Signal clarity and the base rate
To judge how good a signal is, look not at its hit rate but at the difference from what the market did on its own in the same period; that reference point is called the base rate.
In the daily chart above you see the order of the averages and the RSI panel below. When an analysis system looks at evidence like this and says ‘I expect a rise’, the natural question is: ‘How often is this system right?’ This lesson is about why that question is not as simple as it looks.
Say 60 percent of a system's bullish signals worked. Is that good? The answer depends on what the market did in the same period. If 70 percent of all stocks rose anyway in that period, the system's 60 percent success is actually worse than random picking. Conversely, a bullish signal that is right 40 percent of the time can be better than average while most of the market is falling. This reference point is called the ‘base rate’.
That is why the BBL analysis card says ‘signal clarity’ instead of ‘confidence’. Clarity is how much the evidence points the same way; it is not a probability or a promise of accuracy. If the evidence conflicts, clarity falls; if it all points the same way, it rises.
We evaluate our own system with this measure too and write the results without dressing them up: a signal's edge only counts as proven if there is a meaningful difference from the base rate.
How to read it
- Whenever you see a success rate, ask right away: what did the market itself do in the same period?
- Do not read signal clarity as confidence: high clarity means ‘the evidence points the same way’, not ‘it will definitely work’.
- Do not speak with certainty from few examples: a result from 20–30 trades can easily look good or bad by chance.
- Look for consistency across different periods: a method that only works in a single rising period may stop working when the market changes.
We measured
When we measured our own signals against the base rate, in a BIST-heavy period the 1-day hit rate of bullish signals came out at 17.8 percent; in the same period 28 percent of all records had risen on their own. So the sentence ‘it was right 18 percent of the time’ did not by itself prove the system was bad; the market was already falling. Relative to the base rate, no measurable directional edge showed either. Because high confidence bands did not correlate with accuracy in the same measurement, we changed the on-screen label from ‘Confidence’ to ‘Signal Clarity’.
Source: BBL signal report: 12 June – 28 July 2026, BIST-heavy, net edge relative to the base rate ·
Trap
The most common mistake is comparing high hit-rate percentages without context. In a bull market every bullish signal looks good, in a bear market every one looks bad. The second mistake is taking signal clarity for a ‘probability of winning’: an analysis with 80 percent clarity does not mean it wins 80 percent of the time.
Tip
Read the ‘Signal Clarity’ on the analysis card together with the layers on the evidence card: when clarity is low you will see evidence pointing in opposite directions; looking at which pieces conflict makes the decision clearer for you.
Glossary
- Base rate
- The share of outcomes that happen without the signal in the same period; the reference point.
- Net edge
- The difference between the signal's hit rate and the base rate.
- Signal clarity
- How much the evidence points the same way; not a promise of probability.
Mini quiz
Try it now
Compare signal clarity and the evidence in an analysis →For information only, not investment advice.