Jump to content

BenFairfax

Regulars
  • Posts

    299
  • Joined

  • Last visited

Everything posted by BenFairfax

  1. Is this not the perfect time to slim down MLC ranks? At very least postpone appointments and see if anyone notices.
  2. Not going forward, IoM students now pay same 9.2K.
  3. I agree with your earlier post, with our geography being the main reason for this. IoM was certain a good place in see out the pandemic. Regarding the government handling, two issues I have are: (i) AFAIK the four deaths resulting from the SPC outbreak did not even receive a public apology from IOMG, and AFAIK no one was even sacked, Ian Kermode provides more details at: https://mmo.aiircdn.com/61/604b6f9087a3b.pdf (ii) With the roll out of the vaccination program. I was expecting a certain amount of (for want of a better term) fraud, but I cannot tally for example group 4, to be even close to the reported 4K+ in size. Both items put the IoM in a bad light.
  4. Replied to @wrighty detailed as a household we went into lockdown on 26th July until 2nd Aug. Others took similar steps, and some just continued as before. Each to their own.
  5. I mentioned in ideal world R(t) (R function of time), but then you end up with non-linear system of which implementation becomes lot more demanding. You end up having to use Monte Carlo simulation. However as I mentioned I could smudge this and get same effect with R=1.1 at Day 37. An even more refined model would be if R was a function time, cases, acute disease, fatalities. With models they form part of the public messaging, or at least should and a coherent narrative will need to be presented to the public. Regarding the message to 'crack on', I imagine that was the direction of Public Health England gave local representatives regarding message to be repeat to the population, to get the level of infections needed for Herd Immunity. Not saying that right or wrong, just that makes perfect sense for the programs messaging. The reactions are a range on the ground, but that was case with all the waves, excluding first one. What I can say for sure if we went into strict lockdown 26th until 2nd; in accordance with our agreed mitigation to the events occurring in the IoM.
  6. Total agree and regarding messaging IOMG should of treated us like adults and said: We going for a natural wave with mass infection in young, 2+2s please keep out the way (particularly high-risk ones) because 2+2 infections add nothing to 2+2 equivalence required for Herd Immunity.
  7. I always considered the hospital admissions as the only data point I could reasonably believe was a realistic reflection of reality. I completely agree with your points regarding testing and what exactly is in it for the individual to go for a PCR test, rather than just doing decent thinking and isolating until (at least) return negative LFT. Once COVID+ fully resting also reduces the likelihood of developing Long COVID. Regarding the peak, using SEIR model framework possible to get peak a few days after 31st, but soon the wave just runs out of hosts and peaks because of this rather than in part lack of hosts and in part change in behaviour. Now just in terms of model, very very unlikely peak not past, do not take my work for try moving parameter around: https://sites.google.com/webcabcomponents.com/seir-model-of-iom-natural-wave/seir-model-live?authuser=0
  8. They were used as a smoke screen, and the plan was fixed in stone and came down the pipe from Public Health England, as run in British Virgin Islands, Jersey, and soon the UK. Choice was vax kids or mass infection in kids with some spillover, JCVI choose mass infection.
  9. I knew that the case numbers we would see, had good idea when acute outcomes would occur, and by considering past waves behaviors, on 8th July nailed in colors to the date 31st July (and stuck with it). The government I assume at direction of Public Health England were happy to ramp this wave, as much as possible, but clear that at some point particularly when we had acute outcomes people would freak out. The freak out date, naturally ties into the peak of exposure. Once you got a data, next you want to estimate a level. @wrighty told me my 'effective R =1.1' implied a lockdown lite, which I think is roughly what we have. I was expecting a big reaction because of the scale of the wave by 31st. After going around trying to grasp peoples reaction I think 1.1, may underestimate the reaction. In finance you measure freak out level, or the happy/depressed type of metrics, which feed into other things, but here I was faced with trying to predict the level of freak out. So I guess in part at least purely subjective estimate, I was expecting binary response, one minute happy to mix in crowds, the next 50% population gone into lockdown.
