summary: what the gap fill report measures, the fill rate to read, when to fade the gap vs trade the continuation, all 7 variants, a worked example, and how to set your chart up so your numbers match.
a gap is the space between where a session opens and where the session before it closed. the gap fill report answers one question: how often does price come back to that prior close before the session ends?
what it measures
gap up = price opened above the prior close. gap down = price opened below it. a fill means price traded back to that prior close — the fill level.
the report is asking: how often does price return to the previous session's close after a gap? you can also set it to a half gap, where the target is 50% of the way back instead of the full close.
the number that matters
the stat to read is the fill rate — how often the gap fills over your lookback.
edgeful's rule of thumb is ~65%. above that, fading the gap back toward the fill is a data-backed edge. below it — especially on the "high probability not fill" reads — the gap tends to run, and fading it is fighting the data.
so read the headline fill rate first, and check it against the 65% line before you decide whether you're fading the gap or trading the continuation.
the 7 variants — at a glance
standard gives you the headline fill rate. the 6 subreports slice that same data — by gap size, weekday, what happens after the fill, how far price ran first, when it filled, and the prior day's direction.
variant | what it measures | use this when you want to know |
standard | how often the gap fills back to the prior close, split by gap up / gap down | "how often does this gap fill — and should I fade or follow it?" |
by size | fill rate split by the size of the gap | "does a gap this big fill as often as the headline number?" |
by weekday | fill rate by day of week (Monday–Friday) | "does the gap fill more reliably on certain days?" |
by close | of the gaps that filled, how often the day then closed green vs red | "once the gap fills, does it stay filled into the close?" |
by spike | how far price spiked away from the gap before filling (average and max) | "how much heat is normal before the fill — where's my stop?" |
by fill time | of the gaps that filled, how many filled before vs after a chosen time | "when in the day do these gaps usually fill?" |
by prev candle | fill rate filtered by whether the previous day closed green or red | "does the prior day's direction change the fill odds?" |
the subreports
by size. the single most important slice to check before any fade. small gaps fill far more often than large ones, so the headline number can hide a much lower fill rate on the gap size you're actually looking at. find your gap's size bucket and read its fill rate — but check the sample count, since large-gap buckets are often thin.
by weekday. the same fill rate broken out Monday through Friday. some instruments fill more reliably on certain days, especially around regular weekly flows. weekdays are always determined by ET.
by close. a gap can fill and then keep right on going. this takes only the gaps that filled and shows how often the day then closed green vs red — so you know whether the fill tends to hold into the close or whether price often reverses back through it.
by spike. on the days the gap filled, how far did price first run away from the gap before coming back to fill it? it shows the average and max spike, up and down. use it to place stops: a move against your fade that's still within the average spike is normal noise; once price exceeds it, the fill is far less likely and it's time to be concerned.
by fill time. of the gaps that filled, how many filled before vs. after a time you choose (default 12:30 pm ET). most gaps that fill do so early, so this tells you your highest-probability window — and when to stop waiting on a fill that probably isn't coming.
by prev candle. the standard fill rate, filtered by whether the previous day closed green (closed above its open) or red. it shows whether the prior day's momentum changes how often the gap fills.
customizing the report
fill percentage. how much of the gap has to close to count as a fill. 100% is a full fill back to the prior close; set it lower (e.g. 50%) to study half-gap fills, which fill more often.
gap size. filter to a specific gap-size band so you're not mixing tiny gaps in with large ones.
session. the session you pick sets both the open and the "previous close" the gap is measured from — a NY-session gap and a London-session gap are different gaps on the same day.
a couple of subreports add their own settings: by fill time has the time threshold (before/after), and by spike has a dollar-vs-percent measurement and an exclude-outliers toggle.
gap-fill levels and bias
the gap-fill level is your exit target — where you aim to exit the trade. the bias tells you the session's likely direction based on historical data.
these work independently. you can aim for a gap-fill target moving downward even if the overall session bias is bullish. they measure different things.
