Two indicators, same name, different signals: which is right?

You load two indicators that claim to do the same thing. Two “supertrend” tools. Two “order block” scripts. Two divergence finders. And they disagree. Different signals, different levels, sometimes the opposite conclusion on the very same candle. So which one is right, and which one is broken?

Probably neither is broken. That’s the part people miss. They’re both doing exactly what they were told. They were just told different things, because the name they share is hiding a pile of decisions that nobody made you aware of.

The name is the smallest part of the tool

Take “supertrend.” Sounds specific. It isn’t. Under that one word sits an ATR period and a multiplier, and the second you change either, the tool flips state on different candles. One author picked 10 and 3. Another picked 7 and 2 because they liked how it looked on crypto. Same name. Different animal. Neither wrote the choice on the tin.

It gets worse with the vaguer names. “Order block” isn’t a defined quantity. It’s an idea, and every coder who builds one has to pin the idea down: which candle counts, how much displacement is enough, does the block need volume behind it, does it die when price taps it or when price closes through it. Two honest developers will answer those differently and both feel right. So you get two boxes in two places, and the disagreement isn’t a bug. It’s two reasonable opinions wearing the same label.

Why this matters more than it looks

If you don’t know which decisions a tool baked in, you can’t know when it’ll mislead you. You’re trading someone else’s assumptions and calling them your own. And when two such tools fight, the instinct is to add a third to break the tie, which is how a chart quietly fills up with referees who all disagree for reasons you can’t inspect.

The fix isn’t to find the “correct” version. There usually isn’t one. The fix is to open the settings and find out what the thing actually does. What’s the period. What counts as a signal. What event flips its state. The moment you know that, the disagreement stops being a mystery and becomes information: this one’s faster, that one needs confirmation, and now you know which suits how you trade.

The version that fits you

Here’s the quiet truth underneath all of it. The best tool isn’t the one that’s objectively right. It’s the one whose hidden choices match how you actually take trades. A faster supertrend suits someone who wants in early and can stomach the noise. A slower one suits someone who only acts on confirmation. Same name, opposite fit.

So when two indicators with the same name disagree, don’t ask which is right. Ask what each one decided on your behalf, and which set of decisions you’d have made yourself. That’s the one to keep. The other was never wrong. It just belonged to a different trader.

The useful version

Two indicators with the same name can disagree because “same” usually means the label is the same, not the contract.

Compare the hidden choices:

• Does one use close while the other uses hlc3?
• Does one wait for bar close and the other update intrabar?
• Does one pull higher-timeframe data?
• Does one offset a pivot marker backwards?
• Does one use a different session, exchange, or extended-hours feed?

Those are not cosmetic differences. They change the signal.

How I debug it

Add a trace layer instead of staring at the final arrow:

plotchar(rawCondition, "raw", "r", location.top)
plotchar(barstate.isconfirmed and rawCondition, "confirmed", "c", location.top)

If the raw marker agrees but the confirmed marker does not, the difference is timing. If neither agrees, the formula is different. If both agree and the arrow still differs, the drawing logic is the problem. That turns “which one is right?” into a smaller, answerable question.