Streamers, August 11

51 named starters over the next 2 days, ranked by skill, matchup and park.

51 starters are named for the next 2 days. This is a single-start question, not a volume one: who is pitching, who they face, and where. The score below combines the pitcher's SIERA and strikeout-minus-walk rate with the opponent's runs per game and the park's run environment, each scaled to the spread across tonight's actual slate, so it ranks THIS board and is not an absolute rating.

Best plays

Sit these

Bad skills, a good offense, a hitter's park, or all three. Streaming is a choice, these are the starts where the choice is to pass.

What this means

Kyle Harrison at Petco is the cleanest add on the board, and it is not close. A 2.92 SIERA paired with a 25.2 percent strikeout-minus-walk rate is ace input, and the two things that usually ruin a good arm's night are absent: San Diego is scoring 4.22 runs a game and the park plays at 1.02x, essentially neutral. Dylan Cease scores a point lower on better strikeout skill, 26.1 percent, because Boston is the tougher lineup at 4.44 and he has no park help either. If both are available, take Harrison. If you need strikeout upside above all, Cease is the same bet with a slightly worse floor.

The name I would leave alone despite the 82 is Drew Anderson. His score is built almost entirely on context: Cleveland at 3.97 runs a game is one of the softer offenses here, and Comerica at 1.04x is close to a wash. Strip that out and you have a 3.28 SIERA with an 18.3 percent K-BB, which is fine and nothing more. Compare it to Zack Wheeler, who carries the same 82 while pitching in front of a 4.37-run St. Louis lineup in a park at 1.09x. Wheeler earned that number through a 24.1 percent K-BB. Anderson was handed his. In a two-day window, the pitcher whose score comes from skill is the one whose bad process still ends in five strikeouts and three runs; the pitcher whose score comes from a quiet opponent has nothing to fall back on when the quiet opponent hits.

The counterargument is real and worth stating properly. Streaming is a matchup game, not a talent game, and the whole point of a context-adjusted score is that it does not care where the edge comes from. If you are streaming for ratios rather than counting stats, Anderson against a 3.97-run offense is exactly the profile you want, and Wheeler in a 1.09x park is the one who blows up your ERA. That case wins if your team is chasing ERA and WHIP and is already fat on strikeouts. Mine wins if you need punchouts, which most streamers do.

At the bottom, I would not talk myself into Chase Field. Sugano and Feltner both land at 41 for the same reason: mid-fives SIERA, single-digit K-BB, a 4.55-run opponent and a 1.10x park. There is no hidden edge there, just three multipliers pointing the same direction. Merrill Kelly at 35 is the one people will argue for, because the opponent is Colorado and the instinct is that any arm beats the Rockies. The data does not support the instinct: Colorado is the highest-scoring opponent on this slate at 4.77 runs a game, and Kelly's 5.8 percent K-BB gives him no way to escape trouble he creates. The 5.14 SIERA is the story, not the logo across the plate. Same problem, worse, for Mitch Bratt at 25 with a 2.5 percent K-BB against that same lineup.

George Klassen's 19 needs a caveat. A 7.81 SIERA and a negative 9.6 percent K-BB, meaning he has walked more men than he has struck out, is either a broken pitcher or a sample too small to mean anything. This data cannot tell you which, and I will not pretend otherwise. Either way you do not start him against Texas in a 1.08x park.

So: add Harrison, use Cease if you have him, and take Wheeler over Anderson if you can only roster one 82. The falsifier on Anderson arrives fast. If he goes six with seven strikeouts against Cleveland, the context-only read is wrong and I would happily ride him again next turn. The cost of being wrong on him is one start's ratios, which is why I would rather be wrong holding the better arm.

How to read this

SIERA is the ERA estimator that survives a small sample best: it strips out defense and sequencing, so it describes the pitcher rather than the night he had. K-BB% is the same skill seen a second way, and it is the part that holds up when the defense behind him does not.

Opp R/G is the opponent's runs scored per game from their own record, shown as a rate rather than a rank so the size of the gap is visible, the difference between the best and worst offense on a given night is about a run and a half. Park is the venue run-environment multiplier, where 1.00 is neutral and higher favors hitters.

Starters are only announced a day or two ahead, so this covers a short window and skips games where the rotation has not been posted rather than guessing at them.

All articles · Redraft rankings · Dynasty rankings