Breaking the Wall

Hey, I have to disagree with your top 12 runners. No offence to Steve, but Logan should be in front of him. Although you have all of your technical calculations, Logan qualified for trials and that is better than a 2:29 whatever at Ogden. I think Logan needs a little more credit. Just my opinion.

James - I have wondered if I overrated the effects of the heat in Ogden this year. I gave everybody 2:00 for the heat. That may be too much, especially for Steve who is a very good heat runner, but in making the adjustment I cannot give one runner more than another just because one is a heat runner and the other is not. I also wonder if I might be a bit too generous in estimating the slow-down in Ogden on the rolling hills by Eden. To say anything conclusive, though, I need more data. Until then, I'll go with what the predictor is giving.

The fact that somebody qualified for the Trials is mostly of an emotional nature. It does have important significance in the way of publicity and emotional feedback, but the fact that one person qualified and another did not in and of itself does not signify that the one who did necessarily achieved a higher quality mark. For somebody who cannot qualify hands-down on any course in any conditions (I think you could easily find a course and the conditions that will make it impossible even for Ryan Hall to qualify), the matter of qualifying is not only in getting into your top shape, but also in getting into your top shape at the right time, and running the right course at that time. This is an important skill, but I am not giving points for it on the top list. I am only comparing the qualify of performances achieved at any time.

My opinion is the Top Runners is too technical. I think it should be by time only, regardless of the course and conditions. Keep it simple. If you are going by Sasha Science and using the "Flat Sea-Level Ideal Conditions Equvalent" it's not in order anyway.

Personally, I feel, rather, KNOW my performance at St. George was better than my Ogden performance.

*Much better training: better base, better workouts, more volume before St. George

*All of my benchmark workouts preceding St. George were about 10 seconds per mile faster than those preceding Ogden

*I was working with and chasing people at St. George vs. running by myself at Ogden. Therefore, I pushed myself harder and closer to my limits (my sore muscles will attest to that)

I don't think the heat affected me much at Ogden, as my splits didn't really slow down much. I also think Ogden is a faster course than you give it credit for, and should rate about the same as TOU. Therefore, I think my Ogden time was worth about a 2:23-2:24 at flat sea level, and my St. George time was worth about a 2:21-2:22 at flat sea level.

Just my thoughts.

It would be cool to have a "sort" feature on the performance list, similar to the mileage board. Then be could sort on "Sasha" times vs. "clock" times. Personally, "clock" times mean more to me, probably because I am a track guy.

Going on what Mik'l just posted, and my post on the message board, I would still love the see a performance list governed completely by an underlying database. I know you like to calibrate times, but I think most people think a time is a time, and would like something more objective. I for one, don't really care if a time slips in there because of a short course or aided course or user error. In the long run (months and years from now), a database-driven list should also be far less work for YOU (since it is automated), and should be more inclusive (ie - be longer and let more people participate in it, rather than the top tier of bloggers). If the blog expands to 100x its current size, you will not want to manually enter and do Sasha Science on the hundreds of runners that may be competing every weekend. Or maybe you do, I don't know.

I think many people would be motivated and inspired to improve by seeing their names in objective performance lists, even if they are a ways down the list, or have to play a few sorting or query tricks to get themselves to appear.

Again, just a few thoughts. Regardless, we all appreciate the thoughtfulness and time you put into the existing performance list and other blog features.

Mik'l:

Thanks for spotting the error, I have now fixed it.

Paul:

Some circumstances to adjust the quality of your St. George performance vs Ogden. You were sick before St. George, that took away some from your fitness, maybe a minute. You closed way too fast in St. George, which means you were probably a minute too slow through the first half. In Ogden you ran a slightly positive split, which would be a slight negative adjusting for the heat. In the last miles of Ogden you had a PR in sight, but not secure. You also did not know how far behind the chasers were and what condition they were in, and letting someone slip by was worth $550, so there was a bit of running scared as well. In the last miles of St. George you had standard A in the bag, and you knew who was behind you, what kind of condition he was in, plus it was not worth serious money anyway.

I agree with Mik'L and Paul that it should just be changed to straight time rather than adjusted. Especially giving 2 minutes for the heat seems a bit much.

