Stephen Keller’s fatal boat propeller accident on Lake Tahoe Saturday August 27, 2011 was covered by San Jose California’s Mercury News, because Stephen Joseph Keller DDS was a dentist in San Jose. Mercury News’ August 31st coverage of the boating accident made a major error in reporting the number of annual United States deaths by recreational boat propellers.

Mercury News said there were 3 U.S. boat propeller deaths in 2009 and 1 in 2010 “according to annual data compiled by the U.S. Coast Guard”.

However, the U.S. Coast Guard annual reports actually indicate 25 recreational boat propeller fatalities in 2009, and 27 in 2010.

In the Stephen Keller accident, Mr. Keller, a 46 year old dentist from San Jose, was the passenger of a rental boat docking at Chambers Landing on the west shore of Lake Tahoe about 4 pm Saturday August 27. He was onboard with four more men from the Bay area. The boat operator was trying to get the boat in position to dock. The boat lurched forward, Mr. Keller fell over the stern and was struck by the propeller. Stephen Keller died at the scene.

A San Jose Mercury News August 31, 2011 article, San Jose Dentist Dies After Falling Into Boat Propeller, covering the death of Stephen Keller upped the ante on how wrong the media can get the statistics. The media often misstates total injury statistics, Mercury News misstated total fatalities. The U.S. Coast Guard reports boating accidents as a series of events and their annual statistics are based upon the order in which those events occur. For example, Stephen Kelly fell overboard (Event 1), and was then struck by the propeller (Event 2). Mercury News only reported the number of propeller strikes in which being struck by a propeller was the first event (Event 1). USCG only reported one Event 1 propeller fatality in 2010 and 3 in 2009. But they reported a total of 27 propeller fatalities in 2010 and 25 in 2009.

Propeller Accident Report San Jose Mercury News

Propeller Accident Report San Jose Mercury News

San Jose California’s Mercury News, reported Event 1 data and claimed it represented the total number of propeller fatalities.

Before we go further, I would like to express our sympathies to the family and friends of Dr. Stephen Keller DDS. We followed the boating accident in several news reports and observed an outpouring of love from many for their beloved dentist and friend. We wish them comfort in their time of grief.

We are long used to reporters misinterpreting U.S. Coast Guard’s annual boating statistical report. We describe USCG’s reliance on reporting accidents statistics as series of events further in our USCG Propeller Accident Statistics: A Guide for Reporters and Other Media Representatives page and on our Propeller Accident Statistics page.

Most propeller accidents are Event 2 or Event 3 accidents, but reporters keep printing Event 1 data from USCG’s annual boating statistics reports and erroneously claiming they represent the total number of propeller injuries or fatalities.

We started trying to contact Mercury News the morning of September 1st asking them to correct their error. Details of our efforts are in the “Our Efforts to Get Mercury News to Print a Correction” section below.

Below are direct copies of the propeller accident statistics pages from USCG’s 2009 and 2010 annual boating statistics summaries. We highlighted the propeller accident row and total fatalities column in red to make them easier to read.

USCG 2009 propeller accident fatalities

USCG 2009 propeller accident fatalities

USCG 2010 propeller accident fatalities

USCG 2010 propeller accident fatalities

It is very clear there were were a total of 25 propeller fatalities in 2009 and 27 propeller fatalities in 2010 per U.S. Coast Guard’s BARD (Boating Accident Report Database).

If anybody does not believe the pages above, you can view the USCGs 2009 and 2010 annual boating statistics summaries directly from the U.S. Coast Guard Office of Boating Safety web site. Page 35 of the 2009 report indicates there were 25 propeller fatalities in 2009, and page 35 of the 2010 report indicates there were 27 propeller fatalities in 2010.

And if you don’t believe that, you might follow some of the news reports on our 2009 and 2010 propeller accident media reports pages with a red D beside them (U.S. deaths). There are a whole lot more than three of them on each year.

Mercury News reports there was only one propeller fatality in whole U.S. in 2010 and three in 2009. They mistakenly came to those statistics from the 2010 Primary Event (Event 1) table below and the corresponding 2009 table not shown.

