So why this site?
Most of us know this virus is not something to take lightly, but what really is the risk?? Ebola bad?! or, like a bad flu, or maybe nothing!? Plus considerations are multi-dimensional:
As to the risk/odds of contracting it. This depends on: the level of un-contained spread of infection around you; how many people you come in contact with; and, social distancing measures taken when you do have contact with others. What is the situation in my area? Is it safe (relatively speaking) to travel to an area? Is there a risk to others if I travel?
These are the questions I was curious about so set about doing some analysis to get the answers. Over time that developed into a spread sheet, then programming to update daily, and finally this website to maybe help others. This website's intent is to bring out some meaningful, current, measures and risk assessments from the raw numbers and put these in one place for easy viewing daily and to inform risk trade-off decisions.
Lockdowns have been used and are coming back. But these are not risk free. They were and are massively disruptive and harmful -- to unimaginable numbers of people. Not only financially, but also in other deaths. It may be eye opening to see the excess death, over covid, in the prime working age groups (25-65). (Graph>...age)> On the other hand, the risk from the virus, for vast majority of people is not as much as we thought or feared early on. So that means, for many, the choice between risking catching the virus, and a for sure, financial ruin and/or other health problems is not so clear. Nor is the risk calculus the same for everyone. These lockdowns may have been necessary at the beginning, I personally think they were and did what they were supposed to do, but do we all really need to be locked down now? And if so, who, if anyone?
And what about Herd immunity. Will/would that save us as many keep putting out there as a possibilty? Are we close?
We are still seeing outbreaks. The Covid-19 virus is very contagious and will spread like wildfire if given an opening. We also know it is possible to slow the spread and keep it within reasonable levels as long as it does not get too far ahead of us. That leads me to the final question I was curious about: How did we do? and how are we doing now?
Those are some of the questions that I hope this site can provide some insights and numbers to help you answer for yourself.
Things are changing and evolving everyday -- a lot. And so much is dependent on people's behavior that it is impossible to know or predict for absolute and forever sure anything. Instead of showing the racking up numbers though, this site focuses on current trends and information and is locale oriented. I tried to make some of the measures sensitive to our behavior changes so that we can see adverse changes in direction of the numbers early, such that people can take action on an outbreak before the virus spread gets ahead of us.
A lot is uncertain and this site is no different, so I caution using things here in isolation or as absolutes. It is meant to give ballpark numbers and near-time feedback as to the situation on the ground and direction of change. It is not in depth analysis or modeling. The data sources used are publically available and kept up pretty reliably. See About for the list of data sources and below for a few caveats about estimating etc.
Thanks for visiting,
SAM
Personal note to you: hope, opinion and an ask and plea
Hope this site provides some helpful information. And also hope, like me, that you may find things to not be as bad as feared. We bent the curve and the virus spread was fairly steady state or going down until October. Nov/Dec spread is taking off though. I speculate due to a number of factors: some we can mitigate with social distancing and mask wearing; other's like visiting family and friends without social distancing, are sacrifices many are just not willing to make, (understandable and I can empathize so will not judge, however, maybe will just inform if not known, that from all I read these gatherings are by far the places where most infection is spread.); and other's still, like being indoors more of the time, is simply something that is.
It may not be as bad as first fear. In my opinion though, now is not the time to give up or give in to the desire to rip off the masks and let the virus rip. Although personally dubious that would ever be the best alternative and I strongly think/feel this is unacceptable on many levels at the moment, I will give caveat to the "strongly" and add: at least, not until those at risk are protected or have a chance to protect themselves. A vaccine is here. When that become widely taken or at least available, then, ethically and I'd say constitutionally perhaps, taking the vaccine and/or wearing masks etc., may get solidly into the realm of being more of a personal risk decision. But we are not there yet. Everyone's actions, with respect to the virus spread, still indirectly affect the safety of the community at the moment.
Please stay engaged. I know and agree that socially distancing is not a perfect barrier. We can waste time and argue til the cows come home on exactly how much or not. Nonetheless, what is a given, is that whatever and whenever you block the virus's path to yourself or others, that moves the odds in our favor over the virus spread. Let's play hard to win.
