The Week in Review, August 5

The week's top producers from July 27 to August 2, 2026, led by Ceddanne Rafaela.

The seven days from July 27 to August 2, 2026, sorted by fantasy points actually banked. Points, not rates: a hitter with seven plate appearances and a .571 average had seven chances to do something, and a leaderboard that puts him first is measuring luck rather than production.

Batters

Pitchers

The week does not always survive contact with the projection

6 of the 20 players above sit below roster-relevant value on our board despite the week they just had. That is the useful column in these tables. A big week from a top-30 player is confirmation; a big week from someone the engine still will not rank is a sample, and the difference decides whether you are looking at a pickup or at noise.

What this means

Five home runs in 31 plate appearances is the loudest line on the board, and Ceddanne Rafaela's forward rank of 69 is the engine's way of saying it liked him before the week and likes him about the same now. That is the useful signal. A 0.419 average across 31 trips will not survive contact with June, but the two steals sitting under the five homers are the part that carries: Rafaela's fantasy value has always been the combination, not the power spike. Nobody is trading him to you cheaply now. If you own him, the correct move is nothing at all, and if you are shopping, wait for the average to normalize and the price to follow it down.

The week I would actually chase is Gavin Williams. Forty-nine points on 12.2 innings with 22 strikeouts is the rare big week where the components are the whole story rather than the luck, and a forward rank of 27 confirms the engine wasn't reacting to the box score. Twenty-two punchouts in two starts is a rate no defense and no sequencing luck produces for you. Chase Burns sits one spot ahead of him at 21 on 13 strikeouts in 11 innings, which tells you these two are already priced as front-end arms rather than as hot streaks. Hold both. The cost of selling Williams on a 1.42 ERA is that you sell the strikeout rate along with it.

Now the sells, and they are not close. Spencer Torkelson put up 30 points on eight RBI and a 0.423 average, and the engine ranks him 334th going forward. One home run in 29 plate appearances is the tell: this was an RBI week produced by the Detroit lineup turning over in front of him, not by Torkelson doing anything new. Gleyber Torres drove in nine with a 0.261 average in the same lineup, which is the same story from a different seat. Torkelson at 334 is a bench bat in most formats and a waiver-wire piece in shallow ones. The strongest case against selling is that first-base scarcity is real and eight RBI in a week is eight RBI regardless of how they arrived. Fine. But you are buying the lineup, not the hitter, and the lineup can change on a lineup card.

Walbert Urena is the starkest gap on the page and the clearest ignore. Twelve innings, twelve strikeouts, a 0.75 ERA, and a forward rank of 597. Almost nothing produces a number that far down except a projection built on a thin track record and a strikeout rate that does not support the run prevention: twelve in twelve innings is ordinary, and ordinary strikeout stuff running a sub-one ERA is the definition of a week that regresses. David Peterson at 431 and Colin Rea at 286 are the same shape, both of them 12 innings with fewer than a strikeout an inning. If you have a free roster spot in a deep league, streaming Urena against a soft opponent is defensible. Rostering him as a held asset is not.

Here is the test I would set. Urena's next two starts settle it: if he is still striking out roughly one per inning and the ERA climbs toward four, the engine was right and you drop him without a second thought. If he pushes past ten strikeouts per nine, the 597 was built on missing information and the market will be slower to notice than you are. For Torkelson, watch where he hits. Cleanup with runners in front of him keeps the RBI flowing; a drop to sixth or seventh kills the only thing holding his week up.

How to read this

Points are the window total from our own scoring, over the exact dates named above. Sorting on points rather than on rate stats means playing time counts for something, which is how fantasy actually works, a player who sat four of seven days did not help you, whatever he hit when he played.

Our rank is the engine's view looking forward, not a grade on the week. The two disagreeing is the interesting case, not an error: the week is what happened, and the rank is what we expect next.

All articles · Redraft rankings · Dynasty rankings