mission27 Posted May 4, 2020 Share Posted May 4, 2020 1 minute ago, mistakey said: remdesivir wont save lives, just keep hospital throughput up. its a decent start but its not anywhere close to being the solution. all your numbers are meaningless so #1 literally is thrown out NY hospital and ventilator capacity certianly wasnt. it wasnt as terrible as it was projected to be but it was in no way sufficient. uh, i guess you can take credit for something that everyone else thought? There were ventilators in New York that went unused. I'm not saying it was inappropriate to ramp up ventilator production and stock up. But the fears that NYC would be Bergamo and we'd be pulling elderly people off vents because we needed them for younger people didn't really come true. And the CFR thing is indisputable at this point. Every single antibody study has suggested CFR in the range MoL was saying from day 1. It isn't our numbers, it California's numbers, New York's numbers, Germany's numbers, etc. 1 Link to comment Share on other sites More sharing options...
mistakey Posted May 4, 2020 Share Posted May 4, 2020 Just now, mission27 said: The death rate has slowly and steadily declined for the past 2-3 weeks as has the % of tests coming back positive. What explains this, if we have not reduced R0 below 1 in much of the country? what explains this is something that i will admit i was wrong/hasty on in the beginning. the epicenter of hte diesease is obviously in NY/NJ. it has gotten better in terms of caseload here in NY. this is also extremely observable due to less sirens. it is getting worse throughout the country and will continue to get worse. the only thing you could "slightly" say that it is getting better in NY going forward is the % of the city that has already gotten the disease and will have short-term immunity, and potentially a supply of PPE/remdesivir to help out the hospitals in the upcoming uptick of cases. + some epxerience by the doctors. i dont think this will be sufficient enough to keep the death rate in succeptible populations down. Link to comment Share on other sites More sharing options...
mission27 Posted May 4, 2020 Share Posted May 4, 2020 11 minutes ago, ramssuperbowl99 said: I think the most likely scenario is the one where mission doesn't have a model. We dont have a forward looking model. We have a model for tracking point in time snapshots of how quickly the disease is spreading in a given geography, at a given time. The MoL model makes no predictions about future MoL scores or the impact of various policies on future MoL scores. Its similar to looking at daily case growth rate, just more sophisticated and sexy. If your MoL gets low enough you start to relax restrictions. If your MoL then spikes too high, you re-introduce some restrictions. Common sense. Link to comment Share on other sites More sharing options...
mistakey Posted May 4, 2020 Share Posted May 4, 2020 3 minutes ago, mission27 said: There were ventilators in New York that went unused. I'm not saying it was inappropriate to ramp up ventilator production and stock up. But the fears that NYC would be Bergamo and we'd be pulling elderly people off vents because we needed them for younger people didn't really come true. And the CFR thing is indisputable at this point. Every single antibody study has suggested CFR in the range MoL was saying from day 1. It isn't our numbers, it California's numbers, New York's numbers, Germany's numbers, etc. uh nys deaths = ~25k italy deaths ~= 29k deaths Link to comment Share on other sites More sharing options...
ramssuperbowl99 Posted May 4, 2020 Share Posted May 4, 2020 8 minutes ago, mission27 said: We dont have a forward looking model. Which means you don't have a model. 1 Link to comment Share on other sites More sharing options...
mission27 Posted May 4, 2020 Share Posted May 4, 2020 1 minute ago, mistakey said: what explains this is something that i will admit i was wrong/hasty on in the beginning. the epicenter of hte diesease is obviously in NY/NJ. it has gotten better in terms of caseload here in NY. this is also extremely observable due to less sirens. it is getting worse throughout the country and will continue to get worse. the only thing you could "slightly" say that it is getting better in NY going forward is the % of the city that has already gotten the disease and will have short-term immunity, and potentially a supply of PPE/remdesivir to help out the hospitals in the upcoming uptick of cases. + some epxerience by the doctors. i dont think this will be sufficient enough to keep the death rate in succeptible populations down. Tbh I dont think you are too far off. Things have gotten much better in the tri-state area and a few other spots and I would say have hit a plateau in the rest of the country. This is trailing 7 day death rate: But we always expected places would peak at different times and increased testing and changes in death reporting criteria are certainly contributing to some of this. Its just disheartening because we seemed to be turning a corner even in these places 5 or 6 days ago and have seen some less than ideal numbers over the past week. But we need to give it a few more days or weeks to see how these numbers develop. 5 days of less than ideal numbers in a few states is not the end of the world especially when those states are dramatically increasing testing. Link to comment Share on other sites More sharing options...
mission27 Posted May 4, 2020 Share Posted May 4, 2020 4 minutes ago, ramssuperbowl99 said: Which means you don't have a model. So there is no such thing as a model of the past? Link to comment Share on other sites More sharing options...
ramssuperbowl99 Posted May 4, 2020 Share Posted May 4, 2020 Just now, mission27 said: So there is no such thing as a model of the past? Nope, because unless you can predict, you don't know it works. 1 Link to comment Share on other sites More sharing options...
vikesfan89 Posted May 4, 2020 Share Posted May 4, 2020 13 minutes ago, mistakey said: uh nys deaths = ~25k italy deaths ~= 29k deaths What's the population difference? Link to comment Share on other sites More sharing options...
mission27 Posted May 4, 2020 Share Posted May 4, 2020 Just now, ramssuperbowl99 said: Nope, because unless you can predict, you don't know it works. Your definition of "model" is extremely narrow I would say what you are describing is a "forecast" A model is any simplified representation of a real world system Link to comment Share on other sites More sharing options...
mission27 Posted May 4, 2020 Share Posted May 4, 2020 14 minutes ago, mistakey said: uh nys deaths = ~25k italy deaths ~= 29k deaths I dont really see how this addresses my point Half of NYS lives in NYC. Less than 3% of Italy lives in Milan. This is a disease that kills people in cities, primarily. My point wasn't that NYS wasn't hit hard, my point was they had excess ventilator capacity and nobody was denied care. So optimism that we could handle the surge in cases, unlike Milan or Bergamo, was IMO well founded. Link to comment Share on other sites More sharing options...
ramssuperbowl99 Posted May 4, 2020 Share Posted May 4, 2020 2 minutes ago, mission27 said: A model is any simplified representation of a real world system Which requires proof that the next batch of your results will be in line with a real world system. Which you don't have. Not a model. Link to comment Share on other sites More sharing options...
mission27 Posted May 4, 2020 Share Posted May 4, 2020 4 minutes ago, vikesfan89 said: What's the population difference? The vast majority of Italy's coronavirus deaths have been in Lombardy. Lombardy has 10m people and 14k report coronavirus deaths. New York state has 20m people and 19k reported coronavirus deaths. I would argue the hardest hit regions of Italy were hit harder than NYC / NYS but obviously there are large swaths of the south of Italy that had minimal cases. Link to comment Share on other sites More sharing options...
mission27 Posted May 4, 2020 Share Posted May 4, 2020 1 minute ago, ramssuperbowl99 said: Which requires proof that the next batch of your results will be in line with a real world system. Which you don't have. Not a model. Thats called a forecast 1 Link to comment Share on other sites More sharing options...
TVScout Posted May 4, 2020 Share Posted May 4, 2020 Models exist to enable prognostication. Link to comment Share on other sites More sharing options...
Recommended Posts