【summary】

Estimated win rate is a value estimated from past results as ``this is probably the true win rate.'' Since the true winning probability cannot be directly observed, the estimated winning probability is used to determine the actual winning probability. It is an important concept widely used in investment, trading, horse racing, business analysis, etc.

Conclusion

What is the estimated win rate?

Estimated original win rate calculated from actual result data

is.

The true win rate cannot be directly observed.

So we make assumptions based on past performance.

Difference between true win rate and estimated win rate

Let's sort it out first.

TerminologyMeaning
True probability of winningOriginal probability of winning
Observed win rateActual recorded win rate
Estimated win rateTrue win rate estimated from data

for example,

  • 100 races, 60 wins

Then the observed win rate is 60%.

Based on this 60%,

“The true winning rate is probably around 60%.”

This is the estimated winning rate.

Basic calculation method

The simplest estimated winning probability is

推定勝率 = 勝ち数 ÷ 試行回数

is.

Here,

  • Number of wins = number of successes
  • Number of attempts = total number of attempts

represents.

example

If you try 100 times and succeed 60 times

60 ÷ 100 = 0.6

Estimated win rate is 60%.

Why is estimation necessary?

Short-term results are influenced by luck.

for example,

Mr. A

  • 10 races, 8 wins
  • Win rate 80%

Mr. B

  • 1,000 races, 620 wins
  • Win rate 62%

In many cases, Mr. B may be a more accurate representation of his ability.

This is because the larger the number of samples, the higher the reliability.

Concept of confidence interval

In fact, there is also an error in the estimated winning percentage.

For example, If you win 60 times out of 100,

The true winning rate is not always 60%.

In statistics,

“It’s roughly within this range.”

Think of it in this way.

This is called a confidence interval.

Utilization in investment

When evaluating investment strategies,

Beginners tend to only look at the winning percentage.

But what's really important is

  • Estimated win rate
  • Number of samples
  • Average profit
  • Average loss

is.

example

#### Strategy A

*Win 8 out of 10 times

  • Estimated win rate 80%

#### Strategy B

  • 620 wins out of 1,000
  • Estimated win rate 62%

Investors usually rate B highly.

The reason is data reliability.

Bayesian thinking

In practice, it is not just a simple win rate, It also uses past knowledge.

for example,

“The winning rate of general strategies is around 50%.”

Estimates are made taking this assumption into account.

This results in Even with little data, you can avoid making extreme judgments.

Common mistakes made by beginners

Judge your ability after a few successes

Three or five consecutive wins are not uncommon.

It is dangerous to judge only on short-term results.

Ignore sample count

Even if the winning rate is 70%, 10 times and 1,000 times have different meanings.

See only the winning percentage

You should also check your profit and loss rates.

Framework for investment beginners

結果データ
 ↓
観測勝率
 ↓
推定勝率
 ↓
信頼性確認
 ↓
期待値計算
 ↓
投資判断

The estimated win rate is a starting point, not a final judgment.

summary

Estimated win rate is the original win rate estimated from past data.

There are three important points.

  • True winning rate cannot be seen directly *Estimated win rate is calculated from past data *The larger the number of samples, the higher the reliability.

Beginner investors should not only have a high winning rate, but also If you get into the habit of checking how many times the numbers are based on data, you will be able to make more accurate decisions.

This article is for educational and informational purposes only, based on public information. It is not a recommendation or solicitation to buy or sell any specific security or financial product. Although care is taken with accuracy, the content and future investment outcomes are not guaranteed. Final investment decisions should be made at your own judgment and responsibility.