Speedrun Pace Calculator

How this is calculated

EVERYTHING IS IN SECONDS. A 1:00:00 run is 3600, a 15:00 split is 900. Timers show minutes and seconds, so this is the one bit of conversion the tool asks you to do - the printable carries a table for it. YOUR LEAD IS THE EASY PART: the PB's time at this split minus yours. Positive means ahead. Seconds are seconds, and a second saved in the first segment is worth exactly as much as one saved in the last - which is the opposite of how a percentage behaves and worth holding on to. ⚠️ THERE ARE TWO HONEST WAYS TO PROJECT A FINISH AND THEY DISAGREE. The FLAT projection assumes the rest of the run matches your PB exactly, so you keep the lead you have and nothing more. The SCALED projection assumes you carry on outperforming at the same rate, so the lead grows in proportion. Neither is wrong; they answer different questions, and the truth depends entirely on WHERE the lead came from. ⚠️ THE DISAGREEMENT IS BIGGEST EARLY, AND EARLY IS WHEN PEOPLE LOOK. It works out to your lead multiplied by (1 ÷ the fraction of the run you have done, minus 1). A quarter of the way in that is THREE TIMES your lead - a 30 second lead means the two projections sit 90 seconds apart, a spread larger than the lead itself. Half way it equals your lead. At ninety percent it is about a ninth of it, which is why late splits feel so much more informative than early ones. They are. WHICH ONE TO BELIEVE COMES DOWN TO ONE QUESTION: was the time saved REPEATABLE? A lead built from cleaner movement and better execution is likely to keep accruing, so the scaled figure is the fair one. A lead built from one lucky drop, a good spawn or a skip that either works or does not is a one-off - it will not repeat, and the flat figure is the honest one. Most real leads are a mix, and the two projections are the sensible bounds to hold in your head rather than a single number to trust. THE PACE FIGURE IS THE ACTIONABLE ONE. It compares the time you have left for your goal against what the rest of the run took in your PB, as a percentage. Under 100 means the rest has to be faster than your PB was. Over 100 means you have slack. VARIANCE, RESETS AND SUM-OF-BEST ARE NOT MODELLED. This says nothing about whether to reset, because that depends on how much time is realistically left to gain and on your own tolerance for a long run - both judgements. It also does not know your best-ever time for each individual segment, which is a better guide to what is achievable than any projection from a single split.

Common questions

Why does it give me two different finish times?

Because there are two honest ways to project a run from one split and they answer different questions. The FLAT projection assumes the rest of the run goes exactly like your PB, so you finish ahead by precisely the lead you already have. The SCALED projection assumes you carry on outperforming at the same rate, so the lead grows in proportion to the run. Neither is a mistake and neither is conservative-versus-optimistic in any principled way - which one applies depends entirely on where the lead came from, and a single blended number would just hide that question instead of asking it.

Which projection should I actually believe?

Ask whether the time you saved was repeatable. If the lead came from cleaner movement, better routing or generally tighter execution, those things carry on happening for the rest of the run and the scaled figure is the fair one. If it came from one lucky drop, a favourable spawn, or a skip that either works or does not, it is a one-off that will not repeat and the flat figure is honest. Most real leads are a mixture of the two, which is why holding both numbers as bounds is more useful than trusting either as a prediction.

Why is the spread so large early in the run?

Because the scaled projection extrapolates whatever rate you have shown so far across everything still to come, and early on that is almost the entire run. The exact size is your lead multiplied by one over the fraction of the run you have completed, minus one. A quarter of the way in that multiplier is three, so a 30 second lead produces projections 90 seconds apart - a spread three times the lead. At the halfway point the spread equals your lead, and at ninety percent it is about a ninth of it. This is the arithmetic behind something every runner already feels: an early split tells you far less than a late one.

How do I read the pace figure?

It compares the time you have left to hit your goal against what the same stretch took in your personal best, as a percentage. At 98.9% the rest of the run needs to be about one percent faster than your PB managed. Anything under 100 means the remainder has to beat your PB rather than merely match it, which is a much harder ask than a comfortable-looking lead suggests. Over 100 means you have genuine slack and can afford to play the safe strats. If it comes out negative, the goal time has already passed at this split and is gone for this run.

Why does everything have to be in seconds?

Because the formula language is numeric and a timer reading of 12:34 is two numbers rather than one. Converting is a multiply by 60 and an add - 12:34 is 754, 1:02:30 is 3750 - and the printable carries a conversion table so you are not doing it in your head mid-run. The outputs come back in seconds for the same reason, with the flat projection also given in decimal minutes as a readable companion. That companion is deliberately not formatted as hours and minutes: a 3,570 second finish is 59:30, and rounding it to the minute would display it as one hour flat - hiding exactly the sub-hour milestone the run was about. Deltas are the figures you will actually quote, and those are already in seconds in every speedrunning conversation.

Should I reset if I'm behind?

This deliberately will not tell you, because it is not an arithmetic question. It depends on how much time is realistically still available to gain in the segments ahead, on how much of your session you want to spend on one attempt, and on whether finishing a run you cannot PB has value to you - many runners find it does, because the practice is in the back half and resetting means never practising it. What the tool can give you is the honest size of the deficit and the pace the rest would need, which is the input to that decision rather than the answer.

Does this account for my best times in each segment?

No, and that is a real limitation worth knowing about. Your sum of best segments - the total if every part went as well as it ever has - is a better guide to what is genuinely achievable than any projection from a single split, because it is built from times you have actually hit rather than from an assumption about the future. This tool works from one split against one reference run, which is what you can read off a timer mid-attempt. If you track segment bests in your splits software, that is the more informative number to plan around between sessions.

Take it further with AI

Copy this into ChatGPT or Claude with your own numbers filled in. It hands over the figures this calculator worked out, so the answer is built on real arithmetic instead of a guess.

I used the Speedrun Pace Calculator at https://www.bfcbrilliance.com/tools/speedrun-pace-calculator.

What I entered:
- Your personal best, in total seconds (s): ___
- What your PB read at this split (s): ___
- What this run reads now (s): ___
- The finish time you are chasing (s): ___

What it calculated:
- Ahead of your PB by: ___
- Finish if the rest matches your PB: ___
- Finish if you keep outperforming at this rate: ___
- How far apart those two projections are: ___
- How far through the run this split is: ___

Use those figures as given — they are already worked out, so please don't recalculate or estimate your own. Help me turn them into a plan: what to buy or do, in what order, roughly what it should cost, and the mistakes people most often make with this job.

More Gaming tools