The Trends Tool is an excellent way subscribers can find DFS value using historical data. Users can create and follow trends by filtering through thousands of data points over the past several years worth of games.
Below is a picture of the Trends landing page, on which you'll find a list 'Pro Trends' that have been created by our in-house experts. These can be filtered by fantasy site by toggling the checkmark on the left side of the page. You can also sort by Plus/Minus, player count, fantasy site and average ownership by clicking on the respective headers.
To create your own trends, click the green 'Create a Trend' button in the top left corner of the screen. This box will then pop-up.
Before getting started, you'll need to create a name for your trend, choose the sport, choose the fantasy site and the position. NBA trends encompass all positions, but you can create trends for certain positions for sports like NFL and MLB.
After you click 'OK', you'll be taken into the Trends Tool, where you can create and edit trends. New trends all start out with every data point in our system, which you will see under 'Count'. The general goal when creating trends is to improve the 'Points +/-' number, which is how much they over (or under) performed their salary based expectation.
You can begin building your trend by clicking one of the filters on the left side of the page. Each category (Player Filters, Team Filters, etc.) has a number of filters within it. One simple way to get started would be to adjust the salary limits. Let's say you wanted to create a trend to find bargain players who have historically performed well. If you go to the Salary filter under 'Player Filters', you can adjust the minimum and maximum range.
If you make the upper salary range $4,000 on DraftKings, you'll notice that the Plus/Minus falls to -0.87. Since many minimum-priced players play few, if any, minutes, they're typically bad DFS choices. We can find valuable spots by further filtering out the data, though.
Since playing time is a big problem for low-priced players, we can find edges by looking at their projected minutes, which is a filter under 'Projections'.
As you can see, there's quite a linear relationship between expected playing time and Plus/Minus, which is represented by the line in the graph. Players expected to play 10 minutes or fewer are generally horrible plays, but the Plus/Minus goes into the green right around the 20 minute mark.
If we bump the lower limit to 30 minutes, which are likely cheap players thrown into the starting lineup due to injuries, we begin to see some serious value. Our Plus/Minus has risen to 4.58, meaning these players have out-performed their salary based expectations by nearly five points on average. The average ownership, which was just 2.6% for all players at $4,000 or lower, has bumped up to 10.5%, while the sample size has decreased from nearly 65,000 to just over 2,000.
You can try to further improve the Plus/Minus by adding more filters, for example the Vegas total for the game.
Players priced at $4,000 or lower with a minutes projection of 30+ have performed even better in games with totals of 220 or higher. You'll see that the sample size is a tiny fraction of where we began, though, now down to just 336 total past matches. If you add too many filters to your trend, the sample size can get too low to the point where only a handful of players will match the trend each year. Trends with tiny sample sizes are also more prone to variance, as the stats may be skewed because of just a few games.
Once you have your trend set, you can view past or current matches and a visual breakdown of player performances.
This particular trend has far more matches with positive Plus/Minus performances than negative performances. Since this trend has a low salary ceiling, it's not possible for players to have a Plus/Minus worse than -20.
Trends will save automatically, which means you can click into another tool or leave the page and you won't lose any of your progress. You can then find any trends you created on the main page by toggling from 'Pro Trends' to 'My Trends'.
Trends can also be utilized within Models, where you can see how many trends each player fits on a given night.
If you scroll slightly to the right, you'll see how many 'Pro' and 'My' Trends each player fits. You can also adjust the 'Pro Trends Rating' slider when building or editing a model.