  10. I did not, was purely thinking in terms of risk/reward on offer now. Since we have 3 week gap for Biotec for Care Home population, they now also face wane in the vaccine's efficacy, 2nd jab over 6mths ago, at least according to this study: https://www.medrxiv.org/content/10.1101/2021.07.28.21261159v1 The CM favourite "doomed if you do and doomed if you do not", likely applies here.
  11. My mate a fan of Ashford, what amazes me is he always has an answer, and just imagine the unbelievable amount grief he has had to put up with for 16 months. When he signed up, must have thought job will not be that bad, get usual moaners but that part of job, and then COVID hit.
  12. I said we are all 'attention seeking freaks', with celebrity you also have knock on effect of fandom. Where people with no personal connection to the rock star will defend and support them without question. But celebrity is two-sided coin, you get attention (at least short term until next boy/girl band turns up), thereafter you tour small venues endlessly playing songs you wrote 20 years ago, to overweight middle-aged people trying desperately to relive their youth. The Rock Star cannot escape the image the fans project onto them, they become a caricature of themselves, and forever unable to age with grace and dignity. In this context the IoM celebrities will pay the price by forever being frozen in time with their fans demanding endless renditions of their 'COVID Songs', be that waiting in the bank, or Tesco or just sitting on the boat.
  13. We are all victims of Social Media algorithms, even MF has a like button. If MF wanted to focus on the message, rather than turn us all into 'attention seeking freaks' then it would remove the like button. So whether we are, or are not attention seeking freaks, such behavior is being endorsed by MFs. There are now very few places on internet where the message is what counts rather than the celebrity. I get sucked into equating the value of someone's views with the number of followers/likes etc Sorry, bit off topic, but I miss the days when the message was all that counted.
  14. My thinking is at aggregate level you can compare the peak exposure with peak bed occupancy. In Jersey, the difference was 8 days. Which implies mostly young admitted, if same applies here peak in 4-5 days. However IoM bed occupancy is running at a higher rate (per exposure) and would imply more seepage from mass infection in young to older groups (or younger with comorbidities). As discussed in effect unvax regarding risk just scales your age up (i.e. can cannot distinguish between 50yo unvax and 80yo vax). Though reverse out all comorbidities etc, is hard going, we can consider inverse problem at aggregate level. You know I made sure had bit of slack with days in hospital at 9.4, and was thinking when closed nursing homes, would cause Nobles problem with discharging. But my 2 cents is Manx Care did the right thing, where we are now. [As side note: Going back to model you know I put in 'effective R = 1.1' on Day 37 (July 31st), reflecting expectation of people "freaking out", and going around South and how risk averse people are here, we have lockdown lite, and thinking R=0.7 be closer. Just thought throw that in. Here is model in that case, and assume school period more ramped in this case: https://epcalc-ten.vercel.app/?CFR=0.000&D_hospital_lag=6&D_incbation=5.2&D_infectious=2.9&D_recovery_mild=11.1&D_recovery_severe=6.00&I0=1&InterventionAmt=0.2&InterventionTime=37&P_SEVERE=0.0123&R0=3.65&Time_to_death=32&logN=9.601368434322 ]
  15. Main thing uncertain off in the model when put together 8th July, was the degree of "freak out", I knew sorts of numbers we would see people would jump issue was I did not know but how much. Originally, I put in 'effective R = 1.1' from 3.35, but now after cycling around and see very few out I am thinking more like 'effective R = 0.73' now, almost like a lockdown for older groups (at least), here are the number is that case: https://epcalc-ten.vercel.app/?CFR=0.000&D_hospital_lag=6&D_incbation=5.2&D_infectious=2.9&D_recovery_mild=11.1&D_recovery_severe=6.00&I0=1&InterventionAmt=0.2&InterventionTime=37&P_SEVERE=0.0123&R0=3.65&Time_to_death=32&logN=9.601368434322 Nobles numbers sadly will be worse before get better, and so 'effective R = 0.73' likely continue for a bit, and then revert back. The net effect is tail be shorter, wave collapse faster.
  16. I could not agree more, all people need to know is the following, and people can choose what is right for them and their family. I suggested to my 76yo father to 'Netflix & chill' for next few weeks and my 10yo daughter is going on bus each day (in a mask naturally to protect others) to a holiday club party mayhem for next 3 weeks. Which is the right setup for our situation.
