Allstate Insurance is misleading their customers over weather events

Dr. Roy Spencer notes:

Allstate Should Pull this Ad and Apologize for Misleading the Public

I’ve been meaning to comment about this TV ad for Allstate insurance, which enraged me the first time I saw it. Allstate knows better (the insurance business deals with probability and statistics) and they knew this was a lie when they put the ad together:

In the ad, actor Dennis Haysbert says:

“A once-in-500-year storm should happen once every 500 years, right? The fact is, there have been 26 in the last decade.”

Setting aside the fact that we don’t have enough statistics to say anything meaningful about what happens over 500 years (hydrologists prefer to stick to 100 years as justified when talking about rare events), Allstate knows very well that such statistics refer to the repeat period for the same location… not for (say) the whole United States. It is not unusual for once-in-100 year weather events to occur more than once, maybe several times, each year somewhere in the U.S.

I consider this false advertising because Allstate knows better, and is purposely misleading the public to make more money

Here is the text of the ad, from

“A once-in-five-hundred year storm should happen every five hundreds years, right? Fact is, there have been twenty six in the last decade. Allstate is adapting. With drones to assess home damage sooner. And if a flying object damages your car, you can snap a photo and get your claim processed in hours, not days. Plus, Allstate can pay your claim in minutes,” Haysbert says while sitting in an armchair on a field and a storm is about to begin. As he mentions the ways Allstate has adapted, the clouds vanish and a rainbow appears behind him. “Now that you know the truth… are you in Good Hands?” Haysbert asks at the end of the spot.

The 500 year storm claim may have come from this article about Hurricane Harvey in the Washington Post:

Houston is experiencing its third ‘500-year’ flood in 3 years. How is that possible?


88 thoughts on “Allstate Insurance is misleading their customers over weather events

  1. The Houston 2017 flood was not a 500 year event, since there were a number of similar flooding magnitude just in the last few decades.

  2. They’ve gotta do something to stay in business.. ;>)
    insurance companies with the worst reputation? Guess who tops the list?
    Search engine tells the tale.

  3. I’ve recently seen this ad and had a similar reaction to the quote highlighted by Dr. Spencer!
    If Allstate Insurance company doesn’t understand statistics any better than this, I’m going to have to reconsider using them for my home insurance.

  4. Just left the following comment at Roy’s blog:

    The ALLSTATE slogan was “You’re in Good Hands with Allstate.”

    Now they’re doing something else with those hands.


  5. We think about the ridiculous cost for renewables, added cost for the basket of goods made with it, with zero benefits and a host of functionality costs . We think about costs for subsidies for going green so the well to do can buy cars, solar rooves, etc, but you can be sure that other skimmer jackals will be coming out of the woodwork to finish the destruction.

  6. A “once in 500 year storm” doesn’t mean it only occurs once every 500 years. It means that that type of storm has a 1/500 chance of occurring in any year. Allstate knows this, it is a metric used by insurance companies. But like most ads these days, they play on the ignorance of the general public to over react to scary misinformation.

    • ” It means that that type of storm has a 1/500 chance of occurring in any year.” and it means, as Dr Spencer pointed out, it only has a 1/500 chance of occurring in that locality in any given year, not world wide.

      • It goes further than that. A “storm event” is also classified by DURATION. There is a 500-yr storm for a 1 hr rainfall total. There is one for a 2 hr rainfall total. There is one for 4 hrs…6 hrs…8 hrs…2 hrs. There is one for 24 hrs…48 hrs…72 hrs…really anywhere in between, but typical durations are generally used.

        The point is that there is an array of 1/500 chance of occurring storms at a SINGLE location as well.

        • Technology has augmented, or warped, people’s perception of this. Now, they see events that happen 30-50 miles beyond television station antennas because of cellphones. Flooding, except for very major coastal or river flooding, never used to be covered much on TV because it was very hard to cover. No cameras were where it was flooding and by the time a station could get cameras to where it had flooded, the flood would already be over with. There’s a camera everywhere today, so anytime a foot of rain falls anyplace it’s a huge news story even though in the past it wasn’t and in any TV market a foot of rain will fall someplace every few years.