And if Paul says his St. George was a better performance than Ogden, than it was a better performance. I know he had much, much more/better training (no real long runs before Ogden), plus read his description of the last few miles- he wasn't letting up.

If you are comparing performances than time would win over adjusted for sure. If Paul would have run St. George the same day as Ogden he probably wouldn't have quite hit the B standard, and if Logan would have run Ogden last Saturday he would have been around Paul's record if not faster. This is totally my opinion.

I know that Paul gained some very good fitness in that five months, and that his Ogden performance, as good as it was, doesn't come close to his race last week. I think Ogden is faster than we give it credit too. TOU is mostly faster because it is in the fall and people can just plain get in better shape over the summer than over the winter around here. I also think, whether you agree or not, last week was your best marathon performance.

If you ask Hobie Call he will tell you that TOU is faster than St.George for him. Time is time, and some people run better under certain conditions than others do. Although I do like to see the comparisons to norms and standards at times. When I ran my fastest 1500, it was at elevation, converted to sea level it is about 4 seconds or so faster. But I didn't run that fast and who is to say if I would have been at sea level for that race that I would have even come close to that time. If I converted all of my PRs they would be totally different because none of them were at sea level. Some were on more difficult courses and others were on faster courses. A PR is a PR unless it is Altla Peruvian, and it is still a PR in a way. Once again this is all the opinion of James Barnes.

I do not put a lot of weight on "my workouts have been faster" when the workouts are not done very hard. Extreme example is that some days I may be running 8:00 in the warm-up and it feels hard, but then find myself capable of 5:20 in the tempo. Other days, I may be running 6:40 in the warm-up and it feels easy, but then I am struggling to run 5:40 in the tempo. The problem is that the perception of hard/easy in the warm-up is influenced by the agitation of the nervous system more than by the actual fitness level. At sub-maximal intensities, e.g 5:40 pace on a flat surface for Paul, the perception of effort may still not correlate with actual fitness because he has the ability to run the same distance at 5:10 pace.

I will put some weight on the workouts when there is a consistent history of correlation between a particular workout and race performance. I call them the workouts where you cannot cheat. It is different for different people. For me, interestingly enough, it is a 3 mile tempo close to all out except the first mile has to be done at 10 K effort. This will fairly accurately predict my marathon. However, the same workout does not do it for others. E.g. Steve Ashbaker has been able to do that workout with me on numerous occasions feeling about the same way I did, but his marathons have fluctuated big time even though his 3 mile tempo time was fairly steady.

One other argument for changing to straight times- the top runners board right now makes it seem like Paul is slower than Sean based on times alone, since his Ogden time is 8 minutes slower than Sean's St. George, not reflecting that Paul also had a faster St. George time. I imagine Paul would rather have his fastest time on the board, even if Sasha Science says another race had a faster equivalent time. I imagine you (Sasha) would prefer your 2:23 marathon showing rather than a 2:32. I'm with James- just go off raw time. I don't try to look at my 8k cross country races and normalize the times- I just remember my fastest time and quote that as the PR.

And to think that St. George is 9 minutes slower than Ogden this year seems extreme- that is equivalent of over 1.5 miles!

St. George is about 6:40 faster than Ogden when both have good conditions. Ogden was hot this year, the times were slower. I really felt it after 10 miles.

Interesting thing I just noticed- if you rank only by personal best (did you just add that, Sasha???), then there are only 2 changes in order- Steve moves from 4 to 8, and Ted moves from 9 to 11. All the other runners stay in the same order.

Am I crazy, or did you just add that personal best, Sasha? I think the top runners page is a lot better with that column now- that way people can see what their PR was. Thanks for adding that.

Jon- you are crazy, that has always been on there.

I find this funny.

Sasha - On that page you spelled Equivalent wrong, twice. There, I added something constructive to this argument.

Everywon nose runers cant' spel.

It's official, I am crazy. Don't tell my wife.

Sasha,

What does the "V" stand for in VPB? I'm assuming PB stand for potty break, pee break.. or something close to that.

Kind of depressing to "know" I haven't progressed at all since May, and that my evaluation of my own workouts aren't valid. Thanks for setting me straight. ;-)

Oh well, you're a skeptic, but I believe in myself!