Opening up and social distancing measures are not mutually exclusive. Quite the opposite. When deciding to take the risk of going out, it seems only commonsense to also take reasonable extra precautions to mitigate that additional risk. Think of it as going out in the rain. Most people likely would still go out but maybe would take an unmbrella and/or limit time standing in the rain. No different for the virus. It's circulating out there and is more deadly than the rain, so please continue to be part of the efforts of many people in keeping us, both safe and safe enough to stay as "open" as is reasonably possible.
I hope most of you can get out there and back to work, school and life.
Just my reading of the stats and my opinion, the risk of getting seriously sick and death are worth more than a passing consideration for those over 50. For current survivability rates see Graph>...age)> When I look at these, what it means in action to me - just my starting strawman to start from then trade off other considerations: if you are over 50, maybe be extra careful going out and maybe some work accomodations are in order; over 65, start to consider avoiding discretional going out and work accommodations; over 75, be careful, avoid crowds or any close contact in the public; over 80+ consider staying in most of the time. For the elderly, overall odds of getting and dying may not look like that much, but if you do get it and show symptoms the odds of going to hospital are very high, as is death. So the older you are I would focus more on that rather than the 97-99+% being bandied blithely about in media at the moment. Those numbers I think are Recovery% including estimate for asymptomatic cases, the "R% w estm asymp" herein. But, see cautionary foot note about "R% w estm asymp", with respect to and for the older age groups.
May you and your families stay well
Some notes on assumptions and estimating etc.
One of the biggest data issues causing inaccuracies came about when huge numbers of random test positives were put into the case counts. This made period to period monitoring nearly impossible, broke models and keeps us in the dark and hard to assess and communicate the risks to people. To mitigate this as best as I could, this site uses daily hospitalization, which is not random, and backs into a "deemed" case count that matches the hospitalization/case ratio when case counts basically consisted of "clinically sick" early on. The calculations here use this "deemed" case count, so to be able to make period to period comparisons. The deemed case calcuation is done for case counts after 4/30 when CDC guidelines for who could get a test changed. Although case counts are still interesting, the testing is at its core random not only timewise but place wise, which in m opinion makes them inapropriate to use for analysis or making any sort of decisions or conclusions on.
In addition to some intuitive assumptions, some measures use estimates of viral constants: infection period (14 days), presymptomic infectious period before symptoms and testing and precautionary quarantine (3 days), asymptomatic cases to clinical cases (10.68 to 1), hospitalization from clinical cases (.1398), and days to death from reported (14). These were derived from data and studies early on, again prior to massive random testing being included in the case counts. Caveat here is that the studies and serologic testing ratios did have a lot of variability, but well within an order of magnitude.
Reporting standards continue to change and states put large counts from prior periods into the current numbers. Have seen numbers taken out or moved from county to county or combining counties. I removed many of the big anomalies when I saw them and found a press release. Other minor data inconsistences, like counts going down or counties stopped reporting, are smoothed programmatically to maintain the integrity of current rates of change and period stats. So if you notice that total cumulative counts do not exactly match published ones that is probably why.
Finally, I am an engineer not a doctor or epidemioligist expert. Please use this site as presentation of some interesting numbers and maybe ways to look at them in context. I have included some discussions on topics and analysis in as an open-minded and balanced way as I could. Many times took to analysis in order to see for myself if things being bandied about in media make sense and/or fit with the numbers. I tried to be transparent on reasoning and the layman assumptions I made to make something useful in a practical way, not necessarily perfect. I think the experts are still learning. I definitely am still learning as I go and this is a work in progress.
Again, please do not take anything here as personal advise, but maybe it can help with information about the state in your area. Listen to the experts and make you own decisions.
That's it for the caveats.
Top ↩Counts and Estimates
Please keep in mind there are one or two levels of estimating on the many of measures shown in this website so use only as one piece of information in your personal decision making.
The rate of new infection is directly based on the odds of people being infected. This has two aspects, which I'l divide into 5 factors:
Active is, analogously speaking, the pool of fuel for the virus's continuation. When we slow the spread we lower the flow going into the pool. Eventually it gets to a steady state: new infections in are about equal to previous people coming out of their infectious period. As can be seen in the graph of the active pool over time, we as a country did bring down the exponential growth in infections and achieved steady state and then some. Don't think this was by accident. We did really well in the lockdowns and right after starting tp open up. Then summer hit and many relaxed I think. After another round of intense targeted efforts, we got the pool slowly going down again but only after we made it significantly larger. The larger the pool the harder it is then to get it moving down again as that factor in the odds is that much greater, need greater lowering of other factors to get the odds back to being low enough to get the spread to slow below 1.0.
