Winning the Hakone Ekiden: greatest comebacks, 1986-2016

In the last two years I’ve gotten hooked by the Hakone Ekiden relay race due to its mixture of tradition, athletic excellence, and aesthetic beauty. I saw the 2015 race in person and watched most of the 2016 edition on TV. As much as I enjoyed it, I was mildly disappointed by the lack of drama in the race for the top spot. I noticed that the team finishing 1st place at the end of Day 1 (stage 5) was the same team that finished 1st overall at the end of Day 2 (stage 10). This got me thinking: is the winner of the Hakone Ekiden always a foregone conclusion by the end of Day 1?

In some ways the answer is yes: Day 1 results are extremely predictive of the overall winner. But there is significant drama in the race for second and third place, as well as the occasional comeback win, as in the case of Asia University’s victory in 2006.

Here are five observations:

1. Final-stage comeback wins are rare

In the 30 races between 1986 and 2015, the eventual winner has been in 1st place at the end of the 9th stage (the 2nd-to-last stage) in 28 of 30 races. In those two races (2001 and 1986), the eventual winner was in second place going into the final stage. And dramatic comebacks after stage 5 are also quite rare. From 1986-2015, the eventual winner has been in 1st or 2nd place at the end of stages 5 thru 10 in 27 of 30 years. The three exceptions are:

Asia University came back from 7th place at the end of stage 6, and 5th place at the end of stage 7 in 2006
Juntendo University came back from 5th place at the end of stage 6 in 1986 (and from 3rd place at end of stage 6 in 1987)

2. It pays to be good early (especially in recent years)

Not only are late comebacks rare, but recently we’ve seen that eventual winners are generally at or near the top of each stage throughout the race. The following charts shows the stage-places of the eventual winner for the years 2005-2016:

3. The 5th stage of the race is ESSENTIAL for eventual winners

While the first four stages are generally flat, the fifth stage climbs almost 900 meters into the dramatically into the hills of Ashinoyu 芦之湯.

Due to stage five’s difficulty, the difference between the fastest and slowest finisher can be dramatic. Between 2005-2016, the average difference between the fastest and slowest stage 5 finisher was 12 minutes, with a difference of 30 minutes in 2012. By contrast, the difference between top and bottom finisher for other stages is between 3-6 minutes.

Stage 5 is especially important for teams that end up winning the race. As seen in race data from 2005-2015, the eventual race winners used stage 5 as a way to significantly improve their position in the field. Following stage 5, the eventual winner is generally in 1st, 2nd, or 3rd place, often making up significant ground in this stage, as seen by the dramatic convergence following stage 5 in the chart below:

The above chart indicates the impressive comeback by Asia University 亜細亜大 in 2006, which was in 7th place following stage 6, a very unusual position for an eventual winner. An account of the Asia University victory can be read here: 往路６位、復路２位で総合優勝！気まぐれな勝負の女神、最後は亜細亜大に微笑んだ

4. The 5th stage is not as significant for 2nd place teams

Although stage 5 is still important, we don’t see as much of a convergence around stage 5 as we do for eventual winners. The graph indicates that eventual 2nd-place teams are generally in 1st to 5th place throughout the race.

Komazawa University 駒澤大学 posted an impressive comeback from 18th place at the end of stage 1, finishing 2nd in 2010: