8/27/07 - 9/2/07 Like A Chicken With Its Legs Cut Off (270 minutes, 1 tempo run, 1 steady state)

Sunday, 9/2/07 (45 minutes)
I had a cramp so it's probably my period or something. Seriously I need to update this log more often and not run just 45 minutes like a little bitch.

Saturday, 9/1/07 (no run)
It was really hot, and the clicking in my knee was very disturbing. I think all the roads I've been doing are getting at it.

Friday, 8/31/07 (2 x 2 mile tempo, reps)
Used the new Asics DS Trainers for a tempo on the track. I missed team practice after being completely unable to face the prospect of getting myself up on 4 hours of sleep to run repeats on an asphalt hill. My legs simply wouldn't have accepted it. Turned out to be a good decision, since later that day I felt great.

2 x 3200m, 10:33, 10:27. 3 min lap jog in between. I felt controlled, and just focused on relaxation, trying to take as few splits as possible (just first lap, mile, and final time). The new shoes felt great, although my feet felt hot a couple of times. I did 4x200 untimed afterwards, just to get the feeling of cruising along.

That friendly old guy who comes to some of our meets was there, and told me that Matt and I were running too fast. Oh No!

Thursday, 8/30/07 (85 minutes, reps)
Full arroyo in the morning, went ahead by myself because I didn't feel like running Matt's pace. I felt like crap that morning. I haven't gotten the right sleeping schedule down, waking up on a few hours of sleep for practice, then inevitably sleeping during the day, so that again I can't fall asleep when I want to the next night, and the process repeats.

Afternoon I got new shoes and did a very short warm up, 4x200 in 34, 32, 32, 30, and cool down. The goal of the reps is to keep a tiny bit of stimulus there for the fast twitch fibers, so that when I need them to run serious intervals later, they'll be there.

Wednesday, 8/29/07 (45 minutes, strides)
Ran the traditional first-day Huntington with the team, about as easy as possible.

Matt's cousin, Don Kiesz

As you know, he's half-ass.

Tuesday, 8/28/07 (45 minute tempo)

5xarroyo tempo 8:55, 8:55, 8:59, 8:59, 9:00

Riding my bike up to the gym, I remembered that Kenenisa Bekele, the fastest distance runner ever, does not conform to the "skinny runner" stereotype. The reliability of internet data is questionable, and different sites list Bekele as anywhere between 160 and 174 cm. But the predominant stat is 160 cm, 54kg (5'3", 119 lbs for you idiot Americans). So Bekele's BMI (weight/height^2) is 21.1 (technically, kg/m^2, but by convention the units are suppressed).
I spent some time this afternoon avoiding studying by mining for more runners' stats. I didn't find a jackpot reserve of many elite runners with heights/weights all listed right there for me, but I found a few sources, especially the USATF athlete bios page. Here it is:

athlete	height (cm)	mass (kg)	distance	BMI
alan culpepper	185.4	59.0	42000	17.2
yvonne murray	170.2	50.3	3000	17.4
greta waitz	168.9	49.9	42000	17.5
sebastian coe	175.3	54.0	800	17.6
sileshi sihine	165.0	48.0	10000	17.6
gebremariam	178.0	56.0	10000	17.7
dathan ritzenhein	172.7	53.1	10000	17.8
deena kastor	162.6	47.2	10000	17.9
sara slattery	170.2	52.2	5000	18.0
paula radcliffe	173.0	54.0	42000	18.0
abdi abdirahman	180.3	59.0	10000	18.1
hazel clark	177.2	58.1	800	18.5
wendy sly	166.4	51.3	3000	18.5
lauren fleshman	172.7	55.3	5000	18.6
jen rhines	160.0	47.6	5000	18.6
ryan hall	177.8	59.0	21000	18.7
hicham el guerrouj	179.0	58.0	1500	18.7
paul tergat	182.0	62.0	42000	18.7
anthony famiglietti	175.3	57.6	5000	18.8
kara goucher	170.2	54.4	10000	18.8
shalane flanagan	165.1	51.3	5000	18.8
matt tegenkamp	185.4	65.8	5000	19.1
tatyana kazankina	161.3	49.9	1500	19.2
galen rupp	180.3	62.6	10000	19.2
chris lukezic	172.7	59.0	1500	19.8
adam goucher	177.8	62.6	5000	19.8
treniere clement	160.0	50.8	1500	19.8
meb keflezighi	170.2	57.6	42000	19.9
steve cram	186.7	69.4	1500	19.9
dan lincoln	190.5	72.6	5000	20.0
sally gunnell	167.6	56.2	400	20.0
brian sell	177.8	63.5	42000	20.1
ingrid kristiansen	168.9	58.1	42000	20.4
bernard lagat	172.7	60.8	1500	20.4
haile gebrselassie	164.0	55.0	10000	20.4
craig mottram	188.0	73.0	5000	20.7
herb elliot	179.1	66.7	1500	20.8
khalid khannouchi	165.1	56.7	42000	20.8
leonel manzano	165.1	56.7	1500	20.8
steve slattery	177.8	65.8	5000	20.8
liu xiang	188.0	74.0	100	20.9
wilson wipketer	172.0	62.0	800	21.0
jeremy wariner	182.9	70.3	400	21.0
kenenisa bekele	160.0	54.0	10000	21.1
mbarak hussein	172.7	63.5	42000	21.3
alan webb	175.3	65.8	1500	21.4
dan brown	175.3	65.8	42000	21.4
linford christe	189.2	77.1	100	21.5
kip keino	175.3	66.2	1500	21.6
david krummenacker	188.0	77.1	800	21.8
sam burley	175.3	68.0	800	22.2
steve prefontaine	175.3	68.9	5000	22.4
jorge torres	170.2	65.8	10000	22.7
jonathan johnson	175.3	70.3	800	22.9
shawn crawford	180.3	74.8	200	23.0
emil zatopek	174.0	69.9	10000	23.1
andrew rock	185.4	79.4	400	23.1
lashawn merritt	188.0	82.6	400	23.4
kerron clement	188.0	83.5	400	23.6
nick symmonds	177.8	74.8	800	23.7
brian clay	180.3	83.9	0	25.8
breaux greer	189.2	104.3	0	29.1
christian cantwell	195.6	136.1	0	35.6
akebono	203.2	234.5	0	56.8

