Want to pick the ultimate fantasy team or make a more informed decision with your golf bets? Dave Tindall talks to Stephen Richter, the co-founder of FantasyGolfMetrics
The thirst for deeper and more meaningful golf stats grows and grows.
Golf is one of the most popular sports to bet on in the UK while those who traditionally liked a punt are also turning to fantasy games such as DraftKings as an alternative way of backing their own opinions to win money.
Twitter led me recently to an American website called FantasyGolfMetrics, a leader in the fast-growing fantasy golf data field.
Several of their tweets piqued my interest as I’m forever fascinated by piecing together clues and theories in the search for winners so I got in touch with their co-founder, Stephen Richter, to find out more.
Golf365: Hi Stephen. Tell us some initial background on FantasyGolfMetrics.
SR: Hi Golf365. We’ve been operating since 2015 and have mainly focused on digging deeper into the standard summary statistics that most people were used to using for their research. at the time, we were the first site to offer a moving average for analysing statistics, the first to offer stats vs the field, and the first to offer comprehensive putting statistical splits based on turf (bermuda, bentgrass, etc).
Humbling to have so many people support the site and subscribe! We're excited to build upon our current platform throughout the year.
Whether you've already joined or are considering it, we've gotten a lot of questions. Click here to watch our tutorial: https://t.co/jinLo9Jdve pic.twitter.com/wCL35bI7FJ
— FantasyGolfMetrics (@FGMetrics) January 26, 2018
Golf365: You’re a subscription site but, to be honest, if I’d produced something of this magnitude, I’d want a few quid coming my way too! If you like a bet/fantasy, all I can say is that the $119 for a yearly subscription is money well spent.
SR: Thanks, we were free for the aforementioned 2.5 years up until two weeks ago but to bring the site to the next level we needed to start charging unfortunately.
Golf365: How long did it take you to plan and build the site? Was the original motivation to help fantasy golf players?
SR: The site is run by myself and a partner/long-time friend Robert DiNardo Jr. (Bobby). We have run it since August of 2015.
The motivation came from seeing a lack of credible golf analysis in the marketplace. Bobby had been putting together his own excel sheets tracking data and he came to me to discuss how to model it out better. He knew my analytics background and we’re both decent sticks, he’s a 5 and I’m a 1 handicap. That plus the fact I wrote my economics thesis at Princeton University on the efficiency of the gambling marketplace for the PGA Tour and we had a match made in heaven.
We initially thought we could have a huge edge on the rest of the players but the more we saw the flaws in current modeling and player ranking the more it became clear there was a need for true golf data analysis.
We began putting out stats based on a 12-week moving average (which most sites are now considering as standard). We felt it didn’t make any sense to look at year-to-date statistics as most people were at the time.
We also realised that stats at one event varied from those at another, and put together what we call rData (short for relative data). This means we look at stats relative to the field for an event. For example, take Player A and Player B. They both played in two different events and had Driving Accuracy of 60%. So they’re the same, right? Not necessarily. If the average driving accuracy for the event Player A competed in was 55% and the average for the event Player B competed in was 65% then Player A outperformed the field and Player B underperformed. Their rData for Driving Accuracy would be 5% and -5%, respectively. This means that Player A has hit 5% more greens than the field. This is one of our biggest differentiators as it truly helps us see who is playing better than the competition. As with the moving average, many other sites have incorporated this into their offerings.
The final concept we really brought to the forefront, and we won’t take credit for the idea but we will say that we offered the most in-depth research, is Turf Splits analysis. We dug deep into how players performed on different green types, looking at Strokes Gained Putting on Bermuda vs Bentgrass (the two most common types) and saw that certain players had extremely different results. We expanded this to finish position, Fantasy Points, and pretty much any other statistic that could have variance when considering turf types.
What’s all the fuss about Noren? Golf365: You have a stats-loaded spreadsheet called the ‘Survival Guide’ for each tournament, ranking every player in the field. As a European, I notice that Alex Noren ranked second and Paul Casey third for this week’s Genesis Open at Riviera. What puts Noren above the likes of Jordan Spieth and Justin Thomas? One reason, I guess, is that he ranks 1st in the ‘Secret Stat’. It’s probably a daft question, but can you tell us what that is, or at least drop a hint?!