  17. Sorry I forgot to say, you can pass the model parameters within URL, here is my original calibrated SEIR Model of the IoM Wave: https://epcalc-ten.vercel.app/?CFR=0.000&D_hospital_lag=6&D_incbation=5.2&D_infectious=2.9&D_recovery_mild=11.1&D_recovery_severe=6.00&I0=1&InterventionAmt=0.33333333333333337&InterventionTime=37&P_SEVERE=0.01&R0=3.35&Time_to_death=32&logN=9.601368434322 Where the parameters are: CFR= Case Fatility rate D_hospital_lag=Time to hospitalization D_recovery_mild=Recovery in days from mild disease InterventionTime=Days after Day 0 when intervention occurs. InterventionAmt=The percentage of the shifted 'effective R' in relation to initial R. logN=Natural Log of the starting susceptible population IO=Number of initial infections D_incbation=Days after exposure in incubation. D_infectious=Days after incubation are infectious. D_recovery_mild=Recovery in days from mild disease InterventionTime=Days after Day 0 when intervention is taken. logN=Natural Log of the starting susceptible population IO=Number of initial infections When modeling I do not fit the model to data. What I did was try to understand through the SEIR Model framework the fundamental drivers of the IoM wave. Based on the properties of the Isle of Man social networks, expected societal reactions; in addition to estimates of organic properties of the virus and known demographics, I calibrated the SEIR model. If the model diverged significantly from the real-world situation, I would not shift parameters to fit the data, but return to my original assumptions and try to understand what conceptual error I had made or influences I have missed. The purpose of the model is not only to provide predictions for the cases, hospitalisation and duration but also to get a handle on what is driving these numbers.
  18. The real world rate we are experiencing now is running a little over the number I would expect from 1.23% at this stage from the SEIR Model. Though I phrased model in terms of unvaccinated*, which as we know is only thing which counts regarding getting to Herd Immunity, I was anticipating some seepage to vaccinated. My warnings to high-risk 2+2 to take mitigations has been repeated even incessantly for several weeks now. I assume in IoM based on UK data 'because COVID' admissions are 75% of the total 'with COVID' cohort (i.e. 25% COVID admissions are for other reasons). Think you guys in circumstances are doing a great job, newspaper presented a controversy (fair enough their call) but I see difference in view as move of a part of healthy discussions. [* I smudged public model because I choose to not be in a position were I ended up talking about mortality risk.] Here is the latest version of the banner, enjoy:
  19. Regarding the vaccine wane I attach below, from: https://acmedsci.ac.uk/file-download/4747802 Royal society judgement in central case natural immunity last a year (versus 3 years for vaccine), so unless kids vaccinated they will likely go through the same process in a years time. I am concerned about rise of 70+ cases and as you say nearly all 2+2, the total number 70+ active COVID increased by 10 today. Really not good, everything should be done to prevent this group from becoming infected during this wave. In terms of reaching herd immunity a notion implied from SEIR Model there is no advantaged in additional infections of 2+2s.
  20. Not the aim, in few weeks the situations improves, there is light at the end of tunnel. I have not read article, but I hope it gets the main messages out, the 2+2 High-Risk ideally should try to take mitigations and avoid high-risk setting for the next few weeks.
  21. The information to noise ratio does not warrant a reply.
  22. And it don't work. I fixed the banner, thanks for that, whether agree with message or not good to have some sort of minimal standards. That is why I keep going on about SATS, Ofsted and Ofqual; so next generation do not end up with my gramma.
  23. I meant real physical people I know in person in the real world. Half MF, Manx Twitter are anonymous. Why does MF have the social media style "likes" button? The point of forums is credibility is based on the content, not number of followers you have or number of "likes" etc. The social media thing sort of OK, but you end up at the lowest common denominator, whether or not that has any value.
  24. I mean in the physical world, not people on the internet, were decent % are anonymous (100% IoM anti-vaxers are anonymous).
  25. We all vaccinated, and think everyone I know in meat world is. From ONS data, it reduces risk hospitalisation by ~20 times, i.e. someone with 1% risk pre-vax hospitalization (1 in 100), post 2+2 it 0.05% (1 in 2,000), not zero but not something to really be concerned about. But if you are say Blood Cancer patient with pre-vax risk of 40% hospitalization, then after 2+2 you still talking 2% (1 in 50) and it up to the person, to decide, however someone at these sorts of odds with prevalence of COVID in community we have, should (in my view) be informed. At least UK PH thinks so: https://www.gov.uk/government/publications/guidance-on-shielding-and-protecting-extremely-vulnerable-persons-from-covid-19/guidance-on-shielding-and-protecting-extremely-vulnerable-persons-from-covid-19
×
×
  • Create New...