Portfolio Significance

What is it & Why Does it Matter?

In your Portfolio Dashboard you’ll now find a new section called Portfolio Indicators. There you’ll find all the main financial metrics like current bankroll, profit, capital and a new metric: Portfolio Significance.

Performance indicators in your Tradr dashboard

How Can You Judge a Betting Strategy?

It is possible for practically any system, even the most inefficient, to go through a ‘hot streak’ or a ‘lucky spell.’ Bettors often make the mistake of assuming that a system or strategy is a winner or loser based on a relatively small sample size.

This results in an individual either scrapping a potentially profitable system too soon or keeping faith with one destined for losses. While it varies according to various factors, it normally takes at least several hundreds bets in a system before you can assign a great degree of importance to the results.

The more bets in your system, the better. The results of a 3,000-bet system are far more reliable than one with 255 bets, for example.

There are, of course, other factors involved in finding meaningful results. At Mercurius, we have developed an exclusive statistical approach, Portfolio Significance, to help you understand if you’re on the right path.

What is Portfolio Significance, and Why Should I Care About it?

In a nutshell, Portfolio Significance is a means of calculating whether the results from your Mercurius account of choice are in line with the data from our backtest. This metric helps you determine when your Mercurius betting portfolio has completed enough wagers within a meaningful odds range for a specific trading strategy.

Mercurius’ sports trading strategies are based on a sophisticated statistical approach. As a result, you must wait patiently for our bot to make several hundred bets before you can determine the likely success of our approach.

How Does Mercurius Calculate Calculate Portfolio Significance?

This approach takes into account a variety of factors, including:

  • Number of Bets
  • Type of Strategy
  • Odds Range

The next step involves simulating trends and returns of portfolios with the same metrics as the results from our backtest.

Portfolio Significance runs thousands of simulations, including all the matches and their outcomes based on the available probabilities.

It follows this by calculating the average daily return and comparing it with the backtest's theoretical returns.

Finally, it calculates a score that outlines how closely the two results compare.

Why Do I Need It?

We understand that this form of statistically-based sports betting is new and potentially confusing for bettors. The creation of the Portfolio Significance metric is designed to help our clients to actively investigate whether they are happy with their Mercurius portfolio.

It is no different to investing in the ‘traditional’ sense. You would (or should) want to know how any form of investment calculates success. You would also expect to receive tools to accurately track performance.

The Portfolio Significance metric can provide peace of mind for clients if their portfolio is experiencing a downturn. If the score is 80%+, they know that what’s happening is within the expected levels of variance.

Likewise, if they have a high score AND a profitable portfolio, it gives them the satisfaction of knowing that the positive returns are a natural occurrence.

What the Results Mean

Overall, you will see a score of 0-100%. The closer to 100% you achieve, the closer you are to achieving the results of the data from our backtest. We have developed three distinct ranges, and your results will fall into one of them.

GREEN (80% – 100%)

Once your portfolio achieves a score of 80% or above, it is closing in on achieving a confidence interval of 95%, which is statistically significant (we explain this below). As the number of bets grows, likely, your Portfolio Significance score will too.

YELLOW (55% – 79%)

While a reasonable number of trades in your portfolio are statistically significant, you can’t yet have confidence that they’ll reach the outcome outlined in our information document. You can expect a fair level of volatility close to the maximum limit outlined by Mercurius Tradr. The score should increase as your system makes further wagers.

RED (0% – 54%)

In this range, your portfolio’s trades bear little resemblance to what you can expect from the strategy due to an extremely high level of volatility. Therefore, the results are not indicative of what you’ll achieve in the long-term. The score could increase or decrease as you place more wagers.

Understanding the Results

Let us now clear up any potential misunderstandings. It is true that until you reach a score of 80%, your returns can’t be considered as an accurate expression of the Mercurius trading strategy you use. It is also a fact that the score can increase or decrease as the number of bets placed grows. The score is more likely to increase once you reach the yellow zone, but that isn’t a guarantee.

Profit Does NOT Necessarily Mean a High Score!

However, you should not dismiss or acclaim any approach too soon. Nor should you assume that profitable strategies automatically have a high score. The number of wagers placed is hugely important.

You could have a score of below 55% even if you are currently in profit. This generally happens if you make a relatively small number of bets at moderate to high odds that are successful. This is because the trend line isn’t yet indicative of long-term profits. The low score is because the strategy will have upturns and downturns.