SR: The Survival Guide is our bread and butter and what most of our early adopters were and still are excited about. A One-Stop-Shop for all the detailed analysis we do.
In terms of the rankings, yes, technically Noren is ahead of them this week for a handful of factors but you’ll notice we have “Classes” set up. Our 1st Class is always the top notch and we like to consider those players somewhat interchangeable in terms of form and history, etc.
As for the “Secret Stat”, we get this question all the time and it’s funny the term has become well known as we don’t call it that internally. I can certainly drop a hint and we’re working on a more academic style paper for the metric but the reason we have kept it secret is because it’s the only thing we have that is proprietary at this point. It’s extremely flattering to have many of the leading sites use our concepts but it can become frustrating when building a brand. We use the secret stat in both our model and our betting picks and if you ask any of our users you’ll hear we’ve had tremendous returns (2017 season showed an ROI of 193%, almost tripling the initial investment if you followed our system and also helped us pick 50/1 winner Gary Woodland two weeks ago).
Here is what I can offer on the stat: It is a complex calculation that allows us to see if a player is truly giving themselves a chance to perform well. It takes out a lot of the metrics that have significant variance and enables us to say “Hey, this guy may not have had the best finishes recently but he’s not playing poorly and there could be tremendous value”. I know sceptics may hate that we haven’t divulged but I can’t be clearer when I say that approximately 95% of my personal betting picks come from analysing the trends of this stat. More to come!
Golf365: On Twitter, I saw you post this: “Key Stats for #GenesisOpen – Hit it long and stick it from 150-175! The average winner at Riviera takes 14.4 shots from 150-175 compared to the Tour average of 8.2 and hits almost 10% less fairways (but 2 yards further).
Looking at 150-175 stats, leading players relative to par from that distance are:-18 Brian Gay-10 Schniederjans, Reavie, Van Aswegen, McGirt, Stuard.In terms of attempts, Gay (-18 from 72), Schniederjans (-10 from 44), Reavie (-10 from 49) are best. Also: G Murray -9 from 37. https://t.co/TBVhBG391V
— Dave Tindall (@DaveTindallgolf) February 13, 2018
I immediately went to the stat on pgatour.com that measures 150-175 yards. First thought was that I should be playing Rafa Cabrera-Bello as he’s top of the list. Then I noticed that his ranking is based on just three shots hit from that distance! Scrap that! Also, I realised that it’s based on average distance the approach finishes from the flag. All well and good but what if you don’t hole the putt!
Then I saw on the far right of the page a column showing their score in ‘Relation to Par’ from approaches hit from 150-175 yards. Brian Gay, despite being 112th on that list due to not hitting it as close as most on the proximity list, is a way clear on -18 under ahead of -10 Ollie Schniederjans, Chez Reavie, Tyrone Van Aswegen, William McGirt and Brian Stuard. But, of course, again it probably pays to get an average. Gay is -18 after 72 shots from 150-175 yards. Schniederjans (who hits it sixth closest) cashes in strongly as he’s -10 from 44 attempts which is better than Stuard’s -10 from 90 attempts. Question: Are my thought processes valid?!
SR: I think your thought processes are extremely valid! Again, this also brings rData into play as we’re looking at how much closer a player hits it to the hole compared to the fields he has played against. This should help explain some of the variation in scoring.
It’s funny you mentioned the players you did as a few of them are ranking relatively high for us. Each week we put out our official Key Stats on twitter. It has been one of our more popular postings as other sites also use it to build their models. We generally find the most important yardage gap each week and give it a 5-10% weight on our overall model although sometimes the analysis leads us to say all around approach is important and not just one gap.
We run 18 different regression analyses each week to determine which stats should be considered. We try to keep it simple with a basic percentage weighting but I can assure you what we see on the back end would be a whole college semester in statistics to explain and that’s not what anyone wants to try to piece together.
Taking a closer look at Gay’s numbers Golf365: Having identified Gay as the leader, my heart sank when I noticed he’s missed his last five cuts at Riviera! Is it because he’s short off the tee (189th in Driving Distance, averaging 285 yards) and doesn’t get to hit from that distance (150-175 yards) as much as others on this course?