Cody - thanks for reporting the error. I have now fixed it. For those who cannot, spell, download and install Firefox with Google tool bar from the link on the right side of this page, your blog posts will be automatically spell-checked, if you spell something wrong it is highlighted in red, and as a bonus, the Fast Running Blog will receive $1 for every new installation on Windows. Yes, Google really wants to make some inroads on Microsoft in the browser market, and is willing to pay the webmasters who are helping it. I am happy to do it even if I am not paid, though.

Ryan - VPB is a spoof off VPN (virtual private network) and stands for Virtual Private Bathroom, or in other words, a bush.

Jon, James - here is how I calibrated the predictor between Ogden and St. George. I took 2005, there were three reasonably reliable runners that ran both that year - Joe, Paul, and I. Joe ran 2:29:01/2:22:23 (diff. 6:38), Paul did 2:35:24/2:26:35 (diff. 8:51), myself 2:36:04/2:27:21 (diff. 8:43). Both St. George and Ogden were run under ideal conditions. After analyzing how both races went, as well as others leading up to it, I figured Joe was the best runner to calibrate the predictor by, while Paul and I had a bit of an off race in Ogden for various reasons.

Regarding the argument that people run slower in Ogden because they have not been training as well during winter. I do not use those people for calibrating the predictor. Pretty much you have to make top 5 in TOU, top 10 in St. George, or top 3 in Ogden making the standard in at least two of those marathons in the same year to be a calibration candidate. Reason - faster runners train properly through the winter, are much less likely to have an off race, and have already gathered the low hanging fruit off the tree of their potential, so it is extremely unlikely for several to make the exact same amount of breakthrough in the same race.

Paul:

You'll have a chance to settle that issue in the Trials and in Ogden next year. I am also going to run St. Jude marathon, and hopefully this will give us some more meaningful data to munch on. As I already mentioned, there has been only one runner that has met the Sasha Science standard to establish a correlation between Ogden and St. George. We need more data. It is possible that I may adjust the Ogden course a bit in the future.

Holy Cow! I only have time on my lunch to just look at this.

Are there math programs involved in Sasha Science, or is it just your gut feel?

Even though a good runner may train through the winter and run a good race at Ogden in May. That same runner is going to race and train hard all summer and be faster in October. Paul would have ran Ogden on Saturday in way faster than his time in May with or without heat.

The race predictor only had me running a 2:30 at STG off of my Hobble Creek time of 1:09. But in those 5 weeks or so I started doing better tempo runs and feel I got better in 5 week. What could someone do in 5 months?

I'm enjoying this thread. It's fun to see how everyone views different races at different times.

Sasha science acutally only had me running a 2:31:26 based on my 1:09:40 at Hobble Creek.

Superfly - I am not at peak fitness levels yet or anywhere close so maybe I can shed a little light to "What could someone do in 5 months?" I ran the DMN 10K and Provo River Half marathon earlier this year. Based on my DMN time the predictor was within 1 min of my actual finish time in the Provo Half. However, the predictor was almost a half hour off on my actual SGM finish time. So, by training better, eating healthier, running consistently, etc etc I was able to shatter the predictor's prediction. Until a person reaches a real high level of fitness, past performance won't always be a good predictor of future results. For someone like myself - overweight and still building a mileage base - 5 months time can lead to drastic improvements. I know the question was somewhat rhetorical but there you go!

Hey if we don't have enough data to covert times accurately then we just need to go with time. My vote is with fastest time, and make another category for conversion.

Clyde - Hobble Creek started late and was hot this year, so the times were about a minute off to begin with. On top of that your Hobblecreek was a bummer race, as evidenced by your position relative to Mike Vick and Matt Harmer in St. George. You went into Hobblecreek not fully recovered from an illness and an injury. Also, Hobblecreek course was slightly changed, possibly to the slower side, since the tuning of the predictor.

There is a difference between face value output of the predictor and Sasha Science. The predictor does not know about the conditions, nor does it have any idea if this was an off race or not. Sasha Science actually told you to go out with the standard B group just in case you were having a particularly good day and make a qualify or bust attempt in the first 7 miles, then back off if the bust felt imminent. This would have been exceptionally unwise if we were to believe the face value estimation of 2:31 on the predictor.

