How close can I get to three hours for a FM?

Compare yourself to people who don't even run or are obese. Snowflake You're a winner in your own eyes?.

And yes, yes I would if you were blowing off in person like you did in your post after the race. An ego like yours needs to be put in it's place.

Tw@ wrote:

Compare yourself to people who don't even run or are obese.

I compared myself to both elite runners (point one) as well as the population as a whole (point three) to place my performance in proper context. Your evaluation of my response was incomplete and disingenuous. I'm sorry if the nuance of the message was lost on you.

Snowflake You're a winner in your own eyes?.

See point four above.

And yes, yes I would if you were blowing off in person like you did in your post after the race. An ego like yours needs to be put in it's place.

When you're alone, and honest with yourself, and not trying to score points on a game that is nobody else is playing, all on some anonymous internet forum, I'm fairly certain you know you would not have the conviction to call another human being that you don't know a pompous twat. I cannot prove this, but I suspect it to be true.

Regarding my post-race post, I think it was clear to most people who engaged their critical faculties that I was implementing a bit of comedic hyperbole in my overuse of the glorious terminology. With this said, you still seem clearly troubled to me. My offer remains open. Even though you wish to tear me down, I would like to pick you up if you're suffering.

Strava provides plenty of significant digits. That's the point! Significant digits are based on the accuracy of your measuring device. In the case of a GPS watch...not very accurate! So anything past one decimal places is literally and truly garbage. The thing with measurements is more digits isn't always better if you have no way to confirm if those digits mean anything of value. More numbers in don't mean less garbage in. It means more garbage in. Also, a large dataset rounded to reasonably few significant figures is a far more accurate representation of the real data than a couple people you talked to at the race anecdotally rounded to more decimal places. Your number looks more precise, but it is certainly less accurate.

The reason Strava doesn't round data up when syncing is because someone who runs 26.35 miles on their GPS didn't run 26.4. You don't get credit for something you didn't do. Pretty simple rules of racing. (I'll even say you don't get credit for that sub 20 5k you've never run).

I would agree that you can run faster or slower on different courses. I will not agree that the reason you were slower on this course than another course is because your GPS measured it long or you had to turn too much. That is my point. Go run the jack and jill marathon and claim your 10 minute PR since you ran down hill if you want. Don't say a course was long and use GPS as the supporting evidence. You're trying to soften your claims now that "in the right conditions I could have done xyz is all I'm saying. Here's what you wrote:

I went to marathonguide and looked at reviews. Started at the most recent reviews and searched for "long." Only came up with "no long lines for port-a-potties!" and "there were more supporters along the route than some larger races I've seen." The first mention of lots of turns leading to a GPS measurement of the course being too long is from the 2013 running of the Rocket City marathon. Lets read the first sentence of that review...

"Course (not sure it really matters since I understand its changing next year):"

What's that?? No one has complained about the course being measured long by GPS since they changed it in 2014??? well oh my... (that review also states that it's a solid BQ race and relatively flat...not exactly a "slow" or "added difficulty" course).

Findmymarathon.com ranks it the fastest course in Alabama, with a PR score of 98.58. (Andrew Jackson ranks 97.63, Boston 99.62, for reference) There's no certified course out there that would have you running a sub 2:57 on Saturday that isn't huge net downhill. Maybe try Revel Canyon City next year.

Finally, only 5 people on Strava uploaded their results for the apparently exceedingly easy, straight, and reliable Andrew Jackson Marathon 2017. Three 26.4s, a 26.5, and a 26.1. For an average of 26.36 or put it in terms of sig figs.....26.4. lol.

