Today's matchup
Leo draws Los Angeles Dodgers on the road today. He shows up on 12 boards — doubles (31.8/100, #11), h+r+rbi (31.1/100, #15), hits (30.6/100, #15), runs (28/100, #15). We've got him at 0.2 on the doubles board. The model doesn't love the spot — the numbers are below. Open any board row for the full breakdown of what's driving it.
He bats from both sides
Leo is a switch-hitter, which changes how the model reads every one of his matchups. Most hitters carry a real platoon penalty — a lefty bat facing a good lefty arm is a materially worse spot, and the score drops accordingly. Leo takes the favourable side almost every night, so he doesn't collect that penalty. What moves his number instead is the quality of the arm rather than its handedness, and how the two sides of his own swing actually compare, which is rarely as balanced as the switch-hitter label suggests.
Where he lands across our boards
Leo was scored on 12 boards in that matchup, and the spread is the interesting part: doubles graded 31.8/100 while triples came in at 13.6/100. Same player, same game, same opposing arm — different questions. A spread that wide usually means the matchup favours one kind of outcome over another, and it's the clearest argument for reading the board that matches what you actually care about.
How the model reads him
Leo gets scored the same way a backup infielder does — same inputs, same scale, no thumb on it. What he's done against arms of this handedness, what this particular pitcher and bullpen give up, his likely plate appearances, the ballpark, the conditions, his recent form. The weighting behind those is ours and gets re-tuned weekly against what actually happened. The output is public and graded, which is the only claim worth making.
What we'd flag
Read this page knowing what it can't do. It can't show you a line no book posted — coverage peaks around 87% for confirmed starters. It can't see a lineup change that happens after the projection publishes. And it can't tell you anything useful from one game, because the model is only measurable across a season-sized sample. For Leo specifically, we don't have a meaningful graded record on him yet.