          • I believe that “augmented” is the better term.

            McLuhan’s paradigm has been reinforced by the sheer amount of data available in visually digestive formats courtesy of Moore’s Law and the Internet.

            Mainlining information to the brain.

    • To put it another way. If you have 500 locations, odds are you will approximately 1, 500 years storm in one of those locations each year.

      • If you have 500 locations, odds are you will approximately 1

        Err. The traditional way to get this wrong is (499/500)^500 which is less than one. How much is that, I don’t know, bing said it was less than 40%.

        ‘The way to get this wrong’ => it is fairly difficult to get this really right.

    • I doubt the veracity of 3 “500 year storms” included in the linked article possibly being referenced by Allstate and included in the main post above. However, that article does include this paragraph:

      “Here’s the other big caveat: The term is applied to a local area, not to the United States as a whole. So when meteorologists say the Houston is experiencing a 500-year flood, they mean there is a 1 in 500 chance of it happening in any given year in Houston.”

      (“…in Houston” was italicized in the article.)

      So even if Allstate wants to deliberately mislead on this topic, the author of that particular article got is right. also including the discussion it had of having to look at it as a probability.

      • “…So even if Allstate wants to deliberately mislead on this topic, the author of that particular article got is right…”

        Yes and no. A 500-yr storm is not the same thing as a 500-yr flood. It’s possible that the author and/or Allstate don’t know the difference.

    • Yes, that is correct, Tom.

      Also, as the author here and others have pointed out, storm frequencies are all very finely granular location specific statistical calculations. There is literally an unlimited number of “500 year events” as defined geographically across the surface area of the planet.

      Those who actually understand precipitation hydrology and hydraulic considerations, and the statistics used in the field understand that they are entirely dependent upon geography and terrain and prevailing weather patterns and micro-climates defined therein, as well as statistical definition.

      Beyond locational specifics, there are other statistical specifics involving storm intensity and duration. In other words, there are many different “500 year storms” even for a specific location – the 500 year, 1-hour event, the 500-year 6 hour event, the 500-year 24 hour event, even down to the 500-year 1 minute event … and that is just the precipitation. Then the flooding events, defined by flood plain terrain and drainage patters, also come in many different durations as well as frequencies.

      In other words, this is all way too complicated to be understood by Allstate commercial writers, let alone their prospective layperson customers.

      Within any large area – say, 100,000 square miles – there are likely to be many storms and floods of varying frequencies up to the 500 year or even longer durations somewhere within that area in any given year or 5 year period.

  7. Let’s see, the chance of having a 1/500 storm over a three year period is about 1/167. So the chance of having three storms is about 1 in 4.8 million.

    So, it is possible that the insurance industry _has_ done its homework.

    • Much more unlikely things happen every moment. Like you for example Trafamadore. The probability of that particular sperm fertilizing that particular eggcell is of the order of one in a hundred million.

      How many million Houston-sized areas didn’t have three 500-year storms during those three years?

    • And you are strongly recommended to read up on Hurst-Kolmogorov dynamics before trying again. Hydrological processes (like rain) don’t work like glorified roulette wheels.

    • Trafamodore,

      Keep it up; there are 3.8 million places in the USA to have the storm, so there is a 79% (3.8/4.8) chance of having three 500-yr storms in the USA in a single year.

      And, as noted in above posts, there are numerous different levels of 500 year storms (2-hr, 4 hr, … 3 day, etc). So given 8 different types of 500-year storms, you get a 633% (8 x 79%) chance of having any type of three 500-yr storms in a given year. Yes, it is greater than unity … it is assured.

      Please notify Allstate that they need to raise their rates (and then move to higher ground).

  8. You would need 5000 years of weather records to establish with any credibility what a 500 year storm is.
    There would be approximately 10 of those over the 5000 years.
    Even if you had 500 years, which we don’t, there is no baseline to even make a comment about a 500 year weather event. We could have had one, we could have not had one – we don’t even know what ONE 500 year weather event looks like. But there are gullible people who lap this stuff up…

    • Generally my thought as well. It’s an abuse of statistics to generalize that far outside of the sample space.

      • It requires extrapolation, which can be a naughty word. But for rainfall events, it’s not too bad.

