RT @MLB: Trevor Rogers and Freddy Peralta are your AL and NL Pitchers of the Month for August!

RT @Tstokey: Softball assessments today 👀. @DrivelineBB @ogxsoftball

Join millions who have switched to Grok.

RT @brettccook: Quite possibly the most unique pitch I've seen. Knuckleball grip with downward action. At times, up to -20 inches of ve….

Was there ever a doubt?

Lance has been preaching it all year. Great breakdown on why he's right!

cc @srbrown70 @bymichaelrosen.

ICYMI here are the slides from the @DrivelineBB presentation on Projecting Pitch Design at Saberseminar!.

Awesome piece on Mize. @m_kress__ and I spent a lot of time this offseason talking through arsenal effects and what would fit him best. Much easier to talk through possibilities than implement them, but Mize has killed it. Breaking RV from 6th to 68th percentile from '24-'25.

Recreating this "full signal" path isn't perfect, but when working on finding actionable changes it adds an important layer to the public metrics. For more info check out @connerpelletier and @SamEhrlich's presentation!.

Our model suggestions:. Mullins could benefit from being slightly more oppo deep, then finish with a more pull side AD and steeper AA. Moniak could benefit from having a higher AA throughout the swing, while working slightly more oppo in the follow through.

Take the example of Cedric Mullins and Mickey Moniak. Mullins: 11° AA, 6° AD.Moniak: 11° AA, 3° AD. Very similar average metrics, but our path optimizer suggest very different changes.

Average metrics don't tell the whole story. When it comes to training players, it's key for us to understand how guys enter the contact zone, where they're at when they're making contact, and how they finish out in front. Descriptiveness matters with this data.

When this is done we've essentially reconstructed the average path of the barrel. Furthering the assumptions already made, we can add a bat perpendicular to the velocity vector at the player's average swing path tilt.

Our process:. 1. Pick a starting point .2. From that starting Y, find the average attack angle and attack direction.3. Expect the barrel to move in that vertical/horizontal direction for 1 inch at their average bat speed.4. Repeat 2-3 while within the bounds of observed Y values.

So when we have this data, how can we use it?. The above plots are not very interpretable, BUT we can leverage it to make things easier to track. Given a set starting point in 3D space where Y is within the bounds of observed data, we can visualize the avg bat path of a batter.

These measures are heavily tied to depth of contact. This is pretty intuitive, as the swing is a very fast, violent motion that changes very quickly. Given enough sample, the individual relationships for a player become clear. Here is Seiya Suzuki's 2025 first half.

It all starts with understanding what attack angle and direction are measuring. These measures are getting at the angle of the velocity vector in the vertical/horizontal direction when the ball and bat collide or cross paths.

Want to recreate bat path estimations like this?. The math behind the basis of @connerpelletier and @SamEhrlich's killer Saberseminar presentation on optimizing bat path⬇️🧵

RT @srbrown70: For SaberSem I studied how batters track pitches, with what level of uncertainty, to better understand their swing shapes an….

RT @SamEhrlich: Another sabersem in the books. Had a great time listening to presentations, meeting everyone, and presenting for the first….