USCG 2010 BARD Propeller Event 1 ONLY Fatalities

USCG 2010 BARD Propeller Event 1 ONLY Fatalities

Below is a copy of USCG’s description of Primary Accident Type from their 2010 annual boating statistics report:

USCG Definition of Primary Accident Type

USCG Definition of Primary Accident Type

Note the Primary Accident Type Definition above even lists the specific title and page number of the table above it. The table Mercury News incorrectly used to represent all propeller accidents. It does NOT!

Mercury News’ reporting error has since being echoed to their sister papers and around the world. Among my favorites is in a Corvette car forum where the Corvette guys copied the news report and bolded the statistics.

The Corvette forum goes on to discuss personal experiences with propeller injuries, several voice surprise at the low statistics, and some expresses a desire to avoid propeller safety regulations based on the low number of reported fatalities. A few of their comments are below.

“I’m guessing a lot of people get injured, but don’t die………don’t know, just conjecture on my part.”

“Surprised more people don’t die from boating accidents.”

“Wow, that means my friend was one of two who lost their life to a boat propeller in 2009. Kinda hard to believe.”

“Yeah, a couple of people manage to die this way per year. That certainly screams for more government intrusion into our lives.”

The Corvette guys are car guys and they seem to question the statistics. Is everybody asleep at the wheel at Mercury News? Its really pretty amazing how many Corvette guys chimed in over just a few hours with personal experiences with propeller accidents. That could be a message in itself about under reporting of propeller accidents to USCG.

Errors like those propagated Mercury News continue give boaters a false sense of security and encourage stupid comments in forums from readers who are misled by the under reported propeller accident counts provided to them.

Also please note the counts we provided are for U.S. Coast Guard reported accidents meeting their criteria. Many propeller accidents go unreported, many are reported in BARD but not classified as propeller accidents, and many are reported, but rejected for not meeting certain criteria.

Dr. Stephen Keller’s obituary reports his memorial services will be held at St. Christopher Church in San Jose on Monday September 5, 2011 at 11am.


Our Efforts to Get Mercury News to Print a Correction

A log of our efforts to get a correction printed on the web published Mercury News story to report the correct propeller accident statistics.

  • September 1st the morning of – I became aware of the news report in our ongoing efforts to daily review boat propeller accident reports in the media and I immediately noticed the error.
  • September 1st 11am I emailed Mike Rosenberg (the reporter that wrote the story) early today In that email I: described USCG’s Event reporting, pointed him to the correct statistics, pointed him to the annual statistics on our site and to our reports of media coverage of propeller accidents and supplied him an email address for a boating accident statistics person at USCG.

    The actual email is below

    Subject: San Jose Mercury News: San Jose dentist dies after falling into boat propeller
    From: XXXXXXX@propellersafety.com
    Date: September 1, 2011 11:09:01 AM CDT
    To: XXXXXX@mercurynews.com

    From: Gary Polson
    Propeller Guard Information Center

    Hello!

    Thank you for covering the Stephen Keller propeller accident, however there is a MAJOR error in the annual United States propeller fatality counts you supplied (1 in 2010, 3 in 2009).

    The actual USCG BARD reported propeller fatality counts were 25 in 2009 and 27 in 2010.

    USCG reports accidents as a series of three events, like:

    Event 1. Struck object
    Event 2. Man overboard
    Event 3. Struck by propeller

    Propeller accidents are most often Event 2 or Event 3 accidents.

    Many of the pages in the USCG annual statistical report are for Event 1 accidents only.

    This process is explained in depth in our USCG Propeller Accident Statistics: A Guide for Reporters and Other Media Representatives
    http://www.propellersafety.com/propeller-statistics-media-guide/

    The “Prologue” near the bottom of that page shows the 2009 data.

    It appears your errors were created by using State of California (ONLY) Event 1 data.

    So instead of for the whole US, you only have California accidents, and instead of having all the California accidents, you only have the Event 1 California accidents.

    For example, if you go to page number 58 of USCG’s 2010 annual report, you will see 3 people fatally struck by propellers as EVENT 1 in the state of California in 2010. Although your article listed 1 in 2010, I suspect your error was related to this approach.

    But if you go to page 35 of the same report, you will see there were a total 27 reported propeller fatalities in the U.S. in 2010 per USCG’s BARD database.

    USCGs 2009 and 2010 reports are available from their site at:
    http://www.uscgboating.org/statistics/accident_statistics.aspx

    Please correct that article as soon as possible as it is already being copied around the Internet presented these false statistics. Boaters will be thinking they have no risk, when risks are really much larger (note even starting to count all the non reported fatalities or those rejected for many reasons.)

