Percentage of Success vs Population On Trail Over Time - Math and Estimation

[Gambit McCrae]

We can take this concept in increments of a decade. How do you think the % of completion has changed as the decades click by. I want to look at it from a percentage perspective to eliminate increases in population. We all understand that population on trail increases as time goes on, so to say that there are "400 more completions in 2018 vs 2017" doesn't really tell us anything more then a whole lot more hikers attempted it.
What I want to look at is this:
In 2018 4000 hikers started, and 400 finished so 10% completed the hike.
In 2008 2000 hikers started and 400 finished so 20% completed the hike.

Thus over time, more people attempt, but a fewer percentage are successful. I understand that this will most likely be speculation and opinion which is fine but if anyone is capable of digging up concrete numbers that would be fun as well.

[Time Zone]

I don't have those numbers (perhaps someone does - the ATC? Amicalola Falls SP?) but to tie it in to your "going soft" thread, I'd be cautious about reading too much into any changes in completion rate. In your (illustrative?) example, you have 2000 more hikers in 2018 than 2008. Suppose most of those don't intend to complete the trail fully, or that most of them are drawn from a sub-population with very different fitness and commitment levels than the first 2000 hikers of either year. We wouldn't expect as many of those second 2000 to finish.

[Slo-go'en]

Unfortunately, hard numbers are impossible to come by . There is no requirement to register when starting the trail, not everyone stops by the ATC on their way by to get a photo and Baxter doesn't keep track of how many started from Georgia and how many started from Monson.

So, its guessing game with a fair amount of extrapolation. Weather is a big factor like last year which sent a lot of people home.

[HooKooDooKu]

Seems like a number I've heard tossed around is that only about 20% of those that attempt a thru hike complete it. I haven't heard anything that suggests that number has changed over the years.

If any hard numbers can be found, I would not be surprised if 20% has remained constant over the last 20 years. For one thing, we are talking percentage, and when it comes to certain difficult tasks, only a certain percentage of the population has the ability to perform that task. It would seem the resources to aid you in completing the AT have increased over the years, in theory making the hike easier to complete. But I think that is getting offset by more people trying to do a thru with less experience than those that attempted it in the past.

The idea is sort of what I hear when it comes to deaths/injuries due to traffic accidents... As experts have made driving and driving conditions safer, the increased safety seems to usually get offset my more complacent drivers doing more dangers things. The combination of better safety features and wosrt drivers mean that the frequency of deaths and injury have stayed relatively flat over the years.

[Dan Roper]

Interesting question, Gambit. I don't know the answer but I suspect the completion rate might drop slightly as participation increases. Our population is getting heavier rather than lighter, and the trail population is probably graying, so more backpackers would probably mean they're slightly less fit, on average.

In the 1980s, I began hiking in the Cohutta Wilderness Area on the Georgia/Tennessee line. My perception is that usage of the trails hasn't changed very much, including the most popular Jacks River Trail. Since the population in North Georgia and adjacent Tennessee has increased substantially over these 30 years, why hasn't trail usage followed suit?

A few years ago, I met a college student birdwatching at a local natural area. She mentioned offhand that many in her generation like to drive up to a point of interest, exit their car, walk around for five to fifteen minutes taking pictures, and get back in their car to leave. Many of us have probably witnessed that kind of behavior at Newfound Gap, the Grand Canyon, you name the place. It's always been that way, but the college student believed that her generation had really bought into that practice. The more tethered we become to electronics, the less likely we are to leave them behind (or risk losing connectivity, even if its only sporadic).

When I was young (back in the '70s), kinds were always outdoors playing neighborhood football or baseball or basketball. I haven't seen kids doing that in more than ten years, and that one instance was the only time in the past 25 years. All activities now seem to be organized team sports. Most of our entertainment is electronic and indoors.

It's always been a modest percentage of the population that enjoyed untethering to enter "the wilderness." I suspect that percentage is declining over time. Raw numbers might be stable or increasing a bit, but the percentage is probably decreasing a bit.

[moldy]

The completion percentage has gone up over the decades. It has to do with many factors. Improvements to every aspect of a thru hike from equipment to food to the trail itself encourage us all to finish. The improvement to trail guides, communication and trail services, hostels, shuttles and trail magic, make it easier to finish. Now, thanks to advanced information services like Whiteblaze, Trail Journals, Youtube, blogs and computer programs we have hikers who start a thru hike and Know what they are getting in to, before they ever start. Numbers will keep going up for starters and finishers....

[MuddyWaters]

Well the first 25% cant get out of Ga.

