The PGA product is finally here! And to get you quickly up to speed, check out this robust glossary of all of our PGA stats, as well as this playlist of tutorial videos.
PGA Video Tutorials
In my opinion, our golf product could end up being the best one we offer. First of all, daily fantasy golf just sets up really well to model. There are a ton of relevant stats (and we have all of them), and, like baseball, you can create a really powerful model without the use of projections.
The second reason I’m excited about PGA is because we have one of the sharpest people I’ve ever met – Colin Davy – running the product. We introduced Colin on the intro FantasyLabs PGA Podcast. He won Hackathon at the MIT Sloan Sports Analytics Conference last year (and he’s defending that title later this week).
Colin, who works as a data scientist, started modeling golf and tennis years ago in an effort to find inefficiencies in the sports betting market. He’s worked with us to incorporate a lot of those elements into what we’re doing with our PGA data and tools, and I’m very confident you’re going to like the results.
Here’s why Colin is excited about the PGA tools:
One, we’ve made it easier to analyze the data properly. When looking at results, plenty of people take shortcuts, like number of top-10 finishes, or treating all missed cuts the same, whether it’s by one stroke or by six. With proper massaging of the data, you’ll be interpreting it correctly.
And two, we include data from everywhere. There are enough players that play on other tours a lot (Euro, Web.com, even sometimes in Asia) and it’s tough to put data from multiple tours under a single roof.
And even if you could get it there, it’s not necessarily apples-to-apples. We do all that crunching ahead of time so that we make it apples-to-apples, and that way, there are no developments you’re missing.
I was high on this guy Matt Jones earlier in the year, and it’s because he beat both Jordan Spieth and Adam Scott in the Australian Open – an otherwise obscure event that you’d never account for. With the amount of noise and variance in PGA, it actually leans more towards quantity than quality of data, so it’s the one sport where you have a good chance of including data that other people aren’t.
With that said, here are the top five reasons I am in love with what we’ve released for PGA.
1. The Data
As Colin said, golfers participate in different tournaments, many playing overseas on courses for which most people just don’t have data. We have data from every tournament – Euro, Web.com, Asia – and that’s important. For that reason alone, we’re going to be able to make better predictions on the crucial golfers for whom others might not have all the necessary information.
And in terms of the stats we’re offering, here are just a few (see more in the glossary):
- Adjusted Round
- Greens in Regulation
- Driving Distance
- Driving Accuracy
- Putts Per Round
- Scrambling
- Field
- Missed Cut %
- Vegas-Implied Odds to Win
2. The Adjustments
The stats we are going to show in models are raw, but we adjust them to make them more accurate behind the scenes. Here’s a preview of what you’ll see:
So how do we adjust the data? A few ways. One is that we need to adjust for course strength. Guys who are playing very difficult courses shouldn’t be penalized as compared to golfers on extremely friendly courses who are racking up points.
Second, we adjust for the field. Some of the guys playing in obscure tournaments against weak fields might be scoring a lot of points when they wouldn’t be at the top of the leaderboard against a stronger field.
We also adjust for things like weather; why should a golfer who teed off at a time when there were 25 mph winds be penalized for that moving forward?
For all stats, we calculate a z-score to help standardize certain stats; a golfer who finishes -3 but beats the field by six strokes is more impressive than one who finishes -3, tied with five others, for example.
3. The Content
Every week, we’re going to have a number of free strategy pieces, both time-sensitive and evergreen. This will include a weekly article from Colin, as well as a weekly podcast with Colin, CSURAM88, and Bryan Mears.
Our hope is the content supplements the awesome tools we’re going to offer.
4. The Modeling and Backtesting
Of course, FantasyLabs is all about creating (and backtesting) DFS models. No one else provides the ability to not only quickly create a daily fantasy model, but also immediately know if that model has the potential to be predictive of DFS value moving forward.
Further, because we use Plus/Minus to compare actual performance data to salary-based expectations, we can quickly figure out which stats are already priced into DraftKings PGA salaries. Vegas might be important in predicting who will win, for example, but that doesn’t automatically mean it’s predictive of value.
You can also use dozens of stats in your models, including weather. We even match up hourly weather forecasts with golfers’ tee times to make sure we’re accounting for what sort of weather each golfer will see when he’s on the course.
And of course you can build lineups – up to 100, all based off of your specific model – in seconds using our lineup building tools.
5. The Research Capabilities
And finally, we’re the only site that allows you to sift through a massive database of stats to identify which stats are predictive of fantasy success, then create trends to locate high-value plays in the future.
Take a look:
The coolest part? We have course-specific data, so you can easily test certain narratives. Think long courses favor golfers who drive the ball far? You can immediately figure out if that’s true, then adjust your model accordingly.
———————————————
I’m really excited about what we’ve launched for PGA, and I think you will be, too. As with everything we release, every aspect of it will be the most premium offering in the DFS industry.
With the way golf sets up and the team we have on board, this has the potential to be our most useful product yet.
