Forecast accuracy and how it affects the car wash industry.

[Sey-Mour Snow]

I'm sitting here with a car wash owner and he is blasting me and all meteorologists. His argument how is that we are wrong 70% of the time and every time we forecast rain/snow and it doesn't precipitate we are unnecessarily costing him money.

He is so sure that we over dramatize the weather as weather anchors and does not believe me that as a weather anchor I only tell the truth and never exaggerate a forecast.

So that brings me to my question how accurate are we on average throughout the year rain,snow, sunny days, in terms of days the general public knows we are wrong vs days they know we made the right forecast. I'm thinking 85%.

[CoastalWx]

Yeah I love those clowns that hear what they want to hear and then get mad for something they thought was expected/unexpected. It's hard to quantify accuracy in the TV domain. You have a large area and it can be impossible to explain all the nuances of a forecast for different areas. I think we are getting better at communicating forecasts to the general public as a whole. In the 80s, it was "A chance of aftn shwrs and tstms," which does absolutely no good for anyone. Now with better models and technology, we can at least try and time a line moving through and communicate this to the public. Same with snow amounts. We can show with sim radar how the shield will move in and at one time. Granted this sim radar product may only be from one model...but it gets the point across. The individual met can always add comments to fine tune things.

[CT Rain]

I'm sitting here with a car wash owner and he is blasting me and all meteorologists. His argument how is that we are wrong 70% of the time and every time we forecast rain/snow and it doesn't precipitate we are unnecessarily costing him money.

He is so sure that we over dramatize the weather as weather anchors and does not believe me that as a weather anchor I only tell the truth and never exaggerate a forecast.

So that brings me to my question how accurate are we on average throughout the year rain,snow, sunny days, in terms of days the general public knows we are wrong vs days they know we made the right forecast. I'm thinking 85%.

 

Sounds like it's time to find a new car wash to visit.

[Cold Miser]

I am surprised that this thread has not blown up into something big.

[OceanStWx]

Yeah I love those clowns that hear what they want to hear and then get mad for something they thought was expected/unexpected. It's hard to quantify accuracy in the TV domain. You have a large area and it can be impossible to explain all the nuances of a forecast for different areas. I think we are getting better at communicating forecasts to the general public as a whole. In the 80s, it was "A chance of aftn shwrs and tstms," which does absolutely no good for anyone. Now with better models and technology, we can at least try and time a line moving through and communicate this to the public. Same with snow amounts. We can show with sim radar how the shield will move in and at one time. Granted this sim radar product may only be from one model...but it gets the point across. The individual met can always add comments to fine tune things.

 

At the least the car wash owner doesn't exaggerate. Normal people just tell me I'm wrong 50% of the time, this guy went all out for 70%.

 

But I would say for a day 1 forecast, as an industry forecasters are no worse than +/- 2 degrees on temps and > 80% on PoP.

[CapturedNature]

On a related note, the restaurant folks say (or used to say) similar things.  I never wanted to go the TV route but my mentor was a meteorologist on TV and so I interned with him one year and I remember getting calls from restaurant owners starting on Tuesdays saying "please don't mention any rain or storms in the forecast for the weekend".  This was back in 80s and we only did 5 day forecasts but they would insist that if we mentioned storms earlier in the week that people would change their plans and their business would be down for the week regardless of the actual weather.  I don't know if that has changed, but it's interesting that you brought that up about other businesses.

[OceanStWx]

On a related note, the restaurant folks say (or used to say) similar things.  I never wanted to go the TV route but my mentor was a meteorologist on TV and so I interned with him one year and I remember getting calls from restaurant owners starting on Tuesdays saying "please don't mention any rain or storms in the forecast for the weekend".  This was back in 80s and we only did 5 day forecasts but they would insist that if we mentioned storms earlier in the week that people would change their plans and their business would be down for the week regardless of the actual weather.  I don't know if that has changed, but it's interesting that you brought that up about other businesses.