This alone is mind boggling. How anyone can prepare for a year, spend hundreds to thousands on gear, and not be able to make 80 miles.

Theres lots of dreamers. People in poor shape, flawed expectations, or not really committed to something thats actually ... hard work.

I suspect as much as 1/2 may fall in this group, drop out before Va.

Has social mefia attracted more of these? Surely yes.

But it also helps them succeed too

[rickb]

From the ATC website:

[Time Zone]

Well the first 25% cant get out of Ga.

This alone is mind boggling. How anyone can prepare for a year, spend hundreds to thousands on gear, and not be able to make 80 miles.

My hypothesis: they do not train in comparable elevation gain/loss conditions. Why not? I'd guess because that's not the topography where they live.

[Slo-go'en]

Wonder what happened in 2017 to have only have 715 out of 3,839 finish?

I know what happened to me, I caught a nasty chest cold on the way to Florida to visit my Dad, which got worse when I hit the trail in Georgia and barely make it to Hiawassee alive. End of April, cold and wet. But I was only planning on going to Marion, VA anyway, so I don't count.

[imscotty]

RickB,

I was surprised to see the success rate going down the past few years. May have something to do with the electronically wired generation getting more easily bored and distracted?

I am pretty sure I've seen numbers from the 70's, 80's and 90's indicating an upward trend in success rate. I always figured this was due to better information available and lighter equipment. Perhaps the success rate has peaked, and we are seeing the other side of that slope in your posting.

[rickb]

If one were to google up the final reports from each season’s Kennebec Ferry report and compare the number of NOBO thru hikers logged there vs the number of reported thru hikes, who knows what conclusions could be drawn.

I just pulled up one year:


2015 Kennebec Ferry Statistics



We had a record number of hikers this year.
Southbound through hikers, 307
Northbound through hikers, 1,007
Total hikers using the Ferry Service during the 2015 season, 2,181
Confirmed Forders/Swimmers, 43


In this year 1007 Northbound thru hikers crossed the river, but only 677 reported having a thru hike to the ATC.

Why such a big difference?

Did that percentage hold constant over the years?

[Slo-go'en]

In this year 1007 Northbound thru hikers crossed the river, but only 677 reported having a thru hike to the ATC.
Why such a big difference?
Did that percentage hold constant over the years?

Either 330 didn't make it the rest of the way, or didn't bother to report they completed the trail or incorrectly identified themselves as through hikers when crossing the river. Or a little of each.

[rickb]

Either 330 didn't make it the rest of the way, or didn't bother to report they completed the trail or incorrectly identified themselves as through hikers when crossing the river. Or a little of each.

Agreed.

Or they considered themselves to be thru hikers — in the popular sense of the word — but not qualified to sign and submit the ATC’s 2000-Miler application.

As hikers’ values evolve I would expect that subset of hikers has changed over the years.

If so, the percentage of hikers who have enjoyed a great walk from Maine to Georgia over the years may have changed more or less than the AT figures suggest.

[Gambit McCrae]

From the ATC website:

The NoBo trend was as I guessed which is why I wanted to base the discussion off of percentages.
IMO Summary: As time goes on, resources make the trail logistically and ergonomically easier to hike, however society is becoming softer and a result is a fewer percentage of people complete the trail each year from a "total that attempted" perspective.

Thanks for that data Rickb!

HOWEVER I also believe the data is skewed by the ever growing population of people who set out and never make it out of GA. I would like to see the number of thru hikers that register for the smokies vs how many that started in GA

[imscotty]

MapMan is great at analyzing and providing these kinds of numbers. Perhaps MapMan can weigh in on this?

[Slo-go'en]

HOWEVER I also believe the data is skewed by the ever growing population of people who set out and never make it out of GA. I would like to see the number of thru hikers that register for the smokies vs how many that started in GA

A scientific study is needed. At the summit of Springer, every declared thru hiker is issued an "AT passport" with an imbedded RFID chip. The chip is scanned at key locations along the trail to track progress. There would be no personal info, just a number, although age and gender would be useful metrics to associate with the hiker number.

[Gambit McCrae]

MapMan is great at analyzing and providing these kinds of numbers. Perhaps MapMan can weigh in on this?

I have been anticipating his arrival....

[Gambit McCrae]

A scientific study is needed. At the summit of Springer, every declared thru hiker is issued an "AT passport" with an imbedded RFID chip. The chip is scanned at key locations along the trail to track progress. There would be no personal info, just a number, although age and gender would be useful metrics to associate with the hiker number.

