Male Race Results

Female Race Results

Analysis

Before the race the expectation was that Taylor Knibb would build a big lead on the bike (maybe five minutes or more) and that Ashleigh Gentle would run possibly five minutes quicker than Taylor. We had seen something similar happen at least year’s PTO US Open in Dallas when Taylor was able to put 6:03 into Ashleigh on the bike, but then struggled a bit on the run. Ashleigh was able to run 8:07 into Taylor, overcoming a huge deficit in T2 and winning the race in Dallas.

What happened in Milwaukee was a little bit different: Taylor was able to build a lead to Ashleigh, but the lead was only about three minutes at the end of the bike. Then Ashleigh was able to run “only” about two minutes quicker than Taylor. Did Ashleigh (and other participants) “trade” a smaller gap after the bike for a slower run? This question motivated the following analysis.

Adjusting Times

To start with, comparing actual race times is often not a very meaningful way to assess the performance in a race. For example, Taylor’s bike split in Milwaukee was about a minute quicker than in Dallas – but this does not necessarily mean that she rode “better” in Milwaukee. Looking at the other athletes who have raced in Dallas and in Milwaukee, they were often significantly quicker in Milwaukee, for example Ashleigh was almost  four and a half minutes quicker. Did Ashleigh and Taylor ride better than in Dallas – or was the course in Milwaukee overall quicker than in Dallas? (The easiest explanation would be a slightly shorter course in Milwaukee, but it could also be better weather conditions etc.)

To get a better idea about “how quick” the courses in Dallas and Milwaukee were, we can look at all athletes who raced in Dallas and Milwaukee. For the women, this gives us 11 data points (i.e. athletes who raced in Dallas and Milwaukee), for the men we get an additional 13 data points. A simple average shows that the Milwaukee course was just over three minutes quicker than in Dallas, and more sophisticated measures (median instead of average, working with ratios instead of differences, removing outliers etc.) don’t change the picture much. This indicates that a bike time needs to be about three minutes quicker in Milwaukee for a similar performance than in Dallas. For Ashleigh, her “adjusted bike time” was about 1.4 minutes quicker in Milwaukee (three minutes less than the difference between her bike times).

A similar analysis can be done for the run, indicating the Milwaukee run to be one minute quicker than the run in Dallas.

Putting the data into a Graph

We can put these adjusted differences into an x-y graph, and get four different quadrants:

• top right: faster run and faster bike in Milwaukee
• top left: slower bike but faster run in Milwaukee
• bottom right: faster bike but slower run in Milwaukee
• bottom left: slower bike and slower run in Milwaukee

We can also add a line where the bike and run differences cancel out (i.e. a faster bike and a slower run by the same amount or vice versa), shown as a solid green from the top left to the bottom right in the graph below:

The “farther away” from this line an athlete’s data point is, the faster (if to the top right of the line) or slower (if to the bottom left) his or her race has been.

Women’s Analysis

Here’s the graph with the data points for the women Pros in Milwaukee (click for a hi-res version):

A few observations:

• Ashleigh Gentle’s data point is very close to the green line, indicating that she had a similar overall performance in Milwaukee. Essentially she showed a better bike performance but then a slightly worse run performance than in Dallas.
• Similarly, Holly Lawrence had the same performance in Milwaukee and Dallas, and also quite comparable bike and run performances.
• Taylor Knibb may have had a slightly worse bike performance (she is in the top left quadrant), but then ran much better, leading to a better overall performance (and distance to the green line).
• Paula Findlay and Ellie Salthouse raced much better in Milwaukee than in Dallas – also resulting in much better overall positions. (Paula was third in Milwaukee and tenth in Dallas, while Ellie improved from eleventh in Dallas to sixth in Milwaukee.)
• Jocelyn McCauley (19th in Dallas after seventh in Dallas) and Rebecca Clarke (eighth in Dallas, 16th in Milwaukee) didn’t have good days in Milwaukee.

Men’s Analysis

Here’s the same graph but for the men who raced and finished in Dallas and Milwaukee:

Some comments:

• Sam Appleton and Matt Sharpe had the best “relative performances” in Milwaukee – both coming off not-so-good performances in Dallas (Sam was 27th, Matt was 32nd).
• Jason West finished second, impressing with the fastest run split. But he had already run well in Dallas, and his improvement into 2nd place was mainly a result of his better bike performance in Milwaukee.
• Lionel Sanders was able to have a much better run than in Dallas, leading to an improvement from 21st in Dallas to an eleventh place in Milwaukee. His relative performance is in almost the same spot as Taylor Knibb’s.
• Sam Long slightly underperformed compared to Dallas – nonetheless he had great finishes at both events with a third in Dallas and a fifth in Milwaukee.
• Florian Angert and Aaron Royle didn’t race well in Milwaukee after having good results in Dallas – they slipped from fifth to 19th and seventh to 17th.

Prize Money

PTO US Open has a total prize purse of 600.000 US$, paying 30 deep.

Male Race Participants

Female Race Participants

Winning Odds

Male Race Participants

• Kristian Blummenfelt: 65% (1-1)
• Magnus Elbaek Ditlev: 10% (9-1)
• Sam Long: 6% (17-1)
• Jan Frodeno: 5% (19-1)
• Ben Kanute: 4% (25-1)
• Frederic Funk: 3% (28-1)
• Daniel Baekkegard: 2% (39-1)

Female Race Participants

• Taylor Knibb: 54% (1-1)
• Ashleigh Gentle: 23% (3-1)
• Paula Findlay: 10% (9-1)
• Tamara Jewett: 4% (27-1)
• Holly Lawrence: 3% (34-1)
• Katrina Matthews: 2% (42-1)

Prize Money

PTO Asian Open has a total prize purse of 600.000 US$, paying 20 deep.

Male Race Participants

Female Race Participants