Today's matchup
Rafael is hosting Boston Red Sox on the Saturday, August 15, 2026 slate. He shows up on 12 boards — rbis (58.8/100, #68), h+r+rbi (53.3/100, #115), hits (46.7/100, #121), total bases (43.9/100, #122). We've got him at 0.5 on the rbis board. Open any board row for the full breakdown of what's driving it.
Season line
0 starts, 0-0, 0.00 ERA, 2.00 WHIP, 1.0 innings. He's struck out 0 over that span, against 0 walks. This is the baseline the matchup adjusts from — the model starts here and then asks what the specific lineup he's facing does to strikeout arms, how the park plays, and how deep he's likely to go.
He throws right-handed
Rafael throws right-handed, which is the ordinary case and means fewer wild swings in the lineup he faces. Opposing managers aren't restacking against him the way they would a left-hander, so his projection moves more on the quality of the lineup than on its handedness. What does move it: how deep he's been going lately, and whether the lineup in front of him strikes out.
Where he lands across our boards
Rafael was scored on 12 boards in that matchup, and the spread is the interesting part: rbis graded 58.8/100 while stolen bases came in at 6.2/100. 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
Every arm on the slate goes through the same math, Rafael included. Opposing lineup, handedness, expected batters faced, park, weather — weighted, combined, and normalized 0–100. We keep the weighting to ourselves and re-tune it weekly on real results. What we don't keep is the record: every projection here is graded against the box score at the line the market offered.
What we'd flag
Three caveats worth carrying. Not every player gets a posted line — book coverage runs to about 87% for confirmed starters, so a blank line is the market's silence, not ours. Lineup changes land late enough to strand a projection published hours earlier. And judging any of this on a single night is meaningless; the model is tuned on hundreds of settled outcomes.