        For example, for Tampa, FL, here are the approximate 24 hr rainfall totals for 1-in-XX yr events:

        2yr – 4.5″
        2.33yr – 5″
        5yr – 6″
        10yr – 7.5″
        25yr – 9″
        50yr – 10″
        100yr – 12″

        You’ve got plenty of weather records to get the smaller events. The IDF (intensity-duration-frequency) curves get pretty flat when you start heading towards the larger rain events.

        • Yup, Michael – and even beyond the 500 year event, hydrologic statistical analysis identifies the “probable maximum precipitation” and the “probable maximum flood” events for specific locations, which are used in the engineering design of facilities that involve particularly high hazards if a failure occurs, such as a nuclear facility or a high dam that, if it failed, an entire city downstream might get washed away.

          PMP and PMF are barely higher than the 500 year event.

          Again, these are statistical calculations – a scientist or engineer does not sit down and say that it is physically impossible that a rainfall event greater than PMP can occur … he or she is just saying that it is improbable that a greater event will occur based upon the data series and statistical calcs performed.

    • No – actually, you don’t. That’s not how statistics work.

      Rainfall precipitation statistical processes typically deal with less than 100 years of records, and even 30 years is plenty to establish any particular probability of an event with a return of X years. It’s just stats. The shorter the record, the larger the uncertainty or the sigma.

      The stats herein described are not incorrect – it is the interpretation and meaning and significance of the stats in the Allstate commercial that are objectionable.

  9. The map posted above omitted showing the Great Western Drought. Isn’t that as important as heavy rainstorms?

    Climate is more important to people Regionally than worldwide. And like every natural function, it seems to be Cyclical. Which part of which cycle governs my region this year? How can we Be Prepared! …?

  10. Yeah, this always irks me (as did that commercial when I saw it a few weeks ago). You have to start with the total number of weather stations….even within a given city you have many opportunities for a smaller (in area) very intense/large precip event, just given the number or measurement points (of course, as they get closer the precip events are correlated spatially). When you consider the total number of ‘locations’ (where we measure such things), the probability of NOT having a 500 year event in some given area and time-frame approaches zero. Twenty six seems rather low for the entire country, possibly less than one would expect in that time-frame (?)

  11. And since almost all of the damage from Harvey was caused by flooding, normal home owners insurance would not even come into play. My house got flooded and Allstate was never involved

  12. Did they raise Southern US insurance rates 2500% to cover the extra storms? No. Okay, so they know the facts.

  13. The insurance companies are major financiers of Climate Change Alarmism- why? Gee there is money to be made! To hell with morals and the truth, it’s lies and profits!

    • If they can convince their customers (who DON’T shop around otherwise they wouldn’t be customers) that the hypothetical and fictitious 500 yr event really means once in every spot every 500 yrs they they can increase their rates and make 500 times a much money. If there is any real competition they won’t be able to do that.

  14. According to FEMA, the slab of my house in Houston is at the 100-year flood elevation, with the 500-year flood elevation about two feet higher. If there have been three 500-year floods in Houston, why has my house never flooded?

    Clearly the term “500-year flood” applies to very localized areas. Both Allstate and Washington Post should well better, but choose to disinform instead.

      • So what is meant by “Storm” ?

        Inches of rain?
        Accumulated Wind/Cyclone Energy?
        Property Damage?
        Lighting strikes?
        Barometric Pressure?

        • There were 40 USGS streamgage locations in Houston with at least 15 yrs of data and which measured record flooding during Harvey. The 1-in-500 flood charcterization was based on two of those locations. 30 of those others exceeded 1-in-100. The other 8 were below the 100 yr level.

          It wasn’t a 500 yr flood for every spot in Houston.

          Most homes damages in Harvey were outside of the 500 yr floodplain.

  15. If there were an insurance company that came out and denied increasing climate disasters I’d go with them as they should be cheaper. Thing is events like Houston 2017 eat into profits. Ins Cos don’t like that. Never mind they’ve been collecting elevated premia from Florida for a dozen years with narry a check cut.

    Bottom line. CAGW alarmism justifies higher charges. Simple as that.