It should be noted that if we do get the curve going down, all it takes to keep it going down is continuing the same level of social distancing, assuming the virus infectivity remains about the same. Cannot say for sure yet as it is unknown whether, like the cold and flu, it gets worse in winter or not. For now I am assuming not too much difference given it did not seem to die off in the summer. Indications though are that in winter more activities are indoor and social distancing is harder and virus can remain in the air close to people more than it can outdoors.
That being said waves of virus are starting again in Oct/November. I suppose one could attribute to the weather, but that would be ignoring the more and more people not wearing masks and going out in crowds and young people, who seem to be going out and getting the infection on-purpose and if not on purpose, going out without care. So I think the increases is due more to people getting tired of thesocial distancing, defying and/or quitting on the masks, rather than the weather.
It might help to think of social distancing as our collective active fighting the virus rather than personal protection. I think there may be more than one reason why we might want/need to keep up the "war" effort. See "Herd Immunity - Use it or abuse it" under About>Topics for some thoughts on why we might want to keep up our vigilance and energy in social distancing even though it is hard when things are not as dire.
In October, many are not wearing masks as sign of political loyalty, or maybe they take no lock-down to be green light social distancing no longer needed either. This latter entangling of ideas is I think actually opposite to reality. In fact if we want to open up while this virus is still on-going, social distancing is probably even more important and needed.
Measures and viral constants in calculations | ||
Sick Cases | Note, in order to maintain apples to apples analysis and comparisons from prior periods, the cases reported are adjusted by day to contain an equivalent case count based on the "pre-testing" definition of cases. Prior to extra testing only those exhibiting symptoms and sought medical treatment were being tested. Herein that is called the "Sick Cases" or abbreviated to "S Cases" or "S". Hospitalization/Sick cases is assumed to be a viral constant that has not changed. This nominal hospitalization rate from sick cases was estimated from hospitalization of cases prior to 4/30. Current hospitalization counts from reported cases is used to back into a deemed "sick" case count per day. | |
Reported Cases | These are the numbers found in the data sets for "confirmed" cases. This number includes substantial amount of random proactive testing. This makes the number invalid for period to period comparisons, which is why the dashboard calculations used "Sick" Cases as the basis for calculations. | |
Active | Number of people estimated to be currently infected who are not quarantined. This pool of people is the source of new infections. Basically it is an estimate of asymptomatic cases who have not been found by proactive testing. It also includes the pre-symptomatic period of confirmed/tested cases. This is calculated from Sick cases to be the deemed asymptomatic cases per the viral constant. | |
Asymptomatic | It has been found that many more times reported cases were actually infected but did not know it. Several studies have been done to get this multiplier. | |
Death Rate | Death rates are measured from the consistent "Sick" cases to allow period to period comparisons. | |
Dyr/100K | Annuallized current 7day Death rate in population. Takes deaths in a 7 day period divide by (population / 100K) then annualized by multiplying by 365 days/7 days | |
Dyr/100k (21 day alternative) | For small counties, in order to not give a small number of isolated deaths too much weight in annualizing, if the county is less than half the average county population (<500000) a 21 day period is calculated and will be used if that is less than the 7-day. | |
Constants | ||
Hospital / Sick | 0.1398 Ratio calculated from data prior to 4/30/20. This was when confirmed cases were by definition "sick" cases. | |
Infected / Sick | 10.68 - average multiplier found from samples taken early on that were comparable to the "sick" cases early on. See CDC for more detail | |
https://www.cdc.gov/covid-data-tracker/#serology-surveillance | ||
Infectious period | 14 days | |