I like it...a lot...brings out the manufacturing organizational side of me. ORGANIZE the hikers. Process the hikers. Document the hikers LOL

[Slo-go'en]

HOWEVER I also believe the data is skewed by the ever growing population of people who set out and never make it out of GA. I would like to see the number of thru hikers that register for the smokies vs how many that started in GA

I would think the majority of hikers register for the smokies just before starting rather then trying to do it along the way. What you really needed to know is how many registered but never put their permit into the register box at the entrance to the park. That would be a decent indication of how many never made it that far.

But seeing that the box is usually stuffed solid with permits, it's not clear if they are ever collected and counted.

[Gambit McCrae]

I would think the majority of hikers register for the smokies just before starting rather then trying to do it along the way. What you really needed to know is how many registered but never put their permit into the register box at the entrance to the park. That would be a decent indication of how many never made it that far.

But seeing that the box is usually stuffed solid with permits, it's not clear if they are ever collected and counted.


I can certainly agree with that. I was hesitant about registrations done prior to start dates but didn't think it would be that many. I know I would personally wait until Fontana village before I registered but I can see how some ones anxiety of having all their ducks in a row would require them to register before setting out.

[rickb]

One factor that complicates a comparison of completion rates over the years is definitional.

Early thru hikers were required to submit a detailed written report that was reviewed by the ATC before they would be listed as 2,000 milers.

In my day, thru hikers were asked to mail in a trip report to the e ATC— but one that would not see much or any critical review. In my case, my dad suggested that I send one in but I explained to him that I didn’t see value in certificates — he sent in a local newspaper article without my permission and I got listed anyway (thanks dad). The bar had been lowered.

For the past couple decades or so thru hikers are asked to sign a specific form attesting to thier accomplishement — with at least one change to the ATC’s written requirements along the way.

But even more importantly, people’s values change - with regard to how they elect to hike the Trail and define for themselves what constitutes a complete hike (a good thing, IMHO).

Also to the importance of signing and sending in the ATC paperwork— and to interpreting their rules literally or in spirit.

I am not sure of the net impact on the rates each year, but I do believe it to be substantial when looking at the numbers across decades.

[map man]

I've been reading on the ATC website what their estimates have been for thru-hikers starting, thru-hikers completing, and thru-hiker finishing percentages since around 2005 (when I first became interested in backpacking on the AT). They always seem to list the last seven or eight years of hiker classes, so back in 2005 I was seeing their stats for, say, 1998 through 2004, in 2010 I was seeing numbers for 2003 to 2009, etc. The percentage of completing NOBOs has varied some from year to year but always seems to be in the 20-30% ballpark. My memory is that in those earlier years the numbers were more likely to be in the 20-25% range while in later years it's been 25-30%, until the 2016 and 2017 years when their estimate went down to 20 and 19%. I have no idea if those two years are an aberration, a trend, or a change in estimating methodology.

As ATC's concern with NOBO thru-hiker over-crowding in the Spring has increased their efforts to make an accurate starter count at Amicalola and Springer have increased too, it seems. In recent years there have been Ridge Runners in these locations raising awareness of LNT and other issues with departing hikers and they have been doing some head counting on behalf of both the Georgia ATC and the national organization while doing those duties. Or so I understand (I reported myself as a section hiker to a Ridge Runner at Amicalola in 2016 at the start of a section hike that year).

But while that starter number has arguably become more accurate, their methodology for counting finishers has not kept up. They are still using the numbers they get from hikers sending in their 2000 miler certificate applications, and this always undercounts finishers because some hikers choose not to apply for the certificate.

A clue to how significant this undercount is has been provided by the staff at Baxter State Park in Maine. As their own concerns about AT hiker overcrowding have increased they have become very diligent about counting AT hikers in all categories from NOBO and SOBO and flip-flopping thru-hikers to section hikers. So BSP officials have reported that in both 2016 and 2017, for instance, there were over a thousand NOBO thru-hikers in their park. That is a really big difference from the 685 and 715 NOBO thru-hikers from those classes who reported their hike to ATC by 2018. And this flaw in undercounting finishers has existed for as long as the ATC has been using 2000 miler self-reporters as their estimate.

So to summarize, a) it looks like there has not been a dramatic increase or decrease in completer percentages over the last 20 years according to ATC estimates, and b) there is reason to believe that completer percentages over that time have been higher than the ATC has been reporting because they undercount finishers.

One more set of statistics to chew on. The ATC back in the first decade of the 21st century used to report hiker counts for Neels Gap and Fontana and Harpers Ferry, as well as Springer and Katahdin. They still report Harpers Ferry but discontinued Neels and Fontana numbers some years back. I know that they got those Neels Gap numbers because they asked the folks at Mountain Crossings to do a count for them, though I don't know how they got the Fontana numbers.

