This represents a view into the 1999 through 2013 Iron Butt Rally statistics as best I can aggregate them. There are over 800 observations over 8 rallies. This reflects the data available, and a long time follower of the rally knows the history goes back another decade. For the purposes of this think of it as “modern history”. I thought I might present this data at a national Iron Butt Association meet or maybe as a magazine article but I’m not sure it is up to either. I’m not going to go into detail about each graphic but what follows is a series of charts and tables that depict some information. The raw data is available from a few different locations. Click on any of the graphics to see a larger representation. For those that care I could have alphabetized or numerically sorted all of the graphs. Feel free to do that in your own time if you wish.
First up we talk about aggregate data across the rallies. BMW and Honda over the years have dominated the rally. There have been some one-off rides like the Boss Hoss. Notably absent in my view is KTM.
This graphic depicts the trends over time. Though in the previous graphic we see BMW owning the landscape the reality recently is that other brands have caught up. Mainly they have eaten away at the BMW and Honda hegemony showing that other bikes might be well suited to the Iron Butt Rally. Rider selection is primarily shown at this point and riders chose BMW, Honda, and Yamaha by far over the other brands in the 2013 Iron Butt Rally.
We choose bikes and build bikes that we think will go the distance or that we can be comfortable on. Over the years particular myths have arisen around particular brands ability to go the distance. The rear-drive failure prone BMW, the frame breakage issues of the Honda Goldwing platform. Whether in the rally or in the common mind the myths remain. The issue with simply applying or counting of DNF bikes as failures is as follows:
- Many of these motorcycles are high mileage examples far beyond standard life spans.
- Simple counts don’t take into consideration the population of the bikes
- The mix and types of bikes have changed
- Brand identification is only one measure as model and type (Kawasaki Concours v. Ninja 250) make examination by brand alone a less than suitable exercise
That being said, you all know you want to know who fails out the most. The answer is BMW, Yamaha, Honda, and a close following fourth, Kawasaki. Let the flame wars begin.
Looking at a percentage of the bikes that failed from the top three brands over the years shows some interesting values. Where the bike failures are fairly close assumptions could be made. The first assumption being that the rally format itself and not necessarily the bike failed. That would be a good thought but may not be purely true. The chart shows two interesting pieces of information. The first piece of information is the downward trend with slight uptick of the BMW brand. BMW was getting much more reliable then had a slight backslide. I haven’t had a lot of time to dig into the Honda failure skyrocketing in the 2013 but it is interesting. Over half the Honda motorcycles starting in 2013 failed. I’m not sure if that is artifact of the data, testament to the riders, or an error in the analysis. Anything is possible.
For many people the question isn’t so much about failure but winning. What do the top 10 riders finishing the Iron Butt Rally ride? Over the decades the top riders have overwhelming been on BMW motorcycles. Once again looking at the data across the decades doesn’t tell the tale of what is currently happening.
To get a better idea we look at the trend data. With one little blip of a Harley Davidson and a Triumph in the top 10 finishers the rest are BMW, Honda and Yamaha. Most assuredly BMW has been sliding out of the top 10 and Yamaha has been sliding into the top 10. With half or more of the top 10 finishers riding Yamaha in 2013 that is something the Yamaha riders can tell everybody. 2009 and prior the ride of choice in the top 10 was definitely BMW. In 1999 and 2001 Honda and BMW basically split the top 10 spots. Honda again took half the top spots in 2011 only to fall to abysmal 2 in 2013.
So what about BMW failures? In the 2007 iteration of the rally 40% of the BMW brand bikes failed to finish. We know that DNF does not have to mean break down and grenaded motorcycles. The rider could have finished the miles and simply not had the points or missed a checkpoint. That is not the motorcycles fault. BMW did better than average both in 2011 and 2013, but still had a significant uptick in failure for 2013.
I know that the Honda guys want to know all about the dramatic increase in mortality of the Honda brand in 2013. I think, but have not dug into the data, that the increase is partially caused by riders dropping out for other than bike reasons. If there is any interest I can expand the data evaluation and pull the DNF reports from the rally and see what is happening. There is always the possibility that there is a coding issue too.
What follows now is an evaluation of a truism of the Iron Butt Rally community. We all know it. The best riders ride well, they ride efficient, and they get more points per mile. In other words this is proof that it is about efficiency and not speed, or mass miles to win.
A hypothesis might be H: A better finisher position correlates with a better point per mile efficiency valuation.
A points per mile formula applied to finisher position should show a less efficient value as the participant position increases. Looking at the data across each of the rallies shows this premise to be true. We would fail to reject the null hypothesis. There are some interesting points but the following charts and discussion will talk about interesting data more than detailed analysis. The Y axis is points per mile, and the X axis is finisher position starting out with 1 and going up.
The 1999 rally has a few outliers and pulling their ride reports might be interesting. In general (the only relevant test) the trend holds that position placement is predicated by an efficient ride.
In 2001 there were some absurd bonuses available to the riders and a sharp difference was apparent. If you took the chance on the many points (even if many miles) you were extremely efficient in the sense of lots of points per mile ridden. This is the first place that an absurdity is pointed out. If the rally master put a quadrillion points on the table, and you had to ride 100K miles in 11 days (approximately seven times what records suggest is possible) you would be targeting the most efficient ride in history. It also is not possible under the current and previous Iron Butt Rally rules. You’d have to average something in excess of the current motorcycle land speed (376.3 mph) record 24 hours a day (9031.2 miles). Skip fuel stops, and never take a rest break. Space shuttles don’t count as motorcycles.
In 2003 there were some interesting clusters and flatness to the curve at the middle of the pack. Less outliers checking the data shows that 2003 was a very competitive time to be the middle of the pack.
The 2007 rally was a high point rally. Fortunately the methodology takes that into account. Here as in other low efficient outliers there are people who had a lot of points but some major issue resulted in getting smacked at the scoring table. Last place with low points to miles ratio strongly suggests… a bike change or really late arrival to the finish line.
Similar results in shape and skew to the 2007 rally.
I really wanted to see what this rally looked like. Some people found some VERY efficient routs. The actual points to mile ratio was very tight. This is the most interesting graph in the whole set. The premise of this rally was to ride to all 48 states and at particular times grab a state capitol when it made sense. This appears to be a VERY competitive rally. I will be digging into this rally in the future to figure out how people routed.
The rally I rode in was 2013. This was another tight rally with the value of points being less determinant than the actually efficiency of the rides. A few apparent outliers (I didn’t actually test it) give some indication of rides gone awry.
The money ball of long distance riding is interesting to me because I actually ride. I think that some people will get a kick out of digging into the data. I hope you enjoyed looking at the data and let me know if there is something you would like to see in the future.
The dataset: IronButtRally19992013
Caveats: I was just having some fun here so no warranty is expressed or implied. You didn’t do this first so if you want to argue with it go do the analysis yourself. Trolls and such not desired, solicited or asked for.