So far, 2023 has been the first “normal” year in terms of the number of worldwide competitions and matches since that fateful March 2020.
And like we do every year around these times, we’ll show the tennis stats since our launch as well as those relevant to the first semester of 2023.
We’ll make use of all the data generated since the birth of WinnerOdds Tennis, more than 6 years ago!
General stats
We began to log bets in May 2016, and since then, over 65,000 bets have been placed.
This is the overall profit graph in units:
It shows the units gained when betting on opening odds, average odds, and minimum profitable odds.
Units gained from every bet placed:
- The green line shows the units gained when betting the recommended stake on the odds logged by the first user to place a bet.
It’s a “theoretical” graph, as no user could be first to place a bet on every match and have every bookmark available.
- The yellow line shows the units gained when betting the recommended stake on the average odds per match generated by all our users. We create it by calculating the potential profit of every logged bet, the total volume wagered by our users, and drawing out the average odds with the formula 1+(potential profit / wagered volume).
It’s also a “theoretical” graph, as it depicts a user who places every single recommended bet on the generated average odds.
- The blue line shows the units gained when betting on the minimum profitable odds, which can be interpreted as the result that a very slow user would face. A user who bets on every match right as the odds are about to drop below the minimum profitable odds.
The case closest to reality would be a user who places around 40% of the total bets on the average odds.
Only 40% of the total bets?
Yup, but keep in mind that so far in 2023, we’ve prompted over 6,000 bets.
In other words, it boils down to about 400 placed bets per month.
Fantastic.
You see, the total profit generated by WinnerOdds’ users can be seen on our homepage. This graph shows how it has developed as all bets have been placed:
As it stands today, we’ve surpassed the 9 million euro mark across all our users!
That’s a milestone that, if someone in 2016 had told me about it, I would’ve called them insane.
But let’s keep going 🙂
If, instead of using the X axis for the number of bets, we use it to represent time, we can witness the profit generated over time.
We can observe a rather prosaic first few years, climbing up at a slow rhythm that stopped in 2018, a spectacular year that kick-started an updraft only interrupted by Covid in 2020. Later on, we see a different but also positive course in 2021 and 2022.
The rising pattern continues on to 2023, reaching the highest profit ever at the end of the graph and getting us ever so close to that glorious 10 million euro threshold.
Every time the profit graph drops relative to the previous highest value, we can calculate the drop—or drawdown—and notice that the maximum drop experienced by our users happened in late 2021.
Other than adverse results, that’s also due to the fact that the average bank our users work with has gone up in recent years.
As you can see, the deepest drawdown we faced neared the 300,000 euro loss across our users and took place in 2021.
However, by using a positive expected value method, we always manage to recover from temporary downswings (although not all users are able to withstand them).
If you’re the kind of user who may lose all your strength amidst a drawdown, here’s an article I wrote about how to deal with them:
– Five tips to overcome a drawdown
Cool,
Every time the red line hits €0, the highest total profit ever achieved goes up.
While, yes, the magnitude of the losses may sound staggering, it’s important to keep in mind that variance is an inherent property of investing in sports betting.
In fact, as we place more and more bets, the likelihood of facing a maximum drawdown increases, as much as we hope it takes a while to repeat.
Failing to have that in mind causes some of our users to stop using the tool. Recently, I wrote an article to explain why people don’t use WinnerOdds correctly.
– Why people fail to use WinnerOdds properly.
As far as the total wagered volume goes, we can see that it follows a fairly steady growth pattern, although due to the average user bank having increased in recent years, the incline is more visibly pronounced later on.
No trifling matter. A total volume surpassing the 220 million euros wagered is nothing to sneeze at!
Results charts
Following recent results, many users have started to see patterns out of nothing.
“qualifiers aren’t profitable”,
“Fridays are always red!”
“every time I lose a fat bet, the whole week goes straight to the dumpster…”
etc…
All right, if we take the data into account, history dictates that the best decision is…
…to let WinnerOdds say when and how much to bet.
Profit per round
In this first semester of 2023, we’ve greatly improved in 1/64 and 1/32 rounds, which, looking at the global results chart, spells out that it’s an improvement coming from short-term variance.
The following charts showcase the profit obtained per round both since WinnerOdds’ inception and in 2023, drawn with the user-average odds.
Comparing the global stats with the 2023 stats, we can see that the total yield hardly changes.
Which shows, without a shadow of doubt, that the system is still as alive as ever.
It’s been, much to the dismay of those who lose their marbles after 3 negative weeks and are nowhere to be seen during the positive ones…
…a year closely in line with the average, sitting at a 6.37% yield versus the 6.36% generated since our beginnings.
Profit per month
A belief ingrained in our world is that the start and end of the season are less profitable.
That December and January are rough months to bet because players go on vacation, have their minds elsewhere, or are not in shape yet.
Well, looking at the data, we can see that’s a mistaken belief. It’s not the case at all. In fact, it seems to be the opposite.
Having fewer chances due to holidays and off-time doesn’t correlate with lesser winnings.
If we look at the overall panorama (first chart below), we can see that the yield per month is very similar across the full calendar year.
In fact, in the long term, December turns out to be the most profitable month yield-wise…
The same chart in euros paints a slightly different picture, generally a little worse as far as yield goes, but it’s the most realistic outlook we have available, and it still comes out positive across the board.