At any rate, those numbers showed attrition rates varying some year to year but usually were around a 10% attrition rate by Neels Gap, around a 20% attrition rate by Fontana and around a 50% attrition rate by Harpers Ferry. And since the Harpers Ferry rate has continued to be reported and has stayed near that 50% number over the last twenty years pretty consistently, I surmise that it's unlikely the early attrition rates at Neels and Fontana have been trending a lot higher or lower over that time period either (and I have some skepticism about anecdotal claims about early attrition rates being dramatically higher than the ATC estimates).

[rickb]

A clue to how significant this undercount is has been provided by the staff at Baxter State Park in Maine. As their own concerns about AT hiker overcrowding have increased they have become very diligent about counting AT hikers in all categories from NOBO and SOBO and flip-flopping thru-hikers to section hikers. So BSP officials have reported that in both 2016 and 2017, for instance, there were over a thousand NOBO thru-hikers in their park. That is a really big difference from the 685 and 715 NOBO thru-hikers from those classes who reported their hike to ATC by 2018. And this flaw in undercounting finishers has existed for as long as the ATC has been using 2000 miler self-reporters as their estimate.
In 2017 the total niumber of AT Hikers registered by BSP was over 3000, but did not reach the 3150 threshold that would have required some to leave the Park before submitting Katahdin.

I know they break down the total to some extent, but I have missed seeing data that shows how many of the registered NOBOs walked all the way from GA vs. how many started in Monson — or somewhere in between.

Do you know if this information is published on line for any year?

[map man]

In 2017 the total niumber of AT Hikers registered by BSP was over 3000, but did not reach the 3150 threshold that would have required some to leave the Park before submitting Katahdin.

I know they break down the total to some extent, but I have missed seeing data that shows how many of the registered NOBOs walked all the way from GA vs. how many started in Monson — or somewhere in between.

Do you know if this information is published on line for any year?

I, too, have looked a few times in recent years for those numbers on the BSP website and not found them. I only remember them because Teej posts those numbers from time to time here on WB. Teej is the man here when it comes to all things BSP. I will look through some of his past posts when I have time this evening.

A few years ago I speculated that some BSP-reported thru-hiker numbers might include anyone coming in from at least as far as Monson since that is their criteria for allowing hikers to stay at the Birches. But Teej corrected me and posted that they do keep track of "true" thru-hikers in their number counts.

[peakbagger]

I section hiked the AT over mostly a 10 year period. I never signed in at Amicalola Falls state park, the ATC headquarters was not open the day we went through Harper Ferry and didn't sign in at the Ranger Station at Katahdin. About the only official trace besides shelter registers would be the Kennebec River Crossing.

I believe the number of starts is inflated at Spring as ATC is pushing for folks to register their hike for statistical purposes. I expect the BSP numbers are pretty good as thru hikers need to check in at the ranger station when they arrive at KSC. There is (or was) a separate sign in register for day hikers out of the parking lot and the trailhead and the ranger cabin is bit out of the way for a day hiker to go over and sign in so its likely that only thru hikers make the extras steps to sign in at the "last register"

[full conditions]

I section hiked the AT over mostly a 10 year period. I never signed in at Amicalola Falls state park, the ATC headquarters was not open the day we went through Harper Ferry and didn't sign in at the Ranger Station at Katahdin. About the only official trace besides shelter registers would be the Kennebec River Crossing.
I believe the number of starts is inflated at Spring as ATC is pushing for folks to register their hike for statistical purposes. I expect the BSP numbers are pretty good as thru hikers need to check in at the ranger station when they arrive at KSC. There is (or was) a separate sign in register for day hikers out of the parking lot and the trailhead and the ranger cabin is bit out of the way for a day hiker to go over and sign in so its likely that only thru hikers make the extras steps to sign in at the "last register"

My own experiences echo yours - I successfully thru hiked in '76 and '80 and finished up a section hike project in 2003 and never left a record with the ATC or anyone else for that matter (nothing against ATC - just didn't want the certificate) and met plenty of other hikers who I know were successful but have never seen their names on the '76 or '80 rosters. On the other hand, we also don't have a good handle on the numbers of aspirants who never registered at Springer or anywhere else and who quietly quit the trail for whatever reasons. I hiked with a guy named Barry in '76 for a couple of days and I thought he was doing fine and one morning as we got down to Hogpen Gap he said "well, this is where I get off". I said "you got to be kidding me - just yesterday you were saying how much you looking forward to getting to New England" to which he just shrugged and told me that he was terribly homesick and missed his girlfriend and with that he started walking down the hill toward Helen. I think Barry was a perfect example of someone who got on the trail w/o any registration or being included in the statistics and likewise dropped out without being counted.