Jon - Sasha Science is a mix of the predictor (math program), race/training analysis, and then the gut feel on top of it.

I hope you don't feel we are trying to bash you Sasha. You know we love you man.

Maybe I will just make my own web application that sorts by time, gender, and age group. I am learning Visual Studio, ASP.NET, and SQL Server right now, and this would be a good little project for me to tinker with. I'll let everyone know when I have it online.

No, forget about it, I'm too lazy! I'll just keep pestering Sashing and asking him to do more work. :-) :-)

Okay, but just out of curiosity, why do you have Joe Wilson's flat, sea level equiv slower than his real time? Was Austin short that year, or downhill? Or below sea level? Just wondering...

Look what you've started James!

Paul - Austin had 300 feet elevation drop when Joe ran it. Also, if you have Ogden KML file, e-mail it to me, I'll use it to fix KML import, then import Ogden to test it, and we can see what the Course Tool has to say about it.

Regarding the future of Top Runners page. That version will always be manual edit - I do not mind keeping an eye on every sub-2:30 marathoner. However, at some point I will create a self-reporting unverified top list database as well. This is a lot more work than keeping an eye on a few fast runners and applying Sasha Science to them every time they run a good race.

The Course tool does have one problem, though - it is not elevation-sensitive when judging the uphill slow-down. However, for Ogden vs St. George it would not be that big of a deal because you are going uphill at reasonably comparable elevations in both, and the uphill slowdown was tuned at 4700-4800 feet, which in the range, I think a little higher than St. George, and a little lower than Ogden uphill. I imagine it would be rather off on the side of overrating performances on rolling hills courses at sea level.

Sasha, I can generate an Ogden KML and send it to you.

I figured this was going open up a can of worms, but sometimes things need to get thrown around. I still love Sasha too, but we definitely don't see this one the same way, and thats okay. I am not even one of the top runners anyways so I should probably mind my Ps and Qs, but if we are going to state who the top runners are it should make sense. I think Paul is the best judge of all this since he has the two fastest time either way. My vote is still for the fastest time. Also I was trying to kill this one like we did Steve's VO2 max blog a couple months ago.

James - a runner should be able to express his opinion regardless of how fast he is. If you are slower, that does not make your opinion any less valuable.

On another subject, I've been thinking about your training and results last night, and have come to the conclusion that you have standard A qualifier in St. George within your reach once the next window opens. Some things need to change, of course, otherwise you would have gotten it already. The first step would be (once you recover from St. George), 70 miles a week needs to become a staple survival mileage as soon as possible. Fairly soon it needs to become 80. No skipping days or cutting your run short because of not feeling good, be determined to get in your run on a daily basis, skipping a day is like trying to fix your hair while driving 80 miles an hour on a windy road. The diet needs to be improved as well, "pretty good" might carry you barely past 2:30, but for you to run under 2:20 it would need to become exceptionally good, not just "pretty good". Otherwise you'll find yourself too sick to run the appropriate mileage.

I think we should hold a poll/fundraiser. 1$ donated to the blog equals one vote for raw times instead of "course adjusted". You can vote $5, $10 or whatever you want. Polls close by midnight tomorrow. Sasha, if you get at least $100, would that be enough to change it?

Thanks Sasha. I still think you have a couple of good shots too, and I would love to see us both make it one day. I think you still need to make your easy days easier as well as your taper (Paul is a good example of reaping those benefits). I think you need a few more days of recovery after a big race too.

I had a lady that coaches at Ogden, and didn't quite make the trials qualifier last week, get on my blog today. She has been telling me that I should have Paul Pilkington coach me so I can make trials. I say who needs Pilkington when I have friends like you, Paul, Logan, and other great bloggers. Between everyone's theories and examples I should be able to figure something out in the next few years that work for me.

I wasn't there, but my guess is that the Ogden adjustment is overrated just based on the fact that you have 4 adjusted PR's set that day.

I am very interested to see how things go for the Utah runners at the trials and you in Memphis as a comparison of of the speed of high elevation/downhill races vs relatively flat sea level races.