understanding the status bar
the gap-fill status bar is based on 6 months of historical price behavior in similar scenarios (same gap size + same weekday). if the bar signals bearish, it means prices have typically moved lower under these conditions — regardless of what the prior-day close looks like.
screener indicators explained
gap up and bearish: market opens higher than yesterday's close, but historically tends to fill the gap and move down.
gap up and bullish (high probability not fill): market opens higher and historically continues higher — unlikely to fill.
these labels give you quick probability snapshots for decision-making.
what to do with it
the fill rate points you to one of two trades.
high fill rate (above ~65%): fade the gap. trade back toward the prior close. enter near the open, target the fill level, and invalidate if price pushes further into the gap instead of closing it — a gap that keeps extending is telling you it's a gap-and-go, not a fill.
low fill rate / "high probability not fill": trade the continuation. the data says price is more likely to keep going in the gap direction than to come back. don't fade it. trade with the gap, and use the prior close as your invalidation instead of your target.
here's the decision in one place:
bias — fill vs continuation, set by the fill rate against the 65% line
entry — near the open, or on a small confirmation candle
target — the fill level on a fade; an extension in the gap direction on a continuation
invalidation — price extending away from the fill on a fade; a close back through the prior close on a continuation
before you commit, check by-size — a gap this size might fill far less often than the headline number — and by-spike, so a normal move against you doesn't shake you out of a good fade.
worked example
say you're trading NQ in the NY session and price gaps up. numbers below are illustrative — read your own off the live report.
1. gap fill standard shows the gap has filled ~72% of the time over the last 6 months. that's above your edge line, so the lean is to fade — trade back down toward the prior close.
2. you check by-size to make sure a gap this big fills as often as the headline. if the right size bucket still shows ~70% on a healthy sample, you're good. if it drops to 45% on 9 occurrences, you stand down.
3. you glance at by-spike to see how far price typically runs before filling, so you can set a stop that survives the normal heat.
4. plan: short near the open, target the fill level (the prior NY close), stop above the average spike / session high. if price pushes higher into the gap and closes above your stop, the fade is wrong and you're out — it's a gap-and-go.
you're not predicting the fill. you're trading the side the data has favored, with a defined point where you're proven wrong.
set it up right
this is where most "my numbers are wrong" problems come from. get it right before anything else.
set your TradingView timezone to EST. click the clock at the bottom right of your chart and select New York. the gap is measured against the previous session close on EST. if your chart is on local or exchange time, you're measuring a different open and a different close.
then match two more things:
the session. the session you pick sets both the open and the "previous close" the gap is measured from. a NY-session gap and a London-session gap are different gaps on the same day.
the contract. a futures rollover leaves a gap that isn't a real overnight gap — make sure a roll isn't sitting inside your lookback.
fix the timezone first. most discrepancies disappear right there.
why doesn't my number match?
if what you measured doesn't match the report, you're almost never looking at a bug — you're looking at two charts set up differently. work through these in order:
timezone — is TradingView set to EST?
session — same session? it sets both the open and the prior close the gap measures from
contract / rollover — same contract, and no rollover gap inside your range?
full gap vs half gap — is the target the full prior close, or 50% of the gap? they fill at different rates
lookback period — same date range? a 6-month number and a 3-month number aren't the same sample
if all 5 line up and it still disagrees, that's worth reporting — reach out through the chat bubble with the ticker, session, and what you measured vs what edgeful showed.
full walkthrough: why doesn't my number match the report?
limits — when not to lean on it
the fill rate is only as good as the sample behind it.
big gaps and rare gap sizes don't happen often, so their buckets in by-size can be thin — a 70% fill rate over 10 gaps isn't the same as 70% over 150. check the count before you lean on it.
gaps around major news or events behave differently than ordinary overnight gaps. and a fill rate that's held across 1 year, 6 months, and 3 months is far more reliable than one that only shows up on the long lookback. use the date range to confirm the edge is still there.
when the sample's thin or the gap is event-driven, that's the time to stand down rather than force the fade.