 

I think for the most part, when people go about to make plans that's the forecast they remember.

 

We find that our most talked about/remembered forecast is always the first snowfall map, no matter how many times we update or change it. 

[CT Rain]

On a related note, the restaurant folks say (or used to say) similar things.  I never wanted to go the TV route but my mentor was a meteorologist on TV and so I interned with him one year and I remember getting calls from restaurant owners starting on Tuesdays saying "please don't mention any rain or storms in the forecast for the weekend".  This was back in 80s and we only did 5 day forecasts but they would insist that if we mentioned storms earlier in the week that people would change their plans and their business would be down for the week regardless of the actual weather.  I don't know if that has changed, but it's interesting that you brought that up about other businesses.

 

Yeah - they still do. lol

[Typhoon Tip]

but there is some truth to all this...

 

let's work this through logically:   if/when there comes a day when forecast precision lands on 90% (for all objective and subjective interpretations averaged...) accuracy, business' that depend on a sunny -vs- rainy -vs- snowy sensible/impact conditions are going to have to figure this out on their own, and/or "blame the messenger" as it were.  

 

unfortunately for that niche business owner that specializes in patio weather for their three-season cafe' ...they are unfortunately not the priority; the priority is informing everyone.   it would be a problem that cannot even be resolved via boutique forecasting because Meteorologist don't actually control the weather (haha) and won't be able to stop predicting rain just because it doesn't suite said hapless cafe' owners business.  

 

in other words, ...Cape Cod's tourism industry, which fell victim to one of the more horrid summers in memory in the late 1980s, with weekend rainstorms on the button during almost all of them between the tradition vacation times of Memorial and Labor days, they are going to be SOL either way - does the Met community still take the blame in such an era where their D7 forecast warns of another beach-weather apocalypse and is right 90 % of the time?  

 

logically, the messenger should never be blamed of course.  however, i personally have no faith in people as a person that may very well have an actual genetic predisposition to think ALL people are just relative azzholes on a spectrum of measurable douche baggery, so i'd have to go ahead and assume Mets would still get phone calls from self preserving 'me-first' types that think the forecast that goes out to 'everyone' needs to still be tailored to their individual 'me-first' needs...

 

interesting conundrum in the future.   we aren't that sophisticated just yet, as a science, art, or industry though.  At D4, I think the accuracy has improved to 60 to 75%, but falls off pretty dramatically thereafter (...but obviously, forecasting patio weather in the desert is a slam dunk at 10 days lead; not so much in Miami with daily 3 PM hit or miss thunder).  so in the meantime, every time the weather rains on the proverbial parade, our community gets to "weather" the storm of outrage either way. 

[Ginx snewx]

Interestingly our business is much better with bad weather

3 M's Museums Malls Movies do well on bad weather days. Summers though most vacation people wake up and check the weather in the AM then plan the day. Because we have so many outside events from weddings to museum activities it's vital I stay up to date on what could change. Cancellation can be a 10K loss or at minimum a massive relocation of setup. I have weaned out the hypesters the rip and readers.This place is invaluable as well as my learned skills. Ryan, Scooter, Will, Phil, Chris, Eck and others are pretty much the ones I check first. With combined opinions I have been extremely lucky and haven't busted on moving an event here yet. Lots of people depend on me to get it right. Often I get asked who I use for the weather and I tell them I have access to the best at a click of the finger.

[dmillz25]

There are some owners out here who are morons and close even when there's a 30% chance of a shower.

[Cold Miser]

I think the car wash thing with the owner blaming the meteorologists is b.s.  My guess is he has a crappy car wash, or crappy employees working there, or over priced, or all 3, thus people don't want to go to his business. Think about it, how many people actually plan their future days around a car wash? A trip to the beach maybe, but a car wash?  No.  It's something that you think of (or may not) doing in a few days...You wake up and if it's pouring rain or snowing out, you don't go. If the weather is decent you go.

