Ghost Race
The same 45-word passage every time, raced against a marker that advances at the net WPM of your best previous attempt on this exact page — a straight-line projection of a number you already produced, not a live opponent.
No recorded run yet, so there is no ghost to race. Finish this passage once and it becomes the pace you are chasing next time.
QWERTY layout assumed. Backspace corrects; uncorrected errors count against net WPM.
Ghost Race uses one fixed passage, forty-five words long, and it never changes between attempts. What changes is the marker running alongside your own progress bar: it advances at a constant rate equal to the net WPM of your best previous run on this page, translated into a projected character position for however many seconds have passed. There's no adaptive opponent reacting to how you're doing — the ghost is a number from your own history, drawn forward as a straight line.
That distinction matters more than it sounds. A real opponent can panic when you pull ahead or ease off when you're clearly beaten; a straight-line projection does neither. It holds the same pace from the first second to the last — the same discipline the Sustained Pace Drill trains directly, away from any race at all — which makes it a stable target rather than a moving one, since closing a gap against it means you actually typed faster than your own recorded best, not just faster than however the ghost happened to be feeling that second.
None of this exists on your first attempt. With no recorded run stored for this page yet, there's nothing to project forward, so the first time through is raced alone. That attempt still counts once it's finished, though — its net WPM becomes the number the next attempt has to beat, and the ghost only exists from that point on.
How to Play
Type the forty-five words as you would on any test, and watch two things at once: a highlighted bar showing your own position in the passage, and a thin marker showing where the ghost has reached, computed from your best net WPM multiplied by however many minutes have elapsed. The HUD reports the gap directly as 'ahead by' a character count, coloured one way when you're in front and another when you're behind.
The run itself doesn't stop early if the ghost gets ahead of you — there's no separate loss condition tied to the gap. You keep typing until you've finished all forty-five words yourself, exactly like an ordinary attempt, and the gap you watch grow or shrink along the way is the whole point of the format rather than a threat that ends things prematurely.
When you finish, the results screen adds two figures of its own alongside the usual net WPM, gross WPM and accuracy: Ghost, the pace you were racing, and Margin, your final gap in characters, in whichever direction it went. If this attempt's own net WPM beats what was stored before, it quietly replaces your ghost for next time — you are, in effect, racing whichever version of yourself typed this exact passage fastest, and that version keeps getting replaced by whichever version wins next.
Why Forty-Five Words Never Change
Every other detail of this format depends on the passage staying fixed. If a new random draw showed up each attempt, a slower time might just mean a harder set of words rather than a slower typist, and the whole comparison would fall apart. Keeping the text identical removes that variable entirely, so a wider or narrower gap against the ghost can only be explained by how you typed it this time, not by what you were asked to type.
A Ghost That Never Reacts to You
Because the marker is a projection rather than a recording of an actual attempt, it moves at a perfectly even rate for the whole forty-five words, regardless of where the harder words in the passage happen to fall. A human opponent would slow down on an awkward cluster the same way you might; this one won't. That can make an early stretch feel harder than it should relative to the ghost, and a later one feel easier, purely because the ghost doesn't know the passage has any difficulty at all.
The Record Is Just for This Passage
The best time this game compares you against isn't your fastest typing anywhere on the site — it's specifically your best recorded net WPM on this forty-five-word passage, filtered out from everything else stored in your browser. A strong result on a 1-Minute Typing Test or a fast Sprint Mode attempt has no bearing on the ghost here; only a completed Ghost Race run ever updates it.
What a Widening Gap Is Actually Telling You
Since net WPM already folds in an error penalty, covered in full on How WPM Is Actually Calculated, a gap that opens up isn't only about raw finger speed — a mistyped word can cost you ground against the ghost the same way hesitation can. Watching where in the passage the gap actually opens, rather than just the final number, points at whether the problem was pace or a specific error.
Frequently Asked Questions
What happens the very first time I play, before I have a recorded best?
There's nothing to project a ghost from yet, so you race the passage alone. Once that attempt finishes, its net WPM becomes the number your next attempt has to beat, and the ghost appears from then on.
Does the ghost speed up if I start beating it, the way a real opponent might?
No — it holds a constant pace worked out from your best net WPM, for the whole passage, regardless of how the race is going. It isn't reacting to your typing at all; it's a fixed line, not a rival.
Is the ghost my best time from anywhere on the site, or just from this game?
Just this page. The comparison only looks at previous Ghost Race attempts on this exact passage, so a fast result on a different test or game never changes who you're racing here.
If I fall behind early on, is the run over?
Not at all — nothing ends the run early. You type the whole passage either way, and a slow opening can still be pulled back.
How is this different from Sprint Mode, which also races my own best time?
Sprint Mode compares a single completion time once a short passage is finished. Ghost Race compares character position continuously, throughout a longer passage, so you watch the gap move in real time rather than finding out where you stood only at the very end.