Matt's and Katherine's recent posts about runners' weights had me thinking about the question of whether it makes sense to diet for cross country. My gut reaction was, "don't worry about your weight, put the scale away, eat healthily and it will take care of itself." In fact, looking at the list, Bekele, Prefontaine, Torres, and Webb are all accomplished cross country runners (to different degrees). Many of the mile/800 runners (Symmonds, Webb, Krummenacker) are on the heavier side of that list as well.

Basic conclusion: if your BMI is 22 or below, you don't necessarily have to lose weight to be a world-class performer. (Mine, at 181cm and 69 kg, is the same as Bekele's to within the margin of error of the measurement.) It's doubtful that losing weight would help Bekele, for example, because it is very likely that the best runner in the world is already running at near his optimum of just about everything. Running talent is, as Garrett pointed out, highly multifactorial and follows a bell curve. Barring super-mutant heroes, it's unlikely that anyone is far far more talented than the rest of the world, so even if you have the most talented guy out there training well but with a major error (for example, weighing too much), be probably won't be the fastest, because there are plenty of other guys nearly as talented right behind him, and someone else won't make that same error, and will therefor move up and beat him. So I conclude that Bekele is already running at very close to his optimal weight, or else he wouldn't be the world-beater he is. And so I conclude it's at least possible for your optimal weight to come at a BMI in the range of 21-22, or even 23.

For reference here are some heights/weights and BMI's, in the format of a chart I ripped off some random website or other:

Alternatively, use a calculator.

But I have a few more words about BMI. BMI is a dirty fucking lie. The objections to BMI are well-known and common. For example, by strict BMI computation, Bryan Clay and Roman Sebrle, arguably the two fittest human beings alive, are both "obese". The problem isn't that BMI is "bad", just that it's not good enough. It's a one-dimensional number trying to describe a human, which is inherently a multidimensional thing (in most cases, anyway). Other single-number measures of the human body, like body fat percent or hip-to-waist ratio, have their own problems. My main objection to BMI is that it doesn't scale correctly.

A priori, one would think that the BMI calculation should be weight/height^3, since humans are three-dimensional (different use of "dimension" than my previous one), and so make some 10% taller, you also make them 10% wider and 10% thicker.

Not true. If you look at a picture of a person standing alone against a featureless backdrop, you can still get an idea of how tall they are. Tall people just look tall and short people look short, even when you can't compare them to anything. That's evidence that people don't scale evenly in all dimensions. Make someone 10% taller, they get wider and thicker, but not by 10%. So the exponent of height in BMI should be less than 3. But why 2? In other words, why 2.00000000000000? Wouldn't 2.2 or 1.9 probably be closer to the truth?

In order to answer the question, you'd have to do quite a bit of statistics, in order to define what makes a statistic "good". You could, for example, let the exponent float, then optimize BMI's ability to predict certain weight-related medical problems as a function of the exponent. But this procedure, or any other like it, would have its own problems (tall people might have a propensity for diabetes regardless of body composition, for example). You could also try to find an exponent such that BMI and height have no correlation (currently short people have lower BMI's, meaning the exponent is too low.)

So let's say we do that, and find the exponent really ought to be 2.3, but was just made 2 for convenience (BMI is not accurate enough for more than one decimal place in the exponent to be meaningful). Does that make a difference? If I set it up so the exponent is 2.3 and a 170-cm person has the same BMI as the old system, the new BMI formula is BMI = weight/(1.1726*height^2.3). The new "2.3 power" BMI gives a 5% deviation from the old BMI at the heights of 200cm or 143 cm, which encompasses the vast majority of the human population. Final conclusion about BMI: the exponent of 2.000000, while wrong, isn't so bad as long as it's ballpark. However, BMI is still a poor measure of appropriate body composition for a runner, as it can vary widely even at the highest levels of competition, and most likely varies far more widely at lower levels of competition. If you want to race faster, don't force yourself to an unusually-low weight, just train to be faster. That doesn't mean carrying around giant bags of cellulose won't slow you down, but it does mean that Ethiopians aren't winning races just because their country has no food.

One more point - the fact that BMI correlates to height (and it shouldn't) led me to wonder about another catch-all physiological variable runners like to talk about - VO2 max. VO2 max is oxygen consumption/unit body mass. Supposedly, runners with the same VO2 max are equally physically fit, at least in the oxygen uptake system. But that may not be true. If tall runners or short runners are innately more efficient, as measured by speed at a set intake of oxygen per kilogram, then VO2 max really ought to be calibrated to height, which it isn't. Total body weight is another variable we should test in correlation with VO2 max. It total body weight is correlated with VO2 max, then really there ought to be an exponent other than 1 in the formula. There's no way I'm going to find the data easily for that one, though.

Monday, 8/27/07 (no run)
I decided to skip running and get some work done instead. I'm giving some consideration to running 4-5 days a week until I've got my academic situation taken care of. Also, I'll likely be tutoring from 4-6PM many days, so it's unsure how frequently I'll be able to train with the team.