 

I will go out on a limb and say that 99.99999% sane people do NOT look at the weather forecast days before a car wash trip.  

[OceanStWx]

Bad news for the car wash industry: https://www.washingtonpost.com/news/capital-weather-gang/wp/2016/04/11/accuweather-extends-its-controversial-45-day-weather-forecasts-to-90-days/

 

Time to cancel my plans for the 4th of July, looks like a 98% chance of rain and 6.5 hours of it overnight. 

[CoastalWx]

LMAO. Sunny and mid 80s on July 4th in DC. That sounds like going out on a limb there. That will always have a 85% chance of verifying. 

[CapturedNature]

Bad news for the car wash industry: https://www.washingtonpost.com/news/capital-weather-gang/wp/2016/04/11/accuweather-extends-its-controversial-45-day-weather-forecasts-to-90-days/

 

Time to cancel my plans for the 4th of July, looks like a 98% chance of rain and 6.5 hours of it overnight. 

 

That's just ridiculous.  I belong to a community of maple syrup producers and I know several who based their tapping and even ending their seasons based on those 45 day forecasts.  Results were all over the place with people losing out on sap that they could have collected to collecting sap that wouldn't produce consumable syrup.  Most of us used common sense but they are real world examples of people whose business was affected by AccuWeather's decision to run these long range forecasts.  Maybe they should think about those consequences before doing something like that.

 

Truth be told, when I got out of college I started a weather business with some colleagues and one of the "unique" services that we offered was long range forecasts for specific events.  We used a mix of climatology, seasonal forecasts and other trends to come up with odds for precip on days that were about a month out.  We never provided anything specific and only did as we got about a week out.  We never sold a lot of those but most of them were for weddings or corporate gatherings.  Even today I can't see providing anything more than generalities for days more than a week or two out.

[OceanStWx]

That's just ridiculous.  I belong to a community of maple syrup producers and I know several who based their tapping and even ending their seasons based on those 45 day forecasts.  Results were all over the place with people losing out on sap that they could have collected to collecting sap that wouldn't produce consumable syrup.  Most of us used common sense but they are real world examples of people whose business was affected by AccuWeather's decision to run these long range forecasts.  Maybe they should think about those consequences before doing something like that.

 

Truth be told, when I got out of college I started a weather business with some colleagues and one of the "unique" services that we offered was long range forecasts for specific events.  We used a mix of climatology, seasonal forecasts and other trends to come up with odds for precip on days that were about a month out.  We never provided anything specific and only did as we got about a week out.  We never sold a lot of those but most of them were for weddings or corporate gatherings.  Even today I can't see providing anything more than generalities for days more than a week or two out.

 

Accuweather is disingenuous about these products at best, claiming that they are scientific and that the public knows they aren't locked in forecasts.

 

PoP to the 1%? Really? We could probably round those numbers off to the nearest 5 or 10% maybe. To me that's advertising a false precision. And I'm sure there is a section of well informed public that knows this is mostly just using climo and long range models to tweak in one direction or the other, but there is another section that is super pumped they can forget the tent on their July 4th BBQ bash.

[Typhoon Tip]

LMAO. Sunny and mid 80s on July 4th in DC. That sounds like going out on a limb there. That will always have a 85% chance of verifying. 

 

 

 

Forecast:   Today, increasing lightness followed by brightness, diminishing toward evening.  Temperatures above absolute zero ...but not quite hot enough to fuse hydrogen into helium-3.  

 

Tonight: Lingering lightness early, followed by darkness, same temperatures.  

 

...what's my PoP  ?

[bobbutts]

NWS blows every storm 100% of the time for me.  

Never once has a storm played out like forecast.

 

"Overnight: Snow, mainly after 2am. ... and sleet before 2pm, then sleet likely between 2pm and 4pm, then freezing drizzle likely after 4pm."

[OceanStWx]

NWS blows every storm 100% of the time for me.  