  16. If there is a Buck to be made about anything, then someone will make it. The Green myth is perfect for this, as the various so called undeveloped nations now know.

    All the newly gradulated persons with knowledge useless in the real world, gravitate to the Climate change field, , its a well paid government grant place.


    • There is no free market in insurance. The industry is highly regulated with substantial barriers to entry.

      • Not only that, but most of the insurance companies you can name use a re-insurer. And there are few of those, so they effectively set the rates for everyone.

      • The Agencies must advertise or the Independent Agents would go elsewhere.

        Pretty much for the same reason Big Pharma advertise their Rx to people who can’t write their own scripts. Its to generate interest in the product and then go initiate contact with a dealer of that product who can then generate more revenue with add-ons.

        There is also competition in the form of levels of service.
        Insurance is also a function of Peace of Mind, so part of the advertising is to convince people that whatever they are trying to insure is better protected under one brand rather than another.

        • I think that is the logic behind the “You are in good hands with Allstate” slogan. To convey that their customer service is above all others. Whether that is true or not is up to those insured by Allstate to decide.

          As for rates, due to improvements I made to my home, my wind insurance is half of what it was 2 years ago and thanks to FEMA finally redoing the flood zone maps in my area I am no longer in an AE flood zone so I am no longer required to purchase flood insurance. Total savings are over $1700 per year.

          And no, I do not use Allstate for my insurance needs.

          • Florida was not in “Good Hands” nor did they have a “Good Neighbor” when all the major carriers deserted the state because they didn’t have the corporate guts to deal with hurricanes.

            And I am an Allstate customer. (actually their shell company Castle Key) because my contract was too old to have the legalese that allowed them to drop everybody else.

          • Tom, when the FEMA FIRM program started FEMA was supposed to revisit/correct the maps every 10 years. They didn’t/don’t. I don’t very often see re-mapping “move” property out … usually it is the other way around.

            Note to anybody in an un-numbered ‘A’ zone … don’t wait for FEMA to correct their maps. Have somebody that knows what they are doing take a look at your property. If it can be fixed, get removed now (don’t wait till the bill is due to trigger your action), removal during the premium period gets a 100% refund but it takes a few months to process.

            Time it right and you get the premium back plus the escrow … in Tom’s case that may have been as much as $3400.

  17. Meanwhile Prof Nic Lewis has another look at climate sensitivity and finds that multi-decadal warming would be just 50% of IPCC AR5 projections after 1,000 years.

    Here is Nic’s projection of FIGURE 2, with his temp dipping to 50% of their warming after 1,000 years. See bottom ORANGE line added to their FIG 2. Thanks Nic for responding to my request. Needless to say that Co2 levels could remain high for many thousands of years in the future. At least this is what many scientists believe. Nic also finds slightly higher warming at 100 years than the original FIG 2 , see black line just below his ORANGE line at 100 years, but then both temps diverge ( their line rises after 100+ years) after that date. Who knows?

    UPDATE 13 December 2018

    “The below version of Figure 2, produced in response to reader suggestions, marks (orange line) the approximate path of temperature change implied by my observationally-based TCRE and ECRE estimates, in response to the CMIP5 ESMs mean total emissions over years 1-99 of 1919 GtC. Note that the ensemble of CMIP5 ESMs involved (whose mean warming is shown by the green, red and blue lines) had a lower mean TCRE than the ensemble of CMIP5 ESMs and EMICs used for IPCC AR5 Figure SPM.10. Note also that the warming simulated by GFDL-ESM2M (black line), which has a TCRE of 1.03°C, is marginally lower than that per my observational TCRE estimate in year 100, but is much higher in year 1000 than that per the observational ECRE estimate. That reflects this model both having a higher ECS (~3.0°C), and simulating a higher proportion (~20%) of emitted CO2remaining in the atmosphere in year 1000, than per my observational ECRE estimate”.

  18. Yes this ad really pissed me off too. If you take 5000 measuring stations – and I am sure there are a lot more than that in the USA, you would expect a 500 year storm 10 times in a year – am I wrong?

    • Are you thinking adding a 2nd weather station in your home county doubles your chance of a 500 year event?