Presymptomatic period | 3 days - It is assumed that confirmed cases were infectious 3 days before being tested and quarantined | |
EOY EOQ | End of year or end of Quarter. This is used in projection track and numbers are show in the some tables. In 2020 the projection was done to "vaccine" miltestone which was end of year, 12/31/20. Vaccine was available in 2020. In 2021 changed to measure until end of quarter. Dashboards showing the state at the end prior quarters may be seen under History menu. | |
Days2death | 14 days from reported. Note in projection tracks the EOY or EOQ death number counts future death from cases at EOY/EOQ. | |
Legends | ||
Risk Categories | ||
Risk Level | Legend | County's category is 1st it qualifies for |
High | ||
3 | 1 Death% | High risk and High Death rate from cases or in population. In Population: > 850 annualized deaths/100k/year (Note, 850 is nominal w/o coved). From Cases: (rolling rate>10% or >7.5% and not going down(2d D>5)) |
3 | 2 Accel | rate is increasing >100% (and # new s-cases per day >5) |
3 | 3 Hi sprd | rate of spread is >= 17% per day (current rate may be increased up to 2x by accel) |
3 | 4 Hotspot | active (14 days) infection, not quarantined > 5 per 100 pop, i.e. 1/20 (mostly estimated asymptomatic) |
Medium | ||
2 | 6 decel | sprd >= 7.5% and decelerating: accel <-10% or # active going down or( accel + death decel) < -10%) as long as accel<100% |
2 | 5 spread | rate of spread >= 7.5% (and < 17%, # new scases per day > 5) |
Low | ||
1 | 9 <=10 | new scases <= 5/day and not high accel i.e. <20% |
1 | 8 down | low spread rate and decelerating (see above for decel definition) |
1 | 7 Watch | various. Typically low spread or small number of cases but not in decel |
Very Low | Modified Risk Level after categorization. Used for some graphs and the map risk assignment. Does not affect category assigned. | |
0 | Llow | 3 weeks at low risk. |
Death Flag/indicator (Dflg) | ||
Dflg | H igh | Above normal death rate of cases. Usually indicative of nursing home outbreak or the like. |
if > 1/day (7d) | J | Category High death rate as H and risk in county or other entity is otherwise high |
P opulation | High annualized 7dD/100k population rate, >= 850. 2019/2017 total death rates were 857/864 | |
E levated | Elevated death rate in population, >= 200. Comparable to 2 highest causes of death in 2017: heart disease 198/yr, cancer 189/yr. This is considered moderately high. These will not cont against the states overall risk like the higher levels above are. They are listed in HighDR sheet for reference. | |
A ccident | Death rate above accident but less than Elevated. This level is considered slightly higher than normal everyday risk. They are listed in HighDR sheet. >= 53 | |
Note: 2017 next highest rate of death was for accidents was 52.17. Two other points of reference: flu 17.09; suicide, 14.48 | ||
High | This is listed in the county for high death rates. H, J, and P | |
Dflgs | Four weeks of Dth flag. First is most recent then prior periods. | |
Risk Level | ||
Risk Level | 4 | Risk recently got worse, i.e. from last period |
5 | Risk worse and stayed worse for more than a period | |
3 | High risk | |
2 | Moderate risk | |
1 | Low risk | |
0 | Low risk for 3 periods |
Column Descriptions | |
%pos | Percent positivity for current period. Number of positive tests within all tests in last 14 days. Only available at the state level in this dashboard |
Pop | Population |
Days Since max | Find maximum sick cases per day and sees how long ago that was |
% from max | 1 - current sick cases / max |
Sprd Rate | Spread rate. Number of new sick cases per day as percentage of contagious pool (# sick within contagious period regardless of quarantine status). This value is a 7 day average |
Accel | Acceleration of spread rate. Current 7d spread rate / previous period's (7days ago) spread rate |
Asym/100 | Current active pool number / population |
Hot (#/ 100) | Approximate number of contagious people per 100 population this week and into next week. This is the (max of current estimated active pool or the active pool projected one week using spread rate and accel / pop) * 100 |
Dyr/100k | current 7 day (or 21 day for small populations) deaths per 100,000 population, annualized to compare against other CDC published rates for other death causes |
Dth accel | 7 day average death rate / prev (7 days ago) death rate |
7dD | deaths in last 7 days |
