and downs stop feeling personal. This rating underpins the leaderboard standing you see.

Why a win raises it

A win is evidence you are at least as good as the result suggests, so your rating goes up. But how much it goes up depends on who you beat, and that is the part worth understanding.

Beating an opponent stronger than your current rating is surprising evidence, it suggests the system underestimated you, so your rating jumps more. Beating someone weaker is expected, confirming what the system already thought, so it rises only a little. The logic is symmetrical and fair: you are rewarded most for results that reveal you are better than assumed, and barely at all for results that were a foregone conclusion.

Why one loss is not the end

The flip side reassures more than it stings once you understand it. A single loss adjusts your rating a little, but it does not undo your standing, because your rating reflects a long history of results, not one race. One loss is a single data point against many, so its weight is limited by design.

This is the antidote to tilt, the spiral where one loss rattles you into rushing and losing more. If you internalise that a lone loss barely moves your true standing, you can shrug it off and play the next race calmly. The players who climb are not the ones who never lose; they are the ones who lose without letting it derail the next several races. Losing to a weaker player does cost a bit more (it is surprising evidence in the wrong direction), which is all the more reason to give every race clean focus rather than throwing one away, a theme running through climbing the leaderboard.

Reading rating over a session

Over a single session, your rating will wobble, up after a win, down after a loss, and watching it too closely race by race is a fast route to frustration. The useful view is the shape of a session as a whole: did you finish up, down, or roughly even, and does that match how well you feel you played?

Across many sessions, the rating settles toward your true level and then moves as your actual skill moves. That is the signal to care about, a rating trending upward over weeks means you are genuinely improving, while day to day noise means very little. Read your rating the way you read your win rate and best time: as a trend over time, not a verdict on the last race. Zoom out, stay calm, and let the estimate do its job.

Understanding how your rating moves is what lets you compete without tilting, you stop over reacting to any one result and start playing the long game. It is the engine beneath your leaderboard profile, the number climbing steadily nudges upward, and the reason a calm, consistent approach beats a streaky one every time. Play the trend, not the race.