      Or are you saying that is the logic of that Allstate ad?

      I hope you meant the latter.


  19. I guess it might also suggest that maybe Allstate Insurance don’t really know what they are doing? Now perhaps that might lead to excessively low or excessively high premiums for some customers, so take a case by case approach to buying their products. But definitely don’t invest in the company if they can’t allow competent actuaries to stop them from scoring own-goals of such magnitude.

  20. Bottom line , fear is good for business in an area where people buy something not because they know they will need it. But because they fear what may happen if they do not have it..

    • For many people, fear has nothing to do with it. The fact that they have a mortgage, or a HELOC, or some other credit attached to their property requires that they have it insured.

      (If the rate is reasonable, it is actually not a bad idea, unless you live well above any possible non-Biblical flood level. I’ve been through three “century” floods here in Arizona – and have done a fair amount of research on “The Big One,” when Confederate Army scouts were convinced they had sighted the Pacific Ocean when they were still about 300 miles away from the Colorado River.)

  21. Just to make things clearer…a 500 yr storm is NOT a 500 yr flood event.

    A 500 yr storm event is a given amount of rainfall over a certain duration. This duration can be 1 hr, 2 hrs, 4 hrs, 6hrs…24 hrs, 48 hrs, 72 hrs, etc.

    So not only can there be 1/500 probability storms at multiple locations across the country, there can be multiple 1/500 probability storms at a SINGLE location but of different durations. An intense 1 hr deluge can be a 500 yr storm but is obviously a much different entity than a 500 yr 24 hr rainfall from a slow-moving tropical storm or something.

    A 500 yr flood is theoretically a flood elevation with 1/500 chance of occurring. Obviously a short-duration 500 yr storm won’t generate that for a large watershed. It may not take a 500 yr storm, either. It could be two back-to-back 100 yr storms of long durations with a short period of no precipitation in between, for example.

  22. Australia eventually resolved this problem with a Royal Commission

    Just a gem
    ‘Allianz’s misleading and deceptive insurance websites

    Allianz may have engaged in misconduct through each of the 39 misleading and deceptive statements about travel insurance on its website, each of the 14 representations related to home insurance, each of the 4 representations related to car insurance, 3 related to life insurance and one about boat insurance.

    Allianz also failed to report these breaches to ASIC in a timely way, and in fact decided not to report these breaches.

    The insurer may have also engaged in misconduct through poor risk and compliance management.

    Allianz’s behaviour fell below community standards by:

    Not removing the misleading statements from its websites immediately;

    By not being frank and open in its dealings with ASIC, in particular by failing to inform ASIC of similar breaches or of how far back some of the misrepresentations dated;

    By seeking to manipulate the content of an independent report commissioned by Allianz that was intended to meet a regulatory requirement and be provided to APRA.

    Ms Orr says it is open to find that these problems were caused by the following deficiencies:

    Allianz had inadequate processes to monitor its own and partners’ websites.

    The insurer had inadequate processes for monitoring and closing compliance problems. Monitoring and supervision remains an issue at Allianz.

    Until July this year, Allianz had inadequate oversight of its travel insurance partner AWP.

    The insurer’s culture does not prioritise compliance and risk management.’

    Heavy penalties and governance are in the pipeline.

    This company needs to be Trumped.

    • This is not simple to calculate. I don’t know if you can even decide if an event is a 500-year-event, in fact.

  23. EVERY storm is a “50yr event,” at a minimum.

    When SA went black they blamed a 50yr wind event. By my count, the highest wind speed recorded in DPR SA had been exceeded at 57 weather stations, and presumably more than once at some. A peak of 105km/h (not knots) wouldn’t be noteworthy anywhere except perhaps indoors.

  24. Insurance companies, banks, Governments, the UN, UNICEF, the EU, the environmental movement, the WMO, they’re all in on this crazy scam. It doesn’t matter if we win or lose the climate debate. It won’t stop them as their real objective is a Global Power Grab.

  25. And the Angel tree is not the oldest living thing East of the Mississippi. There are Cyprus trees way older.

  26. It appears that the ad company has accomplished it’s goal. Everyone is talking about Allstate Insurance.

Comments are closed.