Dth% | 7 day deaths from 7 day sick cases reported days2death ago |
Category | See Legend for the values. It is intended as description of stage of virus spread. |
Active | Estimate of currently active asymptomatic or pre-symptomatic contagious people who are out and about in the community. It presumes anyone testing positive for the virus, i.e. in reported counts, are quarantining. |
7dS | Estimate of comparable sick cases (comparable to Mar/Apr). This estimate is derived from hospitalization, not testing, in order to remove the uncertainty "more testing" introduced into case counts. |
Cat -1 | Category one week ago |
Actv -1 | Active one week ago |
7dS -1 | 7d Sick cases one week ago |
Cat -n, etc | that data snap shot 2 and 3 weeks ago |
Dflg | Death flag(s). Indicate high death rate with respect to population or high with respect to sick cases. See legend for values. When 4 are given, the first is most current week, followed by the prior 3 weeks. |
Risk Lvl | Risk Level. See Legend. Basically category risk level (3-High, 2-Med, 1-Low) with following enhancements: 4 if category got worse from last week: 5 if risk level got worse and stayed worse: 0 low risk for >3 weeks |
Map Risk | Prior name for Risk Level. |
cHosp | Current hospitalization. This is from daily state reporting. |
7d D /cHos | since deaths are delayed from hospitalization this is current 7-day deaths over the current hospitalization "days2death" ago |
Measurements from and used in calculating Projection tracks | |
proj_a | active pool change rate. It is the main projection factor used in the "sick cases" projection tract. |
proj_dr | the measured deaths rate. This is applied to the projected case counts to get the corresponding death projection tracks. |
~~ had | Very, very rough calculation of percentage of people that have had the virus. This is calculated by # sick cases (estimate) times the asymptomatic viral constant (also an estimate), i.e. how many on average had antibodies versus those reported clinically sick. This constant was estimated from CDC serological studies early on. At this point the modeling in this dashboard assumes all who have had the virus are immune. |
EOY Cases | End Of Year estimate of sick cases from current projection track (or for history, projection from the day the history is shown for) The projection used the proj_a factor of that day. See History for factors and this estimate over time. |
EOY D | End Of Year estimate of deaths using projected sick cases. |
EOY Rptd Cases | To get an estimate of EOY reported cases. Assume the current Rpt/Scases ratio (calc using most recent delta cases) thru EOY. Apply this ratio to the difference of EOY Scases - current Scases, and add to the current reported cases. You can see this calculated number in the Graphs>Workbook graph's projection tracks. |
Data Sources | ||
JHU | Johns Hopkins Data | Time series by county for confirmed and deaths counts |
License | https://github.com/CSSEGISandData/COVID-19 | |
Location | https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data/csse_covid_19_time_series | |
Follow links for US confirmed and US deaths to data pages: | ||
*confirmed*.txt | https://github.com/CSSEGISandData/COVID-19/raw/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_US.csv | |
*deaths*.txt | https://github.com/CSSEGISandData/COVID-19/raw/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_deaths_US.csv | |
The Atlantic | The Covid Tracking project | Hospitalization and testing time series by state. Stopped reporting 3/7/21. Used for saved quarterly Dashboards in 2020. See History for these. |
License | https://creativecommons.org/licenses/by-nc/4.0/legalcode | |
Location | https://covidtracking.com/data/download | |
daily_hosp.csv | https://covidtracking.com/api/v1/states/daily.csv | |
Health Data | HealthData.gov | Hospitalization time series by state. Switched from the "The Atlantic" hospital data to using this 3/13/21 |
License | Public use data from US Health and Human Services | |
Location | https://beta.healthdata.gov/Hospital/COVID-19-Reported-Patient-Impact-and-Hospital-Capa/g62h-syeh | |
daily_hosp.csv | https://healthdata.gov/resource/g62h-syeh | |
New York Times | optional alternate data source for cases and deaths times series. Not used; using JHU data at present time. | |
License | open source | https://github.com/nytimes/covid-19-data/blob/master/LICENSE |
Location | https://github.com/nytimes/covid-19-data/raw/master/us-counties.csv | |
