Written by Alex Vella

To give you a bit more insight into the performance of our users and software, we are going to release an article every month where we go through our overall results, as well as taking a deeper look into our performance with each individual sport.

We will do this with the help of both our big data tool and back-log data. The back-log data was used to find out the number of trades placed, flat ROI and actual ROI. This is because the big data tool has a cap of 10,000 trades to speed up loading time, so for all of our sports it will not show results of every trade placed in October. But for the average edge placed and closing edge data we have used the big data tool.

Using the big data tool to see our edge data, we tried to narrow down the trades we would analyse. Here are the types of trades we calculated the edge data for; bets placed 0-8 hours before kick-off, all odds types, odds ranges 1.0 - 4.0 and 1.5% - 15% edges.

The reason we used those filters is because they are the presets that we recommend to our users. See our recommended presets article here. For the back-log data, there are no filters, it is every trade placed no matter what the odds range, time placed or edge.

Limitations

Before we get into how we performed in October, it’s important to note some of the limitations of the data we will show you:

1. For the Flat ROI calculations, stake sizing of €1 per trade is used to remove the effect of users with large bankrolls skewing the results. Also, this way we can see whether our edge is based on beating the market, rather than the Kelly criterion.
2. There is a chance that some of the bets in this data are duplicated as multiple customers could be placing the same bet. But the chances of this happening regularly are low as our user-base is quite geographically-diverse, which means they have to use different bookmakers.
3. Something else that is possible but even more rare is customers logging their own bets manually within our software. Those bets are not recommended by us, but they will still appear in this data, win, lose or void.
4. Some of the sample sizes will be very low, meaning no conclusions should be taken from those data-sets. The bigger the sample size the more accurate the data. Generally, a sample size of under 2,000 trades will include a lot more randomness and needs to be taken with a grain of salt.

In the month of October, 163,584 trades were placed, €10,824,697 was turned over and €587,950 was profited at an ROI of 5.43%! Here is how each sport performed, broken down into recommended leagues (popular leagues) and non-recommended leagues (lower leagues):

Key Findings for October

Soccer

For both recommended and non-recommended leagues, it is interesting to see the disparity in results when comparing flat vs actual ROI. It shows the importance of first identifying you have an edge (3.05% in this month for recommended leagues) then implementing a successful staking strategy (to make it 5.01%).

As you can see, there are more trades available in non-recommended leagues, but it doesn’t necessarily mean they are more profitable. Although the edges may be bigger in smaller leagues, those markets are less efficient. So even if you find a 3% edge on the German 3rd division, because there is less liquidity in those leagues, a big swing could move it out of your favour.

Tennis

A successful month on Tennis when it comes to the French Open (only the grand slams are included in recommended leagues), but it’s only a sample of 1,349 trades. Over time, Tennis has been one of our least successful sports as it is not the most efficient market and sometimes in the smaller tournaments, players don’t tend to try their hardest to win. This is replicated in our results this month on non-recommended leagues with both flat and actual ROI’s below 1%.

So if you are going to trade Tennis, we recommend a pretty decent minimum edge and trade a bit closer to the start of the game when the markets are at their most efficient. Even with an average closing edge of 3.2% over 7,408 trades, it didn’t equate to a solid ROI.

Basketball

A very successful month with Basketball. Huge sample size on non-recommended leagues (basically every league outside of the NBA and Euroleague) and massive ROI of 8.39%. Seems to be a very profitable segment, the trouble will be getting your money down on some of those smaller leagues. But it’s pretty crazy to see that our edge was virtually non-existent in these markets. It was our staking strategy proving effective to create a decent ROI. Something to keep a look at for next month considering the substantial sample size.

American Football

The flat ROI will have to be tracked in the coming months for recommended leagues (NFL). Could just be the sample size of 3,426 trades and variance playing up but worth keeping an eye on. Especially when looking at the average closing edges of 3% & 3.4% respectively. Non-recommended league data shows another good example of having a successful staking strategy.

Baseball

Another sport with a low sample size, but you can come to some conclusions on the non-recommended leagues data which is basically every league other than the MLB. A nice ROI of 8.20% and one of the biggest average closing edges for the month. It looks like a great place to make some money, but like I said with Basketball, good luck trying to place your bets on the smaller leagues!

Ice Hockey

With no NHL on in October, there were only trades for the non-recommended leagues. But they proved to be quite profitable over a big sample size with a pretty decent average closing edge. Interesting to see that a flat staking strategy would have been profitable. Once again, something to keep an eye on for future months.

Rugby & Rugby League

Can’t make any definitive conclusions with these sample sizes, which is pretty common for these two sports. But if you did place a few bets on Rugby or Rugby League, you probably made a nice profit.

CS:GO (Esports)

Finally we look at the newest sport implemented into the software. And as you can see, CS:GO continues to be a very successful segment for our users with both flat and actual ROI’s sitting at 10% over 10,957 trades. Crazy to see that we are out-performing the average closing edge by so much, which seems to be a common trend on CS:GO.

I would be very cautious betting Esports when it comes to keeping your accounts from being limited though, from my personal experience it has led to me losing quite a few bookmakers. It could be smart to only place an Esports trade every now and then to make it look like it’s not your main sport. But maybe not. My experiences are just a very small size of the population!

--

And that’s all she wrote for our first look at Trademate Sports’ monthly results. Please let myself or the team know if you have any queries about this data and whether you would like us to add some other data in for November.

Start a free week trial (no card details required) of Trademate Sports.