You ignored the vast majority of my points, and now you're conflating significant digits with a significant amount of data. I'm not sure how else to explain it, but I'll try. If you were to consider hypothetical course A that was properly measured at 26.22, course B that was improperly measured and is 26.27, and course C that was properly measured at 26.22 but contains far more tangents, you might obtain GPS averages for a number of runners of something like 26.40 for A, 26.45 for B, and 26.47 for C. Strava is going to keep rounding down, and using your methodology, arrive at no difference. On top of that, when you don't have enough significant digits to measure small difference, you're not going to detect small differences, hence the practice of dropping decimal points beyond the number of significant figures that go into the system.

Agreed, I would not want them to do that either, and in no way is my position reliant upon them doing so. So, that is a logically fallacious straw man if I have ever seen one. I, and my others, wish that they would include the hundredths position and round down. This would a better approximation of how far a runners has gone. That digit would be far less accurate than the tenths digit, but it would certainly be far more accurate than just assigning every runner a zero, which is exactly what they do. And then, you're trying to obtain meaningful information to compare two races from that data. It can't work that way.

I'm not asking for credit, just as I'm not asking for credit in running a sub 2:57 on a properly measured, less tangential course. I'm suggesting that I can be reasonable to assume that I can do both given the data that I have, and that had RCM been more like the races I've done before (fewer tangents), all other things being equal, 2:57 was a solid possibility.

I think you're sorely mistaken by disagreeing that a larger number of turns yields a slower time. Turns slow runners down, just like hills and heat. I thought this was Running 101. This course has thirty-seven turns on the first half alone. That makes a difference, whether you acknowledge it or not. Claiming a PR on a downhill is disingenuous, and you're merely trying to use a logically fallacious analogy by attempting to draw a comparison between this and an attempt to find a more fair, level course operated under optimal conditions. I can very well say that a course was long as and use the GPS as supporting evidence. GPS has been used as evidence for mismeasured courses all the time. It is not the final authority on all matters on which it speaks, but it was recently used to discover mistakes in three different local races over the past three years. People make mistakes in course mapping, even when those courses are Boston certified.

A lot of what I saying when I was claiming to have been gloriously "robbed" was intended as hyperbole. Use that if you wish.

And yet, I can go on findmymarathon.com and find reviews with complaints about the course being long. Your lack of research is not evidence for your claim; your claim of evidence is from a lack of research. This is the logical fallacy of an argument from silence.

Just because RCM is the fastest in Alabama does not make it a more ideal course. There are dozens of faster, non-downhill courses with a higher ranking on that site. AJM has a new course that I was unaware of.

Except that the very next week, there is a non-donwhill one nearby in Jacksonville, FL that the projector suggests would yield a 2:56:49 based on my 2:58:18 in Huntsville. Surprise, surprise. So, you see, your utterly absurd suggestion to run a mile-downhill race at Revel Canyon in order to support my claim is yet again, another example of how you use fallacious logic to disguise your inability to evaluate data properly. I thought that my performance would have been enough to convince you that evaluating data isn't your specialty, but I'm beginning to see the real problem. You're greatly affected by confirmation bias, more so than the average person, in which you ignore arguments and evidence (and in two cases thus far, entire posts) with which you do not agree. Then you shift right on to some other topic. I, on the other hand, will take the time to respond point-by-point to what any person is suggesting.

And so, yet again, let me point out the flawed methodology. One, AJM changed their course this year. Again, I was unaware and had been using their previous course when making my case. If 2016 Strava data were used, you would see a different story, as I came up with 26.31 as an average. Do I say, "SIGNIFICANT FIGURES, 26.3! I'm right!" No. Because... Two, you don't have enough data to draw a meaningful conclusion with those numbers. Three, you're using the same GIGO methodology as before. And yet, I doubt you've learned anything. LOL.

I honestly went through and thought I answered every one of your main points, just as a list instead of interjected with each quote along the way, and then you say I didn't address any of your points. So let me condense your points again and you can tell me which one I missed. I'll use your quotes so it's all clear. Just requote yourself for the ones I miss.

You can't use your GPS data as proof the course was too long on a certified course, unless this year they accidentally put the start or finish or turnaround markers at the wrong location.

