We've joined the rest of the world and now have a newsletter: Letzig's Latest.
Begins January 2nd, 2023.

Using shot-level data we’ve created a new page that offers a thorough analysis of each player’s approach game.

Approach performance can be analyzed across 6 different yardage/lie buckets, which allows users to see where players excel—or struggle—relative to the rest of their approach game. Take Collin Morikawa: he’s an elite mid-to-long iron player who is surprisingly below average from 50-100 yards.

We have also added stats beyond strokes-gained to help analyze performance. Supplementing SG data with proximity, GIR, great shot tendency and poor shot avoidance helps us understand why a player is good or bad in a specific bucket. For example, in terms of SG per shot, Justin Thomas has consistently been one of the best wedge players on tour. It turns out that while Thomas hits a lot of poor shots relative to his peers, he more than makes up for this by hitting more great shots than anyone else. For those interested, we wrote a blog post on Thomas’ iron game and this new page in general.

After a few failed attempts, bet tracking has finally made it to DG!

Users can now effortlessly log bets from the betting tools and the tracker will do the rest. While bets are in play, live probability and/or status updates are provided on the bet tracker summary table. As bets are completed, our auto-grading algorithm tallies up the profit/loss and displays it at the top of the tracker, providing users with a live summary of their current betting week.
A tutorial of the bet tracker is available here.

The course fit page now provides our predicted number of approach shots for each golfer from various distance buckets. These are located in the second tab in the table below the course radar plots. Where possible, these predictions come from models that use historical shot distributions and golfers' driving distance and accuracy skill (e.g. What is the typical approach shot for a long, inaccurate player on #18 at Bay Hill?). When shot-level data is not available, we generate predictions based solely on scorecard yardages. To complement these predicted approach distributions, we've also added golfers' predicted skill at 6 different approach buckets (e.g. 50-100 yards from the fairway). These predictions are generated from models that use shot-level data in addition to our round-level predicted skills for SG:APP and SG:ARG. These approach profiles are available at the top right of players' profile pages, as well in a player's row in the approach shot distribution table on the course fit page.

A new endpoint has been added that returns our fantasy projections for all slates offered on the projection page. Projections via the API will only correspond to those found on the page when our default settings (weights, course fit, course difficulty, wave splits, etc.) are applied.

Expandable scorecards have been added to the live tournament stats table. Clicking on a player's row will expand their scorecard to show hole-level strokes-gained and traditional stats. While potentially overwhelming, we hope this additional data provides users with a more comprehensive understanding of a player's round.

The "hold" for each sportsbook on our Finish Tool is now displayed when you hover on their logo. This provides a quick way to assess how much the odds are skewed in the book's favour across different sportsbooks and markets. The orange bar above each logo is coloured according to the sportsbooks' relative hold percentages: higher opacity indicates lower hold.

An additional table has been added to the bottom of our player profile pages that gives users the ability to filter data on event and/or course. Multiple selections can be added at once and will be saved as users navigate across players. The image below shows Collin Morikawa's short, yet impressive, history at Major Championships.

Non-SG stats displayed on the historical tournament stats pages have been added to the raw data endpoints. For PGA Tour events with Shotlink data we've provided our own versions of driving distance, driving accuracy, greens in regulation, scrambling, fairway proximity, and rough proximity. We are able to leverage the shot-level data to improve these statistics slightly (details on each statistic here). At non-Shotlink events, the PGA Tour's version of driving distance, driving accuracy, and GIR are provided.

Probabilities for any finish position from 1-50 are now available on our live models in between rounds. The dead-heat probabilities toggle is applied to this custom finish position as well.

A stripped-down version of the PGA Tour's My Model page is now available for any OWGR-sanctioned tournament. Users can manually adjust predicted skill levels (from our provided defaults) for every player in the field, set the tournament parameters (number of rounds, cut rules), and add placeholder golfers if our field is incomplete. This page will allow power users to find their own edges in more obscure tournaments that we can't devote much time to.

Expected value on each-way bets is now available for the books that offer them. Given that each-way terms typically differ across bookmakers, and that in general it is hard to set terms that work well for the entire field, this should highlight some unique plus-EV betting opportunities!

Price histories for each player across all books can now be viewed by clicking the button to the right of each row. We highlight every historical price change: orange means the price has shortened (decreased) while blue means it has drifted (increased).

This page leverages shot-level data to provide a comprehensive statistical breakdown of every Shotlink-equipped course on the PGA Tour since 2015. Users have the option to include up to 31 stats in the course table; these range from the basic, like course scoring average, to the more complex, like how difficult putts inside 5 feet are relative to the average course. Easy sorting and filtering functionality cap off the page, giving power users the ability to directly compare courses across many dimensions in a single webpage.

Those of you familiar with our (now deceased) blog home page know that it is was unquestionably one of the worst webpages on the internet (competing with the PGA Championship leaderboard). While part of me is sad to see the page go, the new version's simple layout and functionality should make finding specific blog posts much easier. We have added "tags" to each post; by clicking a tag listed in the sidebar you can quickly filter to related posts.

Since adding DG skill radar plots to the player stat pages, we've added the option to view past versions of each player. For example, shown below is Keegan Bradley's 2013 skill profile (orange) overlayed with his current skill profile (green). Needless to say the anchor ban has claimed at least one victim.

Ever wonder who our model thinks is the best putter on tour? Longest driver? Worst approach player? This page provides a full list of the best (and worst) golfers in the world at 6 different components of skill. Each rating value can be interpreted as a prediction of that golfer's next performance on an average PGA Tour course. API feeds and table downloads are available for Scratch users.