What is an "FKT"?

As a mid-Michigan runner who has seen a huge uptick in people claiming FKTs, I've recently been thinking about what does/doesn't count as an FKT ("fastest known time").  This makes perfect sense in light of all the canceled races due to COVID-19 as well as the rise in popularity of easy verification methods such as Strava.  But the uptick in claims of FKTs raises questions about how far that concept can be stretched.  That's what I'm interested in exploring here.

As I see it, historically FKTs have two sources: 1) long haul hiking and 2) the mountain culture of Colorado, the term "FKT" itself coming from the latter (Buzz Burrell, Peter Bakwin).  I can still remember climbing Longs Peak in the early 1990s and seeing the fastest known time of the ascent, descent, and round trip of Longs at the ranger station by the trailhead.  After having tried (and failed b/c of weather) to summit Longs, I was blown away by how fast someone had done this.  Stats were informally kept all over various peaks/routes in Colorado for years before there was anything like the popular craze that exists today.  Times were kept on an honor system (since there was obviously no GPS tracking websites around then) and no one wanted to cheat since it was antithetical to the whole ethos that existed between the pioneers of the sport.  There's nothing wrong with a fringe thing becoming a popular thing.  It's just that that process sometimes results in losing track of the ethos and function of the original concept.  

Here are some clear cases of FKTs: 

• the double crossing of the Grand Canyon
• 14ers in Colorado (including linkups like Nolan's 14)
• the Manitou incline
• The Appalachian Trail.

Here are some clear cases of things that should not count as FKTs: 

• a flat, paved stretch of rail trail, 
• a track,
• a loop route that already has an FKT but someone new runs from a different starting/ending point than the existing FKT.

So, what makes the difference between these clear cases of FKTs and clear cases of things that aren't?  I suggest the following (partial) normative definition:

A route counts as an FKT only if  we cannot easily convert existing road or track times to that route.

This necessary condition on FKTs makes sense of Burrell's norm of not counting road routes as FKTs.  After all, if you're just running 6-minute mile pace on some flat stretch of road in Anywhere, USA, it is quite easy to imagine how other runners will stack up against you.  In contrast, it is really hard to convert the best mile time on a track to the mile long climb up the Manitou Incline.  It is often the case that some of the best road runners don't stack up against the best mountain runners.  That's the point, I suggest, of an FKT: to provide a metric against which we can see how we stack up against others. The reason we don't need an FKT on a rail trail or a track (much less changing the starting/ending points from an existing FKT!) is that we already know how we stack up against other athletes on such routes.  The function of an FKT is epistemic.

It seems to me that Strava KOMs are meaningful in the same way as FKTs are.  If you want to create a segment on a flat stretch of road or track, then go for it!  I've created numerous such segments since they can really help track things for training purposes and it's also just fun to compete with others.  But not all Strava segments are equally helpful and certainly if you're creating new segments on known routes that have had existing segments for a long time just so that you can get a KOM, then you're being a #stravawanker.  And if that #stravawanker then claims an FKT then that's a another level of wankership.

Almost no definition escapes counterexamples.  Here are some harder cases to consider:

1. Routes that involve more technical climbing (Andy Anderson's Longs Peak FKT) or other disciplines (e.g., Joe Grant's self-supported tour de 14ers)
2. The transcontinental record
3. Multiple loops/laps of an existing route

I think case #2 is the easiest to accommodate.  Although shorter routes on flattish road generally compare easily to known road records, multi-day road routes do not.  You cannot very well take an elite marathoner's PR and figure out how they'd stack up against Pete Kostelnick's transcontinental record.

Although Burrell's FKT site suggests otherwise, I see no reason to rule out cases like those in #1 as FKTs. After all, there's a long tradition of ascents/descents of peaks both in the U.S. and in Europe that dovetail well with the FKT movement.  Many of those routes involve more technical climbing, such as Kilian's Mattherhon FKT.  If the concept of an FKT should match up to existing and historical usage, we shouldn't discount multi-discipline routes.  The same, I think, applies to Joe Grant's tour de 14ers, as there is a long-standing tradition in Colorado of keeping informal time records for completing all 50-some 14ers.

I'm on the fence about multiple loops/laps of the same route.

-------------------------------------------

Pedantic philosophical postscript

Nelson Goodman's account of how we justify the rules of logic is useful in thinking about these matters.  A definition of validity for Goodman doesn't simply normatively stipulate how that word is to be used.  Rather, like a lexicographer, the definition is supposed to reflect our descriptive usage of the term in question.  The lexicographer considers usage and then stipulates a normative definition based on those commonly accepted usages of the word.  That's how I'm approaching this normative definition of an FKT, as well.  Whatever "FKT" means, it should reflect the ways in which the relevant communities use the word.  

Sally Haslanger's notion of an "ameliorative definition" is also relevant.  Part of the point of giving a definition is to bring us a better understanding of the concept in question.  That's also what I've tried to do in the foregoing.