The course measurers and certifiers do not agree.

This is not proof of anything in the face of physical measurement and ceritifciation. Also, not "everyone" because there are many more favorable reviews online than the "multiple" complaints of it being long. There are a few turds in every punch bowl.

You have no proof of this. This is a gut instinct on your part, and gives yourself credit for something you've never done.

To which I replied Hey, since you want to use Strava as supporting evidence, please look at the further study I did on Strava. Instead of taking the range, lets look at the average because when you use the range the one guy that reported 26.2 is weighted just as much as the 30 that said 26.4. That's a better way to look at data.

But...I thought you said to check Strava?

I totally agree!! But wait...weren't you saying some pretty definitive things about the GPS data before?

Yeah, that was it.

But you're willing to make pretty strong claims off of a handful of people you talked to anecdotally instead. I chose 40 people at random and reported the results. You altered the search terms to match or mismatch the bias you were looking for. Which is the preferred method if you were doing a lab study? Obviously random sampling and polls are never perfect, but they're the best we can do when we don't want to spend all day proving a whiny guy on the internet wrong. Especially when you're primarily trying prove that GPS ain't infallible.

I also totally agree with this!!! As long as you've 100% backed off from the claim that the course was not long and just a GPS and some online reviews are not proof of anything. I can find positive online reviews for the movie "Failure to Launch", that doesn't mean it was a good movie.

Too bad Nike didn't invite you to Monza and you had to run a race in the real world like nearly every other person ever, I guess.

No I wasn't. I was saying: 40 data points chosen at random with significant digits is a more credible data source than the few people you talked to even though their number had more digits. And for you to say your source is more reliable just because it has more digits using a faulty measuring system is false. I could have polled more than 40 random data points, but I've already wasted too much time on this topic and the average wasn't changing appreciably as I added more points.

This is the whole point of significant digits in measurement! The difference between Course A, B, and C is 0.3%. The use of GPS easily has a greater margin of error in measurement than 0.3%. If you want to try to measure the difference between these courses with GPS the best you can do is state "There is no significant difference in course length, within the reasonable error of our measurement device." Because look! Strava reports they're all 26.4mi! That's the best you can say! You can't detect that small difference using Strava, but you also can't detect that small of a difference using any GPS device anyway. You get a number, and it makes you feel better because it has more digits, but that number has no basis in physical reality. But wait...all this talk about how useless Strava data is...

Oh, that's right.

This is true, but I think you are overinflating their affect on a guy running 6:49s if it really is running more turns slows you down. But you spent a lot of time talking about how the course is definitely long so I got a little saucy about all this complaingin about turns. (I only bring up the 6:49s because it's harder to make a turn at 5:20 than at 6:49, so it probably slows down faster runners more).

So you're still claiming the course is long because of a few people agree with you and GPS data? I thought it was just turns slow you down like hills now?

Were they certified races or local rinky dink affairs? Was it due to the course being designed and measured wrong as you're accusing, or a start/finish/turnaround marker being placed in the wrong place on that day? I would agree that GPS can be more accurate than a non certified course or poorly managed race markers. That's far less likely to happen year after year on a certified course with over 1,000 participants.

It's also usually when the GPS demonstrates errors greater than 1% compared to advertised length, which is the case here.

But you repeated it again after the first comment...

Naw. It was intended as hyperbole. Use that if you wish.

If you didn't know this, how did you expect me to?

I agree I don't have enough data to draw a meaningful conclusion, and neither did you when you said "the course is long, everyone says it's long, it's what kept me from sub 2:57."

Let me give you an analogy. I'm here, with a measuring tape trying to measure a squirming baby. He won't lay flat, he won't sit still, he's an awkward shape, etc. I have 40 people measure that baby and then I can say "This baby's about 26 inches. Can't say for sure if it's a little more or less than that. If someone else measured the baby with a laser while it was asleep their number is probably right." Certainly more accurate than I can get with this tape measurer. You measure the baby once while it squirms around every which way and you say..."No...this baby is 26.35 inches. Because the last time I measured a squirming baby I got the same number as the sleeping laser method, offset by .17 inches. Also a couple other guys I know measured this baby and said it was bigger than 26, too. That means the sleeping laser system is wrong this time."