Vaccine dashboard. Has state table with state priorities, links to states, % of pop and doses given and other information | ||
link-to | vaccine page | https://www.nytimes.com/interactive/2020/us/covid-19-vaccine-doses.html |
CDC | Various data | |
Published by | National Center for Health Statistics | |
Cases and Deaths | ||
cdc_by_week_age .csv | 2019-2020 Weekly counts of death by various causes including covid | |
Location | https://data.cdc.gov/NCHS/Weekly-Counts-of-Deaths-by-State-and-Select-Causes/muzy-jte6 | |
https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-by-Sex-Age-and-W/vsak-wrfu | ||
cases_by_age_ group.csv | Cases and deaths by age from case records. See cases and deaths by age on CDC data tracker page for demographics | |
https://data.cdc.gov/Case-Surveillance/COVID-19-Case-Surveillance-Public-Use-Data/vbim-akqf | ||
. | https://covid.cdc.gov/covid-data-tracker/#demographics | |
Data is on data.cdc.gov. Note this database is only updated by some jurisdictions so case and death counts will not match those reported on JHU site. This data is used only to calculate estimated survival rates within age groups. | ||
Location | https://data.cdc.gov/Case-Surveillance/COVID-19-Case-Surveillance-Public-Use-Data/vbim-akqf | |
stWkAge Master | Deaths by state and age (2015-current) | |
Location | https://data.cdc.gov/NCHS/Weekly-counts-of-deaths-by-jurisdiction-and-age-gr/y5bj-9g5w/ | |
Vaccine data | Vaccine allocation, distribution and adminstered | |
VAC_day | Vaccinations by state and age | |
Location | https://data.cdc.gov/resource/unsk-b7fc | |
CDC_VAC_date | Manually download and save by date. Found automatic replacement, not used in dashboard | |
Location | https://covid.cdc.gov/covid-data-tracker/#vaccinations | |
Vaccine allocation and distribution by week for each manufacturer: | ||
VAC_dist_M_date | Moderna: | |
Location | https://data.cdc.gov/Vaccinations/COVID-19-Vaccine-Distribution-Allocations-by-Juris/b7pe-5nws | |
VAC_dist_P_date | Pfizer: | |
Location | https://data.cdc.gov/Vaccinations/COVID-19-Vaccine-Distribution-Allocations-by-Juris/saz5-9hgg | |
VAC_age | Vacinated by age (data set is by date and has other demographic sets) | |
Location | https://data.cdc.gov/Vaccinations/COVID-19-Vaccination-Demographics-in-the-United-St/km4m-vcsb | |
Vaers | ||
Vaers data | Vaccine Adverse Events database, includes COVID-19 | |
https://wonder.cdc.gov/vaers.html | ||
Breakthru | Breakthrough tracking | |
Interactive page at CDC: | ||
link-to | https://covid.cdc.gov/covid-data-tracker/#rates-by-vaccine-status | |
link-to | https://covid.cdc.gov/covid-data-tracker/#covidnet-hospitalizations-vaccination | |
Location | vu_yymmdd.csv | https://data.cdc.gov/resource/3rge-nu2a |
US Census Bureau | 2010 - 2019 population by single year ageand sex for US and each state | |
state pop by age | https://www.census.gov/data/tables/time-series/demo/popest/2010s-state-detail.html | |
Location | sc-est2019-agesex-civ.csv | https://www2.census.gov/programs-surveys/popest/tables/2010-2019/state/asrh/sc-est2019-agesex-civ.csv |
US Bureau of Labor Statistics | Seasonally adjusted unemployment by state. Most recent month. Published towards end of month for prior month. | |
unemployment,xls | Unemployment by month. Copy from webpage. .csv has most recent month. | |
Location | unemployment.csv | https://www.bls.gov/web/laus/laumstrk.htm |
Misc Links | ||
link-to | Good mask discussion | https://www.ucsf.edu/news/2020/06/417906/still-confused-about-masks-heres-science-behind-how-face-masks-prevent |
link-to | Excess Deaths | https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htm |
If you have a comment or idea to add information or make this more useful please email.
... only constructive comments please.
This is a personal website done on my own time, so I may not be able to respond all comments or it may take some time. If you do email, your consideration and patience will be appreciated.
Top ↩This website does not collect any information nor is it connected to any advertising provider. The hosting service, ACEWEB.COM may log or temporarily collect minimal non-personal information, like IP address, solely for security and performance purposes.
Top ↩Informed layman's viewpoint. See "About>About" for descriptions and information. Estimates and data smoothing are used to present coherent numbers from period to period. Informational only. Should not be relied upon as your sole source for decision making.