If the strategy is still profitable after 500+ bets for example, it will have a higher Portfolio Significance score.

A High Score Does NOT Necessarily Mean Profit!

Reaching the upper range of the Significance metric (80% – 100%) doesn’t imply you will be in profit. Simply having the ability to compare your portfolio’s results with those in the information document more accurately doesn’t mean you will reach the same metrics.

What Happens to my Portfolio Significance Score in the Event of a System Update?

We have changed strategies in the past, and may do so again in the future. If a strategy you use is changed, your Portfolio Significance score will reset. This is a natural outcome since the new strategy will have to be compared to different backtest results.

However, the score isn’t affected in the event of minor updates.

Portfolio Variance - Practical Examples

We have provided a couple of simulations of different returns generated by portfolios composed of similar bets and metrics. This should help you further understand how a statistical approach works.

What’s striking is the natural degree of variance. Such volatility can easily result in a portfolio generating a negative return. However, it doesn’t suggest that your strategy will continue to lose money.

In our examples, as the number of bets increases, the number of ‘losing’ portfolios decreases.

Conservative strategy outcomes within 12 months-windows

Histogram of a 12 months portfolio
After 12 months and roughly 700 bets on distinct outcomes, 78% of portfolios have positive returns.

Conservative strategy outcomes within 24 months-windows

Histogram of a 24 months portfolio
After 24 months and roughly 1400 bets on distinct outcomes > 99% of portfolios have positive returns.

Once you reach an important milestone of 700 bets, a significant majority of portfolios are in profit. With double the number of bets, almost all of them are. It is also worth noting that the amount of profit has also increased on average.

Compare the number of portfolios with 0.30+ after 700 bets with 1,400 bets.

Mercurius Trading Strategies Are Based on a Statistical Approach: What Does it Mean for me as a Customer?

The short answer is that you need to show patience! As the graphs above outline, profit becomes more likely as the number of wagers increases. This is the opposite of losing strategies which tend to become even less profitable over time. As the bets our bot places offer a positive Expected Value (EV), it will eventually generate long-term profit.

Here is a slightly longer answer if you’re interested!

Our betting strategies leverage a statistical approach to value betting. This means:

  • We use much more than statistics only to forecast outcomes.
  • You must place a very high number of bets to reach a significant number from a statistical point of view.
  • Our profit comes from a simple principle: We bet on every trading option we can find with a positive expected value.
  • If you place enough bets to reach a statistically significant number of operations, you will likely have a profit.

Also, the higher the number of operations:

  • The closer the trend of your portfolio will get to the trend described in the information document.
  • The more likely you are to have a positive ROC.
  • The more your bets will be a representative sample of the entire population of the bets with a positive EV (which we simulated in our backtest). Hence, your return will converge to the expected value.

What is the Confidence Interval?

When one computes a confidence interval, they will find that the true value lies within the computed interval 95 percent of the time. In the long run, if we keep computing these confidence intervals, then 95% of them will contain the true value.

Pragmatically speaking, a 95% confidence interval implies that 19 out of 20 times, your measures are the result of skill rather than random chance.

Why do you Have a Confidence Interval of 95% and Not 100%?

95% is the standard threshold in statistical science to make a claim with enough evidence. Given the high volatility of the sports betting domain, it is considered appropriate.

100% statistical confidence would require making an infinite number of observations (or always observe the entire population we are studying, which is completely infeasible).

100% is for Gods, not for machines.

Portfolio Significance Conclusion

Portfolio Significance is a metric designed exclusively by Mercurius. It involves comparing your selected betting portfolio to what’s in our information document. The key is to focus on the data that’s most closely related to your strategy.

The goal is to find out how close you are to achieving the results from the data sample you’re comparing your information to.

The aim is to achieve a score of 80% and above, as this indicates a strong likelihood that your results will ultimately match those of the backtest over a long period. It is an innovative way to let you know when it’s the right time to begin evaluating your portfolio.

Above all, it emphasises the importance of gathering a sufficient sample size before drawing conclusions. Regardless of the betting or trading strategy, in general, more bets = more accurate results.

Sports Trading powered by Artificial Intelligence

Tradr Bot is the first service for Betfair Exchange able to analyze sports, find the edge, and bet for you.

Learn More