    Here is an example of it being copied elsewhere and they even BOLDED the statistics that are wrong.
    http://forums.corvetteforum.com/off-topic/2904317-man-dies-from-boat-propeller.html

    A google search for “San Jose dentist dies after falling into boat propeller” shows your misleading stats are already echoing all over the world. Please correct the original article immediately. Not only is it being copied, but the copies are being copied.

    If you don’t believe me, please contact XXXXXXXXX the lady that keeps the boat accident statistical records for the U.S. Coast Guard Office of Boating Safety at She can provide you the national statistics.

    Again, thanks for covering the accident, but please correct the stats as soon as possible

    Thank you in advance for correcting this error.

    Gary Polson
    Propeller Guard Information Center
    http://www.propellersafety.com

    (XXX) XXX-XXXX

  • September 1st late in the afternoon the San Jose Mercury New article on Stephen Keller’s propeller accident had been replicated many times online by other news agencies and web content generators. Their errors have been perpetuated a hundred fold.
  • September 1st I tried to post a comment on the actual Mercury News story to point viewers to the correct propeller accident statistics. I spent considerable time creating it, but the system would not accept the post (probably closed to comments after stories are so many hours old).
  • September 1st I contacted SPIN (Stop Propeller Injuries Now) and relayed the error made by Mercury News and the problems we were having in contacting anybody at Mercury News and enlisted their help as well.
  • September 1st having received no response from my first email to Mike Rosenbery (the reporter), I created this post as a tool to try to explain the correct statistics to the San Jose Mercury News, and to try to combat the false statistics they continued to spew.
  • September 1st about 6:30 pm I sent Mike Rosenberg a second email that included a link to this page. A copy of that email is below.
    Subject: Re: San Jose Mercury News: San Jose dentist – PROPELLER STATS
    From: XXXXX@propellersafety.com
    Date: September 1, 2011 6:21:59 PM CDT
    To: XXXXXX@mercurynews.com

    From: Gary Polson
    Propeller Guard Information Center
    Hello!

    We see no updates of the errors in your news coverage concerning the annual number of boat propeller fatalities per our earlier email (reference the 31 August Stephen Keller dentist propeller strike article).

    Misinformation spreads like wildfire online. We try to contain errors in media reported boat propeller accident, injury, and fatality counts as soon as possible.

    Your lack of response has caused us to create a post on our site making the correct data available online and to give us a page to call to the attention of the almost countless sites that have echoed your misinformation, before it gets echoed even further.

    Media Under Reporting of Propeller Fatalities Hits New Low – San Jose Mercury News
    http://www.propellersafety.com/2947/propeller-statistics/mercury-news-misreports-propeller-fatalities/

    If we don’t see a response by tomorrow we will be contacting your supervisors.

    I do not like being this aggressive, but we have never seen stats that low printed before. When we try to correct others normally mistaking full US Event 1 data for the total number of US fatalities they will be pointing to your article and its copies and saying not only are we (Propeller Guard Information Center) wrong, but they must have been wrong themselves because the actual stats must be the ones you (Mercury News printed) with low single digit fatalities per year.

    This has to be stopped now. Thank you for your assistance and I apologize for the harsh tone.

    Again, if you don’t believe our statistics, contact the U.S. Coast Guard boating statistics person at the email address I supplied earlier (or your state boating law administrator).

    Have a nice day, and thanks for your attention to this matter.

    gary polson
    propeller guard information center
    http://www.propellersafety.com/

