How InDecision Scores a 9/10 Conviction vs a 5/10

A 9/10 conviction call and a 5/10 conviction call can be looking at the exact same chart. Same asset, same day, same candle. The difference is not what the eye sees. It is what the framework counts.

Most traders treat conviction as a feeling — a gut sense that a setup looks "clean" or "obvious." That approach fails silently, because a gut feeling has no audit trail. When it works, you can't repeat it. When it fails, you can't diagnose it.

InDecision does not score conviction on a feeling. It scores conviction on five weighted factors plus one override layer, run through the same math every single time. A 9/10 is not a stronger opinion than a 5/10. It is a higher weighted sum, produced by a specific combination of signals stacking in the same direction. Understanding how that number gets built is the difference between trusting a score and just hoping it's right.

The Five Factors and What They're Actually Measuring

Every InDecision score is a weighted sum of five components: Daily Pattern Analysis (30%), Volume Analysis (25%), Timeframe Alignment (20%), Technical Confluence (15%), and Market Timing (10%). Each factor produces its own sub-score before the weights get applied. Nothing is eyeballed.

Daily Pattern Analysis carries the most weight because daily structure is the least noisy timeframe available. It is measuring whether price behavior over the last several sessions matches a known, statistically repeatable pattern — not whether the chart "looks bullish" to a human observer.

Volume Analysis is where the framework separates real conviction from retail noise. The threshold that matters here is 4.2x average volume. Below that, a breakout is a suggestion. At or above it, the move has participation behind it, not just price movement.

Timeframe Alignment checks whether the signal on the daily agrees with the signal on the 4-hour and 1-hour. A pattern that only exists on one timeframe is a coincidence. A pattern that holds across three timeframes is structure.

Technical Confluence measures how many independent technical signals — support/resistance, moving average relationships, momentum indicators — are pointing the same direction at the same time. One indicator agreeing with itself is not confluence. Three unrelated methods agreeing is.

Market Timing is the smallest weight, but it is not decorative. It accounts for the 8-hour funding reset cycle and where the current window sits relative to it, since positioning pressure changes measurably around each reset.

A 5/10 typically means two or three of these factors are contributing, and the rest are flat or contradicting. A 9/10 means four or five of them are stacking in the same direction, with volume and pattern doing the heavy lifting they're weighted to do.

Why a 9 Isn't Just "More of the Same" as a 5

This is the part most people get wrong: they assume a 9/10 is a 5/10 with slightly better numbers across the board. It isn't. The weighting is nonlinear in effect, because the two heaviest factors — Daily Pattern Analysis and Volume Analysis — make up 55% of the total score between them. A setup can't reach 9/10 territory without both of those firing near their maximum.

Picture a setup with strong Technical Confluence and good Market Timing but mediocre volume — say, 1.8x average instead of 4.2x. That setup might score a 5 or 6. The confluence and timing are real, but the framework is structurally incapable of rating it a 9, because the two largest weights are underperforming. No amount of strength in the 15%-and-10% factors can compensate for weakness in the 30%-and-25% factors. That's not a rounding quirk. It's the point of the weighting.

This is also why InDecision's conviction bands aren't just labels — they're accuracy tiers with real separation. High conviction (80%+) calls hit 91.2% accuracy. Medium conviction (60-79%) calls hit 78.4%. Below 60%, the framework doesn't publish a directional call at all. It abstains. The gap between 91.2% and 78.4% isn't noise — it reflects the fact that a high-conviction score requires the two dominant factors to be doing real work, not just a majority of factors leaning the same way.

The Risk Context Override

Here is the layer that doesn't show up in the weighted math but can override everything above it: Risk Context.

Risk Context isn't a sixth factor competing for a percentage of the score. It's an implicit override that can suppress a high raw score when the broader risk environment doesn't support it — elevated macro uncertainty, thin liquidity conditions, or a setup occurring immediately around a funding reset window where positioning is unusually volatile.

This is the mechanism that prevents a technically perfect 9/10 from getting published as a 9/10 during conditions where that number would be misleading. A setup can nail Daily Pattern, clear the 4.2x volume threshold, and align across timeframes — and still get pulled down a band because the Risk Context layer flags the environment as unstable.

This is also the most common reason a trader looks at two structurally similar setups from different weeks and sees different conviction scores. The five weighted factors might be nearly identical. The risk environment around them was not.

Where the ABSTAIN Discipline Comes In

The most important number in this entire system might not be 82.5%. It might be the fact that below a 60% weighted score, InDecision does not round up to a soft call. It abstains.

This matters because the alternative — publishing every setup with some directional lean, just with lower confidence attached — is how most retail-facing analysis quietly destroys its own track record. A string of low-conviction 5/10 calls that hit slightly better than a coin flip still get counted in the same bucket as the 9/10s, and the aggregate accuracy number becomes meaningless. It stops measuring skill and starts measuring how many calls got made.

By abstaining below the threshold, the framework keeps every published band honest. The 91.2% on high conviction and 78.4% on medium conviction are numbers earned by setups that cleared a bar — not numbers diluted by forced calls on days when the market simply wasn't offering anything with real structure behind it. A 5/10 does get scored internally. It just doesn't get published as a signal, because a 5/10 sits inside a band the framework has already tested and found to be closer to a coin flip than a call worth acting on.

What This Means in Practice

A 9/10 and a 5/10 are not two points on the same confidence spectrum you can interpolate between. They represent structurally different setups: one where the two heaviest-weighted factors, Daily Pattern and Volume, are both firing at or near their maximum, aligned across timeframes, with a supportive Risk Context — and one where the total weighted evidence is present but thin, missing confirmation from the factors that matter most.

Understanding this changes how you should read any conviction score, not just InDecision's. Ask what's actually driving the number. A high score built on strong pattern and volume behaves differently than a high score built on confluence and timing alone — even if the framework in front of you doesn't break it down. The weighting is the model. The number is just what falls out of it.

Weekly InDecision signals include the full five-factor conviction breakdown for every call — pattern, volume, timeframe alignment, confluence, and timing, band by band. Subscribe to see exactly how the framework reads the market each week.