Concern #1: So far, SAVA's efforts to respond to this allegation, i.e. that it was Dr. Wang's lab, and not an outside independent lab which conducted the re-analysis of p2b data seem ineffective. -- Sava has stated that re-analysis was done by outside lab indeed, which makes sense given’s sava’s limited staff. Whatever associations this lab may or may not have with Wang, even if remote, are being used by accusers to suggest fraud. Their (implicit) rationale is that the association indicates fraud because Wang’s western blots done years ago look funky (an accusation). So, the accusation of fraud relies on a guilt-by-association allegation (not evidence of fraud), which in turn relies on accusation of fabricating several blots among dozens if not hundreds (again not evidence of fraud but instead accusation regarding an issue far-removed from most recent trials). Accusations based on other accusations constitute evidence of fraud? Not even close. $SAVA, Cassava Sciences, Inc. / H1 Concern #2: Quanterix, mentioned by SAVA in their original response, indicates that they are not the lab that analyzed the part of p2b data that was presented in a poster at AAIC. -- This old issue again. SAVA never stated that Quanterix had analyzed or presented the data. Instead, both SAVA and Quanterix had stated that the role of Quanterix was to generate/produce the data. Critics are mis-reading the press release sentences, and then making inferences that fit their narrative that sava is trying to cover up fraud (and an elaborate one) by shifting blame. Making inferences based on mis-readings is hardly convincing. Concern #3: Only the placebo data is the same, while treatment groups data differ between CTAD presentation slide 28 and Cinicaltrials.gov "8 secondary outcome"), as if they are two sets of results.Trade SAVA on T2BF. -- This is easily explained by clerical error (and consistent with sava’s position that clerical errors exist in reporting but not in actual results), because right below the table for spatial working memory in the clinicaltrials.gov website, the effect sizes are consistent with 100mg appearing at least slightly more effective than 50mg (0.46 and 0.25, respectively); in the case of effect sizes, larger is better. FYI, as a rule of thumb around .50 for Hedge’s g is considered a medium effect size, whereas .20 would be considered a small effect. So, the effect sizes are telling us that 100mg has a medium effect whereas 50mg has a small effect. Even if we were to entertain the possibility that the two effect sizes as reported in clinicaltrials.gov website are wrong and should be flipped around, this is less likely because its probability is a function of the multiplication of two events (i.e., P(x1) * P(x2)), where P is probability, x1 is incorrectly reporting the effect sizes on the clinicaltrials.gov website, and x2 is incorrectly reporting change in total errors (lower is better) on the CTAD presentation slide. In contrast, for the scenario where the incorrect reporting is limited to change in total errors reported in the clinicaltrials.gov table regarding the spatial working memory issue, its probability is a function of just one event (i.e., P(x3)), where P is probability and x3 is incorrectly reporting the change in total errors. No multiplication with another event’s probability required. Given such, it is more likely that P(x1) * P(x2) is greater or smaller than P(x3)? And by the way would masterminds of an elaborate fraud all these years really leave such trails of clerical errors? Concern #4: An additional concern is that any analysis of change between treatment and placebo appears to be compromised by inequivalent baseline measurements. Simply put, the baseline measurement (n=64) includes those who were not in the day 28 measurement (n=7=64-57) and therefore the analysis of treatment groups' changes from baseline is allegedly compromised. -- The difference of n=7 is already explained in the very same table for spatial working memory on the clinicaltrials.gov that the accusers are focusing on. Specifically, it states: “Removed from analysis were the 3 patients with no detectable drug in plasma, 2 patients with ≥25% non-compliance by pill counts, one patient with no baseline test and one who did not understand instructions per rater notes.” As one can see, 3 + 2 + 1 + 1 = 7. Concern #5: However, the Sep 3 responses, written and spoken, contain little to no "facts that can be evaluated and verified", but mostly general statements about AD drug development and CEO's views on stock market dynamics. -- This is not true. In the most recent sava press release, the CEO did in fact delve deeper into the facts and even made statements about clerical errors (e.g., Figure 5). Of course, some may want even more details, and details, and then even more precise details, to the point where they’re looking at individual data files and questioning every single cell of data. But this view/desire must be balanced with the fact that (1) the more involved sava gets, the less time and energy they will have left for actual p3 trials and other possibly ongoing developments (e.g., partnership); and (2) the more they reveal, the greater the fuel given to accusers (e.g., critics latching on the phrase “I think”). Statements about stock market dynamics (e.g., shorts, history of petitions being used as anticompetitive tactics) are entirely appropriate because, whether one likes it or not, this is not just science but also strategy/war. When one is a war, does one believe in any and every accusation hurled at their own team, or does one question whether that could be propaganda and exaggerations spread by enemies to weaken morale? The strategic context must not be ignored when evaluating the plausibility of the data fraud versus clerical error theories; contrary to the author’s view, this is not simply a purely scientific dispute. I personally would’ve preferred sava waiting until making a milestone announcement, but perhaps sava was concerned their silence might be interpreted as weakness. They were stuck in a rock and a hard place, and they chose to take a middle-of-the-road approach by giving another press release that offers more details, but not more.Trade SAVA on T2BF. Conclusion: If the petitioners and accusers want to convince the FDA or other key entities that fraud is the answer, they’re gonna have to try a lot harder than this. So far, all the accusations I’ve seen are quite circumstantial, superficial, and likely the result of Type 1 errors from engaging in fishing expeditions (i.e., combing through hundreds if not thousands of page of material in hopes of finding at least something remotely suspicious to make accusations). None of the accusations go in-depth to speak directly about fraud (where are the insider nonpublic unique information that a whistleblower would reveal?), and rather increasingly resemble a throw a spaghetti at the wall and hope something stick approach (“Look at these tables figures and associations, this can’t be just errors, they must be fraud please believe us!”). Sure, the supplement may raise new accusations (which are weak or explained away as described above), but keep in mind they also include laughable “technical” material such as desperate-sounding emails from a lawyer to the fda in a supplement to the petition. Now I don’t have a crystal ball, so yes anything is possible theoretically. But simply giving into fear and vivid imagery of a good scary story is the same mindset of some folks who never invest or try to constantly time the broad market because, practically every year or month, there’s a bear yelling that the market’s gonna crash for sure this year, and the next, and the next, and so on and so on. Rather than being emotionally sucked into panic created by the vivid stories and spicy accusations that tickle one’s innate fear/loss aversion (the intuitive rather than rational human mind is more sensitive to seemingly negative info), the rational thing is to continue updating the plausibilities of competing theories about a company (i.e., in this case, fraud versus clerical error) based on available facts and available facts only.