    a copy of my previous email was also attached

  • September 2nd I tried calling Mike Rosenberg (the reporter) (number listed on their site was answered by a recording with someone else’s name), I tried calling the guy listed online as their online editor in the newsroom (guy answering the phone says that guy does not work there anymore). About 4:15pm I reached a lady at Mercury News. I explained the error in the propeller statistics, that I had sent two emails to Mike Rosenberg, and how to get to this page. She was unable to view it from where she was. She said Mr. Rosenberg would be in to work in a few hours and she would pass our phone number and info on to him.
  • September 2nd about 6:30 pm I called the office again. They said Mr. Rosenberg was out and they would try to get him to call me in an hour or so.
  • September 2nd about 9pm I called the office again. A guy said Mr. Rosenberg was there and he would ask him to call me.
  • September 2nd shortly after 9pm Mike Rosenberg of Mercury News and reporter of the story with the wrong statistics called. He was very cordial and willing to visit about my complaint. I asked him if he had read my emails. He said he had not had time yet. (He is a night reporter and comes in in the evenings). I explained USCG’s Event 1, Event 2, Event 3 reporting methods. I learned he had wanted to report accident statistics for those accidents in which propellers were the primary cause of death, so he elected the “Primary Accident” table from USCGs annual report. I tried to explain to him the “Primary Accident” means Event 1 to USCG. Even Mr. Keller’s accident would not be in there because he was killed in Event 2. (Event 1 = man overboard, Event 2 = struck by propeller). I encouraged Mr. Rosenberg to contact the USCG’s Office of Boating Safety statistics person I had supplied him an email address for earlier if he did not believe me.

    I told him it is very hard for us to convince reporters like him of the correct propeller fatality statistics with so many articles like his out there with the wrong data. We want him to correct his article. He said that would be up to his editor and after some discussion, he gave me a name of a web editor for that particular story. I told him I was certainly not trying to ride over him, but if we didn’t see the proper numbers in a few days, I would be contacting the editor.

    When I kept bringing up the 25 and 27 USCG reported propeller fatalities from 2009 and 2010, he said he did not want to get into the minutia of it. He just wanted those deaths that were primarily caused by propellers. It seems like he still does not totally understand USCG’s Event system of reporting boating accidents.

    I tried to be nice, he tried to be nice, and he seemed honestly interested a bit, but also interested in moving on to his next story (as all reporters should be). He did say he would keep the Event system in mind the next time he covered a boating accident.

    I thanked him for his time.

    I am not really expecting to see a correction from him or his editor without some more pressing by us or others. It looks like the real problem may be in making the USCG annual statistical report impossible to misinterpret.

  • September 3rd – about 10pm I relayed to SPIN we had visited with the reporter but it did not look promising.
  • September 3rd – about midnight, got the idea reworking this portion of USCG’s statistical report might be a good job for a university class project in communication and/or safety somewhere. We may try to create a group of post of potential class projects here as our other theories about creating a place to promote propeller safety class projects has yet to get underway.
  • September 4th – about 7:45 am, we notice another propeller fatality reported in the Mercury News this morning (Martha Acevedo on Lake Havasu on Saturday 3 September). I wonder if any San Jose Mercury News readers think it might be a bit odd to have two local propeller fatalities reported in their paper that happened on consecutive Saturdays when their paper reported there was only one in the whole United States last year.

    What if the deceased lady read Mercury News’ report last week and thought being around boats sounds pretty safe with only one prop fatality a year, and now she is killed, mislead by the wrong statistics printed by Mercury News?

  • September 4th about 6pm I spent some time composing a response to the coverage of the Lake Havasu propeller fatality this week being covered on Mercury News. I noted it seems odd that they would have fatal propeller accidents there on the last two consecutive Saturdays when there was only 1 in the whole U.S. last year per their misquote of USCG statistics. I tried to point them to their coverage of the Stephen Keller accident and to this post and encouraged the news staff to print the correct statistics. After trying to post my comments unsuccessfully a few times, their system finally just blew it away like they keep blowing off my requests for a correction.
  • September 5th about 9pm I continue to update this page. I just added the Primary Accident Type table from USCG’s 2010 annual report and copied USCG’s definition of “primary accident type”. Mercury News still refuses to post a correction. They seem not to believe our information. We will be escalating our efforts after the holiday (is Labor Day today).

We encourage our readers to contact Mercury News and ask for a correction of the statistics on the September news story by Mike Rosenberg titled “San Jose Dentist Dies After Falling Into Boat Propeller”. Please feel free to point them to this web page. We honestly don’t have a suggested contact point. It seems pretty disjointed there every time we call, plus they don’t read our emails. You can try your luck with their contact Mercury News page.

We also encourage you to leave comments below and we are certainly open to reasonable suggestions about how to get their attention.

Plus if anyone at all out there still believes the Mercury News statistics we would love to hear from from you.

We will stick with this and we will eventually succeed in getting them to print a correction. But it may take a while. We welcome anybody’s assistance in the effort as long as you play nice.

Print Friendly, PDF & Email