Never once has a storm played out like forecast.

 

"Overnight: Snow, mainly after 2am. ... and sleet before 2pm, then sleet likely between 2pm and 4pm, then freezing drizzle likely after 4pm."

 

Another satisfied customer...

 

I'm sure there is no hyperbole involved in our 100% failure rate.

[Cold Miser]

NWS blows every storm 100% of the time for me.  

Never once has a storm played out like forecast.

 

"Overnight: Snow, mainly after 2am. ... and sleet before 2pm, then sleet likely between 2pm and 4pm, then freezing drizzle likely after 4pm."

lol.  I always think that pinpointing times like this is a dangerous move.

[Typhoon Tip]

folks shouldn't take the point-and-click forecasts word-per-word, just as much as the technology probably could use some attention to detail. 

 

i was having this conversation with a Met that does afd's for some Plains office at some point over the last couple of years, and he was saying they (meaning the NWS et al) are/were actually aware that the database-to-consumption needs work, and is a work in process.

 

i get the sense, tho, that folks either are unaware of that, or, are just some kind of Asperger string-literate types that can't process information with understanding.

 

even if Jo Jack consumer didn't factually get the memo and understand it, imho there should be a modicum of common sense when reading a text that is discretely laid out in such a way that is obviously perposterous, anyway.  like, ...gee, maybe that isn't exactly what is intended? dip schits... right, it's going to be sunny, high near 54, with blizzard conditions toward evening, exactly like it's said.  what are you, some kind of useless f-tard!

 

but, in a society run by litigation larva just waiting to for their chance to chrysalis...that blaming attitude rears its chance to do so at least excuse imaginable, and probably detrimentally blocks people from thinking much, too.     

 

whatever.  anyone blaming their business losses on a forecast is gets what they deserve.  end of discussion - 

[OceanStWx]

lol.  I always think that pinpointing times like this is a dangerous move.

 

It's a casualty of the 1 hour grids. The computer formatter will just read exactly what the grids show, rather than ballparking or rounding off times. 

 

I'd say, more often than not the trend is right.

[powderfreak]

It's a casualty of the 1 hour grids. The computer formatter will just read exactly what the grids show, rather than ballparking or rounding off times. 

 

I'd say, more often than not the trend is right.

 

Yeah the trend is right but folks do laugh and chuckle at the "precision" of it all... I don't think that helps sentiment about forecasts being correct.  I still think the zone forecast is the way to go.  None of this X before 1am, then XY from 1-2am, then Z from 2-5am, then X and Z from 5-6am followed by W after 7am.

 

Where as the zone forecast will just read, "Cloudy with snow, sleet, and freezing rain developing after midnight...then changing to rain during the morning."

 

I know what Bobbutts was saying though... unfortunately it probably is a very high failure rate with the hourly precision stuff that jumps back and forth every hour, though what you say is right that the trend is always right. 

 

At what level do you consider the forecast correct or wrong?  Do you look at a day as 24 hourly forecasts that can either be right or wrong?

[powderfreak]

I think the car wash thing with the owner blaming the meteorologists is b.s.  My guess is he has a crappy car wash, or crappy employees working there, or over priced, or all 3, thus people don't want to go to his business. Think about it, how many people actually plan their future days around a car wash? A trip to the beach maybe, but a car wash?  No.  It's something that you think of (or may not) doing in a few days...You wake up and if it's pouring rain or snowing out, you don't go. If the weather is decent you go.

 

I will go out on a limb and say that 99.99999% sane people do NOT look at the weather forecast days before a car wash trip.  

 

I don't know man...I don't like going to the bank when its sunny out.  And I only grocery shop when its partly cloudy.  Dry cleaners only get a visit if its raining lightly but not if its higher than 0.1"/hr as I don't want to get out of my car in that rain.

 

You mean you don't plan your daily errands around the weather forecast?!

