Projections

Daily Pitcher Strikeout Leaders

The first Mithrandir Metrics projections surface. This page turns the existing V1 model output into a daily leaderboard with just enough supporting context to explain why each arm rates where it does.

Latest Projection Date

2026-04-09
pitcher_k_v1_regular_16f 11 starters

Overview

Projection Snapshot

Leaders sorted from the same saved artifact the projection pipeline writes each run.

Projected Starters 11
Training Starts 10050
Skipped 1
Model pitcher_k_v1_regular_16f

Leaderboard

Projected Strikeout Leaders

Pitcher Matchup Proj K Pitcher K% Last 5 K% Opp K% 14d Outs / Start Rest Starts
Jack Flaherty headshot
Jack Flaherty
Probable starter
DET vs MIN 5.83 27.1% 25.6% 22.4% 15.2 5 1
Nolan McLean headshot
Nolan McLean
Probable starter
NYM vs AZ 5.77 27.3% 30.7% 22.4% 16.4 6 2
Max Meyer headshot
Max Meyer
Probable starter
MIA vs CIN 5.10 23.0% 19.5% 22.4% 15.6 5 2
Seth Lugo headshot
Seth Lugo
Probable starter
KC vs CWS 4.93 21.4% 22.0% 22.4% 16.6 5 2
Ryan Weathers headshot
Ryan Weathers
Probable starter
NYY vs ATH 4.83 23.7% 24.0% 22.4% 14.4 5 2
Eduardo Rodriguez headshot
Eduardo Rodriguez
Probable starter
AZ vs NYM 4.52 19.3% 18.4% 22.4% 16.4 6 1
Jeffrey Springs headshot
Jeffrey Springs
Probable starter
ATH vs NYY 4.37 22.4% 21.9% 22.4% 16.0 6 1
Rhett Lowder headshot
Rhett Lowder
Probable starter
CIN vs MIA 4.23 21.7% 19.0% 22.4% 15.8 5 2
Mick Abel headshot
Mick Abel
Probable starter
MIN vs DET 4.01 20.6% 20.8% 22.4% 14.1 5 1
Anthony Kay headshot
Anthony Kay
Probable starter
CWS vs KC 3.98 22.2% 22.6% 22.4% 15.4 10 1
Randy Vásquez headshot
Randy Vásquez
Probable starter
SD vs COL 3.36 15.2% 18.2% 22.4% 15.4 5 1

Context

What V1 Shows Publicly

Only fields already produced by the current model output.

Pitcher Skill

Season and Last 5 K Rate

These explain whether the projection is being driven by the pitcher’s current strikeout baseline or more recent form.

Opponent Context

Opponent K Rate Last 14 Days

This is the cleanest public-facing opponent context from the current feature set without overloading the table.

Workload Context

Outs Per Start, Rest, Starts

These help frame expected length, freshness, and how stable the current-season sample is for each pitcher.