summary: the data filter, the frequency filter, and the bias filters work together to narrow WIP down to setups that match your criteria — here's how to use them.
by default, WIP shows you everything that's in play across your selected session and lookback period. the data filters let you narrow that down — so instead of scanning through every setup, you see only the ones that meet your criteria.
there are 3 filters in the left sidebar:
data filter — sets a minimum and maximum historical rate
frequency filter — sets a minimum number of historical occurrences behind a card
bias filters — limits results to long, short, or neutral setups
all three apply at the same time, so you can combine them to get very specific about what you're looking for.
data filter — set your minimum and maximum rate
the data filter lets you set a floor and a ceiling on the historical rate for the setups shown on WIP.
set a minimum to cut out low-conviction setups — if you only want to see data backed by a 65%+ historical rate, set your floor there. set a maximum to exclude anything that looks too extreme, or to focus on a specific range you trust.
for example: setting 60–80% removes both the noise at the bottom and the outliers at the top — what's left is a focused list of setups with a consistently strong historical edge.
a few things to keep in mind when using this filter:
a high % on a small sample size isn't as meaningful as the same % on a large one. pair this filter with the frequency filter below rather than reading the rate on its own
setting the floor too high (e.g. 85%+) may leave you with very few setups — especially on shorter lookback periods where sample sizes are smaller
a 60–70% range is a reasonable starting point for most traders. adjust from there based on how you weigh edge vs. frequency
frequency filter — set a minimum sample size
the frequency filter sets the minimum number of comparable past sessions a card has to be built on. slide it up and any report backed by fewer occurrences than that drops off the dashboard automatically.
this is the direct fix for the sample-size problem. a strong-looking rate off a handful of past sessions and the same rate off dozens of them are not the same read, and before this filter existed you had to catch that by eye on every card. set a floor once and the thin ones stop competing for your attention.
the slider starts at 1, which shows everything. raise it and the list narrows quickly — especially on shorter historical periods, where every card has fewer occurrences to draw on. if the dashboard empties out as you raise it, that's usually a sign to lengthen your historical period rather than to drop the filter back down.
the filter applies to the bias bar as well, so the bullish / bearish / neutral split recalculates over the cards that survive it.
bias filters — long, short, or neutral
the bias filters show only setups that lean in a specific direction based on their historical data.
long — setups with a bullish historical bias. the data shows price has tended to move up in this setup historically
short — setups with a bearish historical bias. the data shows price has tended to move down
neutral — setups without a strong directional lean either way
leave it on all to see every direction at once. this filter is useful when you have a directional read on the day: if you're leaning long based on your pre-market analysis, filtering to long setups quickly surfaces the data that supports — or challenges — that bias.
using the filters together
all three filters apply simultaneously. combining them is where WIP gets powerful as a pre-market tool.
a practical example: it's 9 AM, you're leaning long on NQ for the NY session. set the bias filter to long, the data filter to a 60% minimum, and the frequency filter high enough that nothing thin survives. WIP now shows you only the bullish setups that clear your rate and have enough history behind them to be worth reading — a clean, focused list to build your bias around.
you can adjust any filter at any time without losing the others. they're independent controls that work together.