[OceanStWx]

Yeah the trend is right but folks do laugh and chuckle at the "precision" of it all... I don't think that helps sentiment about forecasts being correct.  I still think the zone forecast is the way to go.  None of this X before 1am, then XY from 1-2am, then Z from 2-5am, then X and Z from 5-6am followed by W after 7am.

 

Where as the zone forecast will just read, "Cloudy with snow, sleet, and freezing rain developing after midnight...then changing to rain during the morning."

 

I know what Bobbutts was saying though... unfortunately it probably is a very high failure rate with the hourly precision stuff that jumps back and forth every hour, though what you say is right that the trend is always right. 

 

At what level do you consider the forecast correct or wrong?  Do you look at a day as 24 hourly forecasts that can either be right or wrong?

 

The P&C we're just at the mercy of. But theoretically if we're making the right updates, we can get those precip transitions down pretty close. Obviously it's hard to be correct for all 2,500+ grid points, but the P&C will always be verbose and it does just seem bad to the naked eye.

 

As far as right and wrong, we break things down in 12 hour chunks for hi/lo temps. But we can track just about everything, all the way from day 7. So you can see 14 separate shifts of forecasts leading up to T=0. The idea is that we want to consistently be at or near the top of guidance sources.

 

It's why the long term may be going away. It's getting harder and harder to add value to the model consensus blend. The value will more come from messaging rather than the actual creation of the forecast.

[Ginx snewx]

Lol this dude at Ryan on FB

Yeah it is we were having this convo at work you get paid alot of money to guess and be wrong and also get your info from the actual weather people we dont even see at the NWS. Dont try to justify your joke of a job to me lol your paid to be an actor please i know people who work in the industry.

[Cold Miser]

Lol this dude at Ryan on FB

Yeah it is we were having this convo at work you get paid alot of money to guess and be wrong and also get your info from the actual weather people we dont even see at the NWS. Dont try to justify your joke of a job to me lol your paid to be an actor please i know people who work in the industry.

Too many people expect the weather "forecast", which is not a guarantee by the way...but a FORECAST to be just that...a GUARANTEE.  They want the weather to show up just as forecasted, the same way that their pizza gets delivered to the door or their coffee shows up at the drive thru every morning, or any other product shows up on the store shelves.  People want everything boxed up with a nice bow around it, with no issues.

[Cold Miser]

I don't know man...I don't like going to the bank when its sunny out.  And I only grocery shop when its partly cloudy.  Dry cleaners only get a visit if its raining lightly but not if its higher than 0.1"/hr as I don't want to get out of my car in that rain.

 

You mean you don't plan your daily errands around the weather forecast?!

lol.  

Actually, I plan the grocery trip around a sunny day.  Since I park way in the back of the lot, the walk into the store allows me to bronze my dome.  All it takes is about 48 shopping trips and I have a nice tan on the top of my scalp.

[CapturedNature]

Lol this dude at Ryan on FB

Yeah it is we were having this convo at work you get paid alot of money to guess and be wrong and also get your info from the actual weather people we dont even see at the NWS. Dont try to justify your joke of a job to me lol your paid to be an actor please i know people who work in the industry.

 

That's not even worth responding to and normally I don't let people like this get to me but every once and a while it does when I'll hear that old adage about it being the only profession where you can be wrong and still have a job.  People are such idiots sometimes and now big events are being politicized which is causing it's own problems.

 

On the other end of that is the fact that just about every time there is something science related in the news they usually put it near the weather segment....mainly because the meteorologist is a scientist.

[Ginx snewx]

That's not even worth responding to and normally I don't let people like this get to me but every once and a while it does when I'll hear that old adage about it being the only profession where you can be wrong and still have a job. People are such idiots sometimes and now big events are being politicized which is causing it's own problems.

On the other end of that is the fact that just about every time there is something science related in the news they usually put it near the weather segment....mainly because the meteorologist is a scientist.

I had to say something