How good is it, really, and how sure are we?
One standardized verdict per drug, graded against its own division, so a best-in-class pill is not punished for not being an injection. Every efficacy number is placebo-adjustedPlacebo-adjusted: the drug’s result minus the placebo group’s. Every trial runs a diet-and-lifestyle program in both arms, so the placebo group already loses a few percent. The drug’s real effect is the gap above it., uses the figure that counts everyoneCounts everyone: obesity trials report two numbers. We always take the conservative one, which counts every person randomized, including those who stopped the drug. Sponsors name it the “treatment policy” estimand (Novo) or the “treatment regimen” estimand (Lilly and Boehringer); we verify each trial’s analysis plan. We never headline the flattering on-treatment number. prescribed the drug rather than only those who stayed on it, and is time-standardized. The grade only counts what the evidence can back: a big press-release number never outranks a confirmed one.
Reductive on purpose. Rigorous underneath.The grade: a tier, within division
Confidence: how sure we are
One bar, a hard zero
This strip is the quick key. The methodology page defines every threshold and is linked from every drug and comparison page.
The board
% body weight lost vs placebo · everyone counted · monitored for new evidenceInjectable
frontier ≈21%5 ratedSurvodutide is the cautionary tale: the 16.6% headline is the efficacy estimand. Once everyone who stopped is counted it is 7.6%, and about 1 in 5 stopped for GI side effects. Certified evidence, lagging verdict.
Oral
frontier ≈11%2 ratedAgainst the hype: oral semaglutide out-does orforglipron on both efficacy and tolerability. Orforglipron’s real case is convenience: no fasting, room temperature, scalable. See the head-to-head.
Efficacy bars show % of body weight lost vs placebo, so longer is better. Bars share one absolute 0 to 25% scale across both divisions, so an oral and an injectable at the same length really did lose the same amount. Each division’s frontier is its strongest drug, the top bar, which is what anchors the A tier. * Retatrutide’s figure is inferred from topline and ADA, not yet peer-reviewed.
What could move the board
the next reasons a verdict changesOn the horizon
projected only · kept off the ranked boardThese are not graded on the same scale. The evidence is still Phase 2 or topline. The faded, dashed-ring grade shows only where each looks to be aiming, at low confidence. When one earns a Phase 3 readout, it graduates onto the board.
Recent changes
the log is seeded with real historyGet an email when a verdict changes
A few a year, no more. When a drug graduates to the board, gets approved, or moves a tier, you hear it here first.