Schnierderjans is a big hitter (19th in DD, averaging 309.5 yards) so, in theory, is a good fit at Riviera. His one start there backs it up as he was T8 on debut last year and shot all four rounds in the 60s. As for the others, Reavie T7 in 2016, McGirt T6 in 2014, Van Aswegen T20 in 2016. I think we know this isn’t an exact science but do some of the above results back the 150-175 theory up?
SR: What we find is it often becomes easy to fit a narrative into any scenario. What you described all makes sense. We have shown that being longer off the tee at Riviera has helped players perform and that’s not a forte for Gay.
I think you’re on the right track in terms of checking off boxes for players but I wouldn’t make it my main focus. I would consider looking at more encompassing statistics like Strokes Gained Tee to Green and recent finishes (and Secret Stat!) and then once you have some guys you like, narrow it down using the more specific categories like 150-175, or even considering how they perform on the turf. That’s generally what I do.
Golf365: Using examples to illustrate what you have to offer on FantasyGolfMetrics, let’s look ahead to The Masters. If I was thinking about putting money on Jason Day, how can the site show me he’s doing well in the key categories for success at Augusta National?
Is Day the right pick for the Masters? SR: Great question. As we don’t offer the key stats for every single course all at once you may just need to go off the basics.
We generally know that Par 5 scoring, Scrambling, and Putting are important at Augusta so you could check out these stats on both our StatSearcher and Course Snapshot tools. If there are players not currently in this week’s field but you know will be at Augusta you can change the dropdown above the player names to “Tour” and it will show all the players who have competed on Tour this season, not just the field.
Golf365: You have a section called ‘Turf Splits’, showing how players compare on different grasses (as mentioned above). Next week we switch back to Bermuda for the start of the Florida Swing. How do users find the best putters on Bermuda or is it a case of thinking of a player to back/invest in and then checking to see if he has good Bermuda stats?
SR: We love when players switch turfs from week to week because that’s when we can sometimes find value. If a player has been striking it well but putting poorly and then they head to a course that has a turf they have putted well on then you could have an opportunity.
It’s very simple to rank the players by Strokes Gained putting and a handful of other stats on that tool. You also want to check the variance from one turf to another. A player may be a poor putter regardless of turf so the SGP could be bad all around but it could be a lot better on a specific green type.
Look at Tony Finau, he was a poster boy for us when we developed this tool originally. He prefers Bentgass over Bermuda and it’s extremely clear. Also, if you click on any of the player rows it will show a pop up with more detailed information. We love using this tool when we’re between two or three players and need to make a final call when filling out a lineup!
Finau: Has a clear grass preference Golf365: Talk us through what you have to offer in the Player Profile section.
SR: We love this section because it’s the only place you can get all the information you need on a single player. You want player background and news? Check. You want to see how they perform in the Key Stats? Check. You want to round-by-round and event-by-event data for Recent Form and Course History? Check. How they perform on different types of courses and what their DraftKings pricing and ownership has been? Got it. Then we have this neat tool on the right that lets you hover over a golf hole and show the hole by hole filterable stats. It’s really a wealth of information at your fingertips!
Golf365: Do the metrics show that course form is worth more at certain venues? Augusta National would be an obvious example.
SR: Oh boy, the “Course History” debate. There’s a running joke on Twitter on whether you’re #TeamCourseHistory or not. For the most part, we are on that team but we don’t think it should be a significant consideration.
The way you have phrased it is perfect. It matters more at certain courses. When we run our Key Stats analysis each week we see that certain factors are more consistent at one course but maybe not as consistent or predictable at another. The more data we have for a given course, the better we’re able to tell.
This week it was pretty clear what matters and I think the history of top performers shows that. Like you said, Augusta is a no brainer. Why do you think there hasn’t been a first-time winner since 1979? Course Experience and Course History may not be the main factors but they certainly do factor.
Golf365: Thanks Stephen. Good luck with those bets!
Full-site tutorial: https://www.fantasygolfmetrics.com/blog/full-site-tutorial-188a6c5f
The post Crunching the numbers with FantasyGolfMetrics appeared first on Golf365.com.