In 2023, we obtained a yield slightly above average, but the variance per month has been significant, starting off with a fairly weak couple of months and continuing with a superb stretch in March and May.
Profit per day of the week
These charts aren’t so different from the round stats, but qualifiers are sometimes played on Sundays, so let’s take a look at the results classified by day of the week.
In the first 6 months of the year, 6 out of the 26 weeks have been negative.
Or, to put it differently, 76% of the weeks have been positive.
If you’re the type of person to say, “this week is in shambles…” “we can’t get out of the pit…”
…widen the scope and keep placing bets 🙂
Frequency of bets by day and hour
Here are a handful of useful charts to help you figure out the prime time to bet and when you can expect to receive the most notifications:
Knowing when you’re likely to receive new picks can help with creating productive routines and habits.
Frequency of matches by day and hour
Mostly just to satisfy your curiosity—especially if you’re the kind to watch every live match on your three-monitor setup—this chart shows the day and time that relevant matches start.
Evolution of the total profit visualized in candlestick charts
Even though we’ve already looked at the evolution of the total profit over time, using candlestick charts to depict the same growth is better to highlight the number of positive and negative periods as well as the maximum swings within them.
Yet again, this shows that the first half of 2023 has been spectacular, as every single month has ended up positive.
Drawdowns sting, but they’re commonplace in an investor’s journey.
Yearly evolution
If we establish that tennis runs at yearly intervals based on the vacations in December and the different court surfaces, analyzing our results per year can give insight on how the system has evolved, although the progression is also influenced by how bookmakers have adapted.
These charts reveal that from 2018 onwards, the system was able to spot many more valuable selections than before, and that despite the slightly lower yield, the profit increased due to the much greater wagered volume (except in 2020 due to the nearly 6-month long Covid pause).
Even though we’re just halfway through the year and these results are subject to change, we’re on course for another great year. To be more specific, if we keep up the pace, we’ll be on our way to the 3 million euro mark across all our users.
We may just be lost in reverie—we don’t know what the future awaits—but these results encourage us to keep working hard to beat the yearly profit record.
Results per competition
Inquisitive scrutinizers and those with their eyes always glued to the ongoing matches love to ask…
Are Challenger games really profitable? Is it worth it to bet on WTA and ATP games?
Here’s all the data classified by competition, both since WinnerOdds’ launch and in 2023.
Bets per week
The increase in bets can also be observed by looking at the number of weekly bets placed over the years:
Since mid-2018, we’ve prompted up to 350 bets per week.
If we also consider the number of weekly matches, we can compute the percentage of matches we bet on.
This year, the percentage of matches we bet on increased a little relative to 2022 because some competitions have gradually healed from the pandemic, especially in Asia.
The algorithm adapts to the number of matches, and if it finds value in all of them, I’m not gonna go and argue it!
I’m just a mere human being 🙂
Total profit per bet
Among all users, how much do we gain or lose per bet?
Take a look. This is the distribution, and as you’ll clearly see, there are more winnings than losses.
Often, we make very little back in a single match (few bets on a few bookmakers), but as a whole, we nickel-and-dime our way through.
The max we lost in a single bet was €56,000, and the max we gained in a single bet was €23,000.
This is because the average weighted odds we bet on are inferior to 2.
However, the hit rate hovers around 70%, which, collectively, lets us make more money than we lose.
Real yield vs. expected yield
This graph shows the yield obtained based on the relationship between the odds we bet on and the minimum profitable odds.
It shows various interesting facts:
Even after all these years, betting on the minimum profitable odds returns a yield above 0%. In periods where the expected yield was 2%, we obtained an even higher one, close to 4%.
The higher the odds relative to the minimum profitable odds, the higher the yield obtained, but it’s not a proportional measure, and a massive difference doesn’t ensure a massive yield.
It’s a sign that when WinnerOdds’ probability estimation varies wildly with the bookmakers’ estimation, both of us are likely overestimating the probabilities of a certain result and its counterpart.
While, indeed, the greater the difference between the odds and the minimum profitable odds, the greater the yield, we have to apply a fractional Kelly criterion to generate the stake size, as an unmodified Kelly criterion wouldn’t take into account any potential errors in our calculations.
It’s most common to spot small discrepancies with our minimum profitable odds, which is only logical considering both WinnerOdds and bookmakers employ similar information sources.
Conclusions
The data paints a clear picture: the system works better than ever, it’s managed to adapt to the post-pandemic circumstances, and all competitions are entirely back to normal.
2023 has been a good year so far, carrying on with the great results we’ve amassed historically.
Any user who makes good use of the tool is certainly obtaining a great profit.
As far as more specific data goes, we can’t conclude that any round, day of the week, or month isn’t profitable, but we can make out more and less favorable patterns.
Tuesdays, Wednesdays, and Thursdays are the most active days, as they comprise 57% of the weekly bets, which mostly take place between 5-7 a.m. and 5-10 p.m. (UTC +0).
We hope 2023’s second semester is as good or even better in terms of number of bets, profit, and yield, and we’d appreciate not having to undergo a drawdown like that of October 2020… but that’s impossible to know.
Variance has no memory, so we’ll have to keep believing in the long term.
Best wishes, and let’s conquer the second half of the year! 🤖
Cheers,
Miguel Figueres
