Swim
2006: 161 broke an hour. Field average: 1:17:29
2007: 41 broke an hour. Field average: 1:25:41
Bike
2006: Field average: 6:33 (conditions were hot)
2007: Field average: 6:37 (conditions were windy)
What has more impact on the bike? Heat or wind?
Can’t compare the bike courses directly. TOTALLY different course except a short out and back.
Jodi
How about the run?
dude you have way too much time to do nothing with.
I was wondering the same thing. I bet the run splits were WAY faster this year. I think the run is where you see the effects of the heat.
Jodi
Agree. I’d like to know the run split average.
How about the run?
2006 Average Run: 5:23:08
2007 Average Run: 4:57:47
What did we ever do before spreadsheets?
I race Cd’A both years and can say that while the temps were 30 degrees different (for the high), the wind on the bike was not much different either year. I didn’t think either year was bad, I’ve ridden the course many times both years, and had much worse winds on some training days, but people griped race day both years. Guess that depends on what you are used to, but it was not significantly worse either year.
Besides, a headwind SOMETIMES means a tailwind in parts, but heat is bad through out the course. I know the slow times definitely had to do with the heat.
What did we ever do before spreadsheets?
Spreadsheets, schmedsheets.
With that large a sample size, all you have to do is go to the race website, results page, select the results for the whole field, and list by run time (or swim time, or bike time). For that sample size, the average value is equal to the median value, which is the person who finished that segment in the middle of the field.
For that sample size, the average value is equal to the median value
That’s a good approximation if the time distribution is gaussian. But, this is ST, man! We can’t have such sloppiness here.
The 2007 median value is 4:53:18, a full 4:29 faster than the 4:57:47 mean time.
What were you thinking? 
Not in my experience. Did you double check that the result generator correctly line up the results? Sometimes it sequences the results different due to a change in the number of decimal places, e.g. 9:xx hrs vs 10:xx hours.
Not to mention that last year was a pro womens race and this year was a pro men’s race…so the pro men’s times are going to throw off any average anyway…I say throw out the first and last 20 and then take a look.
2006 Average Run: 5:23:08
2007 Average Run: 4:57:47
Wow! I’ll chime in here again and note that if you are really serious about improving your place performance in an IM the run is where you can really have some impact. Running even slightly “faster” is going to yeild a huge jump up in your overall placing.
**The 2007 median value is 4:53:18, a full 4:29 faster than the **4:57:47 mean time.
When I look at the 2007 CDA run results, I see 1941 finishing run times, making the 970, 971 ranked run finishing times the median finish times, which were 4:48:36, well below your average value. How did you calculate your average?
Edit: I sit corrected. I went back and resorted the data on IMCDA and got the 4:53:18 median value, with 2086 run finishers. Something about having the bib numbers in the screening screws up run data.
I also did the Excel exercise and got the 4:57:47 average value. Bizarre that there is a 4 minute difference (1.3% difference).
I think looking at averages isn’t a good approximation of the course, given the amount of newbies at these things. The “100th” seems to be a good way to do it, as you did with the swim. In a dozen IM’s, I have never seen as many people finish between 11 and midnight as this year’s IMCdA.
It would be interesting to see the 100th place time for the bikes and runs (and a comparison against other IM’s, particularly LP as some said it was a harder ride than LP, which I find hard to believe…and I certainly don’t think it was “windy” having done IMAZ…)
In terms of judging your overall performance, comparison to the median might actually be more useful than the mean.
But the two values are not likely to be equal, even with a very large number of samples. In the case of an IM run, the median is likely to be faster due to walkers whose slow times distort the distribution curve into a non-normal dumbell shape and subsequently pull down the mean more than the median times. For an extreme example, consider a field of 1001 runners who finish in 3:00 and 999 walkers who finish in 5:00 (case 1) or 7:00 (case 2). The median will be the same in both cases (3:00), while the mean will be 4:00 (in case 1) and 5:00 (in case 2).
Agreed if judging your overall performance to others within the same race.
But for a comparison to of one race vs. another, I think the mean is MORE statistically relevant than the median. The median eliminates the impact of statistically relevant data. Example: In 2006, I walked the marathon at CDA with 2 other people, it took us all over 7 hours. Each of us are capable of going much faster and in some way we all were impacted by the heat. Our results SHOULD significantly alter the data, we were all 2 hours+ slower than expected and it was related to conditions. The median is useful when there are statistical anomolies (outliers), there is something that doesn’t make sense about the data point. In this case there is a reason for the data AND you have somewhat (albeit non-scientific) spec limits. If someone finished the course at 4am, because a mountain lion chewed off his arm, you should maybe consider that an outlier, but anyone inside the 17 hour cutoff is relevant data.
2006 Top 50 age groupers avg. time = 10:19 (took out pro women)
2007 Top 50 age groupers avg. time = 10:19 (took out pro men)
2006 M4044 6th place finisher 10:11
2007 M4044 6th place finisher 10:11
Looks to be identical race, except slower tougher swim, faster cooler run '07 and faster swim and brutal run '06…
Pick your poison
Note: above are approximations, but close enough to serve the purpose
If someone finished the course at 4am, because a mountain lion chewed off his arm, you should maybe consider that an outlier…
I love it! But the real question is…
If “mountain lion guy” lost his arm while running over the mat between swim loops, would Monty concede that he could rightfully complain about the swim conditions? 