TL; DR

David, you've won long ago in the eyes of the people that matter, but you'll never convince this narcissist he didn't run a sub 2:57.

Go enjoy you day; life's too short for arguing with people who have personality disorders.

I will disagree. I do not think you can honestly claim that you went through and answered my points, or you are playing some sort of game where you do not consider it a "main point." I specifically referred to your disingenuous correction of my 0.52 final split to 0.2 when referencing your behavior of avoiding a point-by-point response, and you once again failed to address it. I specifically referred to your disingenuous attempt to ignore online reviews for a site in my most recent post, and you did not address it. I specifically referred to logical fallacies that you committed in your argumentation of your most recent post. None of that was addressed. I pointed out that a hundredths digit on Strava would be more useful than trying to draw a meaningful conclusion on the data with everything rounded down to zero. Ignored. With you, it seems to be a consistent case of clear cut-and-paste with what you think are your strongest points. I could go back to all the old posts in this thread pre-race where you failed to respond when I challenged your methodology on predicting my times and probably pull out 20-30 instances in which you didn't respond to the point I was making. I'm not going to do the work for you. I consider the onus to be on *you* to go back and respond point-by-point to each line I offered on interpreting the data for a race projection. It should not be my responsibility to cut and paste. If you don't want to do it, fine. It will take a lot of time. And you might just consider it water under the bridge at this point. I will just continue to think that you don't respond point-by-point because you can't.

Proof? Who is talking about proof? I'm using it as *evidence* to support a hypothesis. There's an enormous difference there. If you are referring to my "definitely long" statement, I wholeheartedly recant. I never intended on that to be taken in a literal, scientific sense.

Fair enough. I overstepped with my claim by saying "everyone," which I did not intend to be taken literally. All the runners I spoke with in person believe that there is something strange going on with the distance.

There you go with "proof" again, misrepresenting my position with another straw man. And of course you once again commit the logically fallacious argument from silence by suggesting that the absence of a statement about the length of the course is somehow evidence that these people do not hold the belief. Some people don't know; some people don't care. It is not incumbent upon the reviewer to list every thought that they have. If you don't see the fallacy here, study logic.

And now you're bordering on ad hominem.

Straw man after straw man after straw man. When did I say I have proof that I can go sub 2:57? I have said time and time again that I can be *fairly confident* that I could go sub 2:57 without the issue, and you continue to mischaracterize my claim as that I have proof.

As I have consistently argued, the Strava data tend to consistently show that the distances run at RCM are longer than other races in the area. I draw no conclusions and claim no proof of anything.

Jesus.

Pretty definitive things. Jesus.

I have already addressed the point of the statement was an exercise of hyperbole.

And I have addressed the reason. GPS watch data includes hundredths. Strava does not. My data is anecdotal, yours is fundamentally flawed in its method of recording. I can't make you understand this.

Unbelievable. You wrongly criticize me for making me definitive statements that were not intended to be proofs and definitive statements BY MAKING A DEFINITIVE STATEMENT that you cannot possibly prove, and is furthermore wrong. You answer my claim of cognitive bias by making a claim of cognitive bias that CLEARLY SHOWS your own cognitive bias, namely that you merely want to be correct because of your bias, and then suggest that I have done something that I can tell you with all sincerity that I have not done. I even went so far in my objectivity as to say that when I re-ran the search using a different method, I got the same results you did.

Beneath you David.

