This is where the state’s refusal to be forthcoming really causes issues. We have to guess about how to match breakthrough cases with the data they occurred, so when we do that things like the Labor Day decline in testing and school opening rises in testing, as subsequent changes in “case” levels, screw up those methods. The state has the date data but won’t release it. And I suspect they are in no hurry to identify vaxed cases. They don’t want people to see the truth about breakthrough events. In any event, while there appears to be a decline in proportion of breakthroughs, that is likely an artifact of our estimation method. Thanks again to Dave Dixon, who is working on just a version that shows proportion of reported events, which isn’t as meaningful.