Episode 2:

Reasons to Reject the Illusion of a “Win Rate”

Why can you win with that number, yet still not?

Last time, under the theme of “Learning AI vs Non-learning AI,”
we explained why a future-proof logic is necessary.

And this time is the continuation of that.

So, what criteria should you use to choose a logic?

Before getting into the answer, there is something you must break first.

---

■ Does that “win rate” display really have meaning?

When you start trading, you inevitably see it.

• Win rate 80%
• Win rate 90%
• Highly accurate signals

At first glance, it looks strong.

But, here is one question.

Do you know how that win rate is calculated?

■ The true nature of simplified backtesting

Many indicators or tools display

“recent win rate”

but

the content inside is surprisingly simple.

---

■ Common evaluation methods

• After entry, if the next bar’s open is positive, it’s a win
• If the price rises even slightly a few bars later, it’s a win

With this,

even a small price movement counts as a win

in the win rate.

---

What happens with this?

As I discussed in Episode 1

Win rate can be increased arbitrarily

＝Overfitting

Thus, the conclusion is.

■ Why that still doesn’t lead to winning

This is the most important part.

The problem is that you are not looking at the

“risk-reward.”

Or rather, you are ignoring it.

That is the core issue.

---

For example,

• +1 still counts as a win

• -20 still counts as a loss

If you trade in this structure repeatedly,

Even with an 80% win rate,

your capital will keep shrinking.

This is a trick of numbers

In reality, the win rate is

“the easiest indicator to tweak”

it is.

---

■ Why is that?

• You can change the decision criteria midway

• You can change the time frame
• You can change the viewing period itself

In other words, with a slight tweak

you can make it look “better” you want

as a metric.

■ So, what should you look at?

Here, for the first time, we reach the main topic.

What is truly important in trading is

not whether the direction was correct, butthese are:

“What happened after that?”

■ Key points to actually look at

• Reactions from the initial move
• How far it extended
• Where it stalled

Did it reach the set take profit or trailing trigger?

Or was it a BE (break-even) exit?

Was it hit by a stop loss?

Only by looking at this can you determine

whether it’s a usable logic.

■ Why does the outcome change every time

Here is the sensation many people feel.

“It looks like the same signal, but the result varies every time.”

This is not a coincidence.

Because the unseen conditions are different.

■ What’s visible vs what’s invisible

What’s visible

• Signals
• Chart patterns
• Indicators

What’s not visible

• Why it stopped there
• Why that direction was negated
• Why it extended afterward

This is the part of trading that cannot be gauged by win rate alone,

the essence of trading.

■ The more you do it, the harder trading becomes

Let’s return to reality here.

“The conditions that should be traded are

not that frequent.”

Because of that.

■ Therefore, what tends to happen is ↓

• You can’t wait for a chance
• You end up making a trade anyway
• Losing leads to emotional, reckless trading

This is simply repeated.

So, what should you choose?

Here, the answer seems within reach.

■ What makes for a good logic

✔ It is simple for everyone

✔ It can be reproduced anytime

✔ It reduces unnecessary entries

And most importantly, this:

“Knowing when not to act”

■ Preview of next time

“So, what are those ‘unseen conditions’?”

This is where the real core begins.