I have already said repeatedly that GPS and reviews are not "proof" of anything. I have even suggested that the longer average distance could be a result of the extraordinary number of tangents required to be run. I do not back off my claim that the course could be long, but I should not have definitively said that the course is long. FWIW, there was traffic along some of the roads as well as parked cars, making it impossible to run tangents. For whatever reason, the GPS provided a longer result than other races in the area.

Too bad you can't respond point-by-point when I address your statements instead of making these absurd quips. Plenty of people in the real world run on more favorable courses. I could have done better on those. I thought this course was one of those, hence my 2:57 best-case prediction. And just as I demonstrated with the Jacksonville course, I feel that I am likely correct.

Yes, you were.

I have already explained to you the fallacy of relying on less precise information when attempting to draw a conclusion that requires precision.

I'm literally banging my head against the wall. Do you have any grasp of what a confidence interval t-test would look like using the data as more data is applied to it? You keep parroting this Strava quote I made about 26.2-26.4 to 26.4-26.6, but you fail to understand that if the difference is significant enough, it will show with even less significant digits put into the equation. It's much more reliable to reject the null than it is to confirm it when you handicap yourself with not having the hundredths. I ran a test and got a 0.1 difference when accounting for significant digits. I make no definitive claims about that, but is certainly supports my view.

Fair enough.

Sigh...

Agreed, but yes, they were certified courses. And IIRC, two were due to misplaced turnaround and one had a conflict with the measuring, and they were not small local rinky dink affairs.

Which says what exactly?

Then I apologize. It's just really hard to tell online. I'm glad that you've backed off your insinuation that running a mile-downhill course is somehow analogous to looking for a more favorable flat one .

Did I say that I expected you to know this? No, I didn't. I pointed it out as a reason for the discrepancy. Where do you get the idea that I expected you to know this? It was nothing I said. You seem to be seeing what you want to see. I was focused on pointing out the problem with your methodology. I did not criticize you for not knowing it. I pointed out that it was a problem.

Fair enough.

Please allow me to restate my position in a more serious manner.

Prior to the race, I suggested that 2:57 was a best case scenario. I finished in 2:58:18. When I made this prediction, I was unaware that there were thirty-seven turns in the first half alone. My GPS ended up reading 26.52 when most of my HM/FM GPS data would suggest 26.35 would be an expected result. There are a number of possibilities that made this course less optimal than I had planned: the inordinate amount of turns make its more difficult to run the tangents, the course could have been mismarked, the course could have been mismeasured, the course could have been long. I can say nothing definitive, but the course has a reputation for making the GPS data look longer than other races.

Now, I'm tired, and I have real world stuff to address. Thanks again.

This is honestly the most pathetic dick waving contest I've ever witnessed.

bkrunner wrote:

This is honestly the most pathetic dick waving contest I've ever witnessed.

Then you never saw my work over at "Disappointed at NYC, should I go back to back?"

I was gonna congratulate you, but you ruined it. You must be either a troll, or incredible stupid (course long? haha). I'm not even convinced you ran sub 3, and more importantly, I don't care. Also,

three hour attempt wrote:

I can be fairly certain of a few things:

3. The fact that I have completed a marathon in under three hours places me in the top 0.02% of the population.

it's the top 2% of marathon runners only, and in case you're a prime age male not even that. That's it from my side.

above_average_joe wrote:

I was gonna congratulate you, but you ruined it.

This.

While I won't pretend this thread hasn't become massively entertaining in a car-crash-TV kind of way, it's a bit disappointing - I checked in this week after your marathon hoping to see if you'd done it - maybe it's sometimes better if these kinds of threads end unresolved, haha.

Davids S. Pumpkins wrote:

bkrunner wrote:

This is honestly the most pathetic dick waving contest I've ever witnessed.

Then you never saw my work over at "Disappointed at NYC, should I go back to back?"

Oh memories...this brings 2016 back. What a thread it was! Was following it with deep interest as I had similar goals as outsiderunner. I even posted to help, and god I regretted it. Guy was not easy to deal with, and could just not take any advice!