Psychology

What is Conviction Score?

Reviewed by Karthikeyan Anbarasan (Founder, TradeLyser) · Updated 19 Oct 2026 · Editorial policy

DefinitionA conviction score is a rating, usually 1 to 5, that you give a trade before entry for how closely it matches the best version of your setup. Logged across 100 or more trades, it shows whether your 4s and 5s really earn more per rupee risked than your 1s and 2s, for example +0.46R against +0.04R per trade in an illustrative record, before you size them differently.
Formula
Expectancy per conviction level = Sum of R-multiples of trades at that level ÷ Number of trades at that level
Example
Illustrative high-conviction level: 35 trades totalling +16.1R, so +16.1 ÷ 35 = +0.46R per trade
Illustrative · 120 trades scored before entry · average R per trade by conviction levelAll trades +0.23R+0.04RLow (1–2)34 trades+0.21RMedium (3)51 trades+0.46RHigh (4–5)35 trades
Figure 1. Illustrative record of 120 trades, each scored before entry. Low conviction (1–2): 34 trades, +0.04R per trade. Medium (3): 51 trades, +0.21R. High (4–5): 35 trades, +0.46R. All trades together average +0.23R (dashed line).

How to score conviction

A conviction score only means something if each level has a written description. Otherwise it drifts with mood: a 4 after a winning streak and a 2 after three losses for the same chart. Tie the levels to checklist items you can see before the order goes in. The score then becomes a summary of the setup’s quality rather than of your mood, and two people looking at the same chart should arrive at roughly the same number. Record it in the trade entry at the moment you place the order, alongside the setup tag, and do not change it afterwards, even if the trade turns out very differently from what you expected.

ScoreExample description (opening range breakout)
5Every checklist item met, higher-timeframe trend agrees, clean level, volume above average
4Every item met, one supporting factor missing, such as average volume
3Checklist passes, but one item is borderline, such as a level tested many times already
2Minimum criteria only; trend neutral or against
1Taken for a reason outside the plan; tag it so it can be reviewed separately

Reading the record: a worked example

In Figure 1 the levels separate cleanly. The 35 high-conviction trades made +16.1R in total, the 51 medium trades +10.71R and the 34 low trades +1.36R. Together that is +28.17R over 120 trades, or +0.23R per trade. At ₹2,000 risked per trade, the low-conviction trades earned about ₹80 each on average, before charges, and the high-conviction trades about ₹920. Once charges are taken off, the low-conviction trades may well be losing money, which is a question worth answering before deciding whether to keep taking them at all.

High (4–5), per trade
+0.46R
Medium (3), per trade
+0.21R
Low (1–2), per trade
+0.04R
All 120 trades
+0.23R

Even this tidy result needs care. Around 35 trades per level is a small sample: with 30 trades, a measured 50% win rate still has a 95% range of about ±18 points. A gap that shows up in one quarter and disappears in the next was probably noise. Look for the same order across two separate reviews before acting on it. Check two more things before you trust the split. First, are the levels spread across setups, or are all the 5s one setup? If so, you are comparing setups, not conviction. Second, how are the scores spread over time? If high scores bunch up after winning weeks, the score may be tracking confidence from recent results rather than the quality of the chart.

Using it for sizing

Clear gap that holds across two reviews
Your judgement adds something. A modest size step for high conviction, decided at a review and written into the plan, is reasonable. Keep risk per trade within your normal limits.
No clear gap
Conviction is not predicting results yet. Keep size equal across levels and tighten the written descriptions.
Low conviction does better than high
The score is picking up something other than setup quality, often excitement. Keep size equal and rewrite the criteria for a 5.
Rule of thumb
Score before entry or not at all. A conviction score added after the trade closes will match the result and teach you nothing.

Conviction score vs emotion grade

Conviction score

How good the setup is, judged against written criteria before entry.
Question: is this my best version of the setup?

Emotion grade

Your state of mind when you took the trade, such as calm, rushed or frustrated.
Question: was I in a fit state to trade it?
For F&O traders in India
Sizing by conviction is coarse in index options, because you can only trade whole lots: with a Nifty lot of 65 (check the current NSE circular), the smallest step up from one lot doubles the risk. On a small account, conviction is often better used to decide which trades to take than how big to make them. In TradeLyser, the Strategy Map tab on a trade has a confidence score and a note for each rule, which can record the same judgement.

Reference guide

ContextValueReading
Scored before entry from written criteriaUsable dataThe score reflects what you knew at the time, so it can be tested against results
Scored after the trade closedContaminatedHindsight pushes winners up and losers down, so the score just repeats the result
Size raised on high conviction with no record behind itUntested sizingWait until each level has 30 or more trades and the gap holds across two reviews

Questions

What is a conviction score in trading?

A pre-trade rating, often 1 to 5, of how well a trade matches your best version of the setup. Its value comes from comparing results by level after enough trades.

Should I trade bigger on high-conviction trades?

Only if your own record shows high-conviction trades earn clearly more per trade, across at least two review periods with 30 or more trades per level. Until then, equal size is the safer default.

How many trades do I need before conviction scores mean anything?

Aim for 30 or more at each level, so 100 or more in total for a three-level split. Fewer than that and differences between levels are easily due to chance.

Is high conviction the same as overconfidence?

No. High conviction backed by written criteria and a record of results is calibrated judgement. Overconfidence is rating trades highly without that record, or after a winning streak.

Related terms