[Maineiac64]

The advent of social media and the build up of romantic notions of hiking in recent movies and books has likely swelled the ranks of the ill prepared.

[soilman]

I would guess that not all NOBO thru hikers sign in at the rangers station at KSC. I can see if one is not camping at KSC they would have no reason to go to the rangers station. I started an unsuccessful SOBO in 1976. I did not sign in at the rangers station. I walked 1300 miles before quitting. I stopped in at Harpers Ferry and never signed in there. In 1977 I started NOBO from Springer Mt. Never signed in there. In 2010 I signed in at Amicalola Falls (#550), but not at Springer Mt. Signed in at Harpers Ferry (#344) and KSC (#242).

[perdidochas]

Well the first 25% cant get out of Ga.

This alone is mind boggling. How anyone can prepare for a year, spend hundreds to thousands on gear, and not be able to make 80 miles.

Theres lots of dreamers. People in poor shape, flawed expectations, or not really committed to something thats actually ... hard work.

I suspect as much as 1/2 may fall in this group, drop out before Va.

Has social mefia attracted more of these? Surely yes.

But it also helps them succeed too

I often wonder about that. Why would you go to all the trouble of starting a thru-hike, without doing prep work before hand--things like practice backpacking trips or simply getting in reasonable shape.

[Koozy]

The numbers as published by the ATC are likely the best you'll ever get for this estimation. There are so many factors that go into under or overcounting hikers at any given section of the trail.
IMHO, I think the % completed is bias high; many that start and quit early are never accounted for, and I personally knew about 10 hikers that skipped up to 500 miles of the trail in 2014 yet still claimed completions.

[Emerson Bigills]

The real numbers are only as accurate as the data collection. The internet and social media have opened up thru hiking the AT to people who are not hikers and may not even be outdoor types at all. I hear it from people several times a week. This puts a lot of inexperienced people on the trail and often they are unprepared for what lies ahead. Having said that, the internet also provides prospective hikers with valuable information that would never have been available before to hikers. The improvements in lighter weight equipment also aid hikers in moving faster and suffering less.

To conclude, I think there are enough puts and takes on each side that a completion rate that has not moved too far either way in the last couple decades would not surprise me.

[evyck da fleet]

I see two opposing forces at work. As time goes on and more services and info is available the hike becomes easier to undertake.

On the other hand, most younger people who ask me about thru hiking usually do so having referenced social media stories of trail magic, parties etc. I see a group interested in an adventure but not necessarily a thru but who will register as such to be part of the group and enjoy the benefits of claiming to be an aspiring thru

[KnightErrant]

I was thru-hiker 2173 to get a tag at Amicalola on April 12th. 1262 at HF on July 1st. And right around 1300 at Katahdin Stream campground (didn't make note of exact number) on October 4th which led me to believe that in Baxter, some people had claimed to be thru-hikers that weren't, because otherwise literally no one dropped out between HF and BSP. I'm sure I also got passed by a lot of people in the second half of my hike, because I had fairly serious health issues in the last two months that slowed me down.

But even using Amicalola's number of 3862 Nobo starters and the Kennebec Ferry's number of 1213 Nobo thru-hikers, that's a 31.4% completion rate of 2018 Nobos. Even rounding up to 4000 starters and down to 1200 for finishers, which assumes 138 Nobos didn't get their tags and 13 Nobos who crossed the Kennebec and didn't get to Katahdin, you have a Nobo completion rate of exactly 30%.

There's more guesswork when you include SOBO (who might not be finished yet) and flip floppers, but I would assume that the Sobo completion rate is at least as high as Nobo, since you have to already be a more serious hiker to even attempt Katahdin and the 100-mile wilderness to start out.

I guess I just don't see the evidence that the completion rate is actually going down? I think with the Hiker Yearbook becoming so popular, a lot of finishers register with them and plan on getting that as their keepsake, rather than applying for the certificate with ATC. Meanwhile, having the ATC hangtag is almost a status thing on the trail, so I think most people take the time to register at Amicalola even if they then skip the approach trail. That leads me to believe that the ATC statistics underestimate completion rather than overestimate it. I think Amicalola registration and Kennebec numbers are probably the most accurate counts we can get, for Nobos at least.

[peakbagger]

I wonder how many folks sign in as thruhikers at BSP so they can stay at the Birches?.

[George]

I wonder how many folks sign in as thruhikers at BSP so they can stay at the Birches?.
Plenty, but unlikely more than a 100 a year