There is only one thing more annoying than spotting an error in my statistics shortly before publishing a race analysis and that’s not spotting it. Unfortunately, an error slipped through in my Lake Placid analysis on Monday: I included results from the 2014 race forgetting that due to stormy conditions many athletes were pulled from the swim after a single lap. These athletes continued the race and their times are included on Ironman.com complete with their apparently faster swims. It’s why Kyle Buckingham appears to be 9th overall in the Ironman.com results and it’s why I never did an analysis of the 2014 race.
I’ve been informed of other sources for the Ironman Lake Placid 2014 results, but unfortunately my current tools can’t easily extract the details from them. To ensure a fairer comparison it’s easier to simply drop the 2014 results. Knowing times from 2014 were affected by this issue there was a chance that they were skewing some of the charts so I thought it best to produce new medians and distributions covering only 2002-2013 (excluding 2005).
Fortunately it appears that 2015 didn’t have much impact on the age group medians. Swim times are more comparable now and they were always going to be the most heavily affected part. Other than that the 2015 race still comes out slower on bike an run.
It’s the same with the distributions too. The swim comes out closer, but the rest of the race remains that much slower than usual.
So not really a huge change to my earlier analysis, but I wanted to be sure that the drop off in times wasn’t just the result of an anomalous results set. You can read the original analysis here.