The Head of the River Race - Intermediate times and statistics
The statistics described here are included in the
text file of results sorted by crew number. These notes will help you to interprete the figures.
Actual Split Times
Intermediate times are recorded at Barnes and the ARA office at Hammersmith, dividing the race into three sections. The split times for the crews which finished within 30 seconds of the overall winner are recorded in the first table below. This list of crews is defined to be in Group 1.
The total times ("Total") taken by all the crews in Group 1 have been calculated for each section, and from these totals, the average proportions ("Group 1 Prop") of the overall time taken by crews for each section of the race have been calculated.
The total times taken by ALL the crews in the race have also been calculated for each section, and from these totals, the average proportions ("Group 0 Prop") of the overall time taken by ALL crews for each section of the race have been calculated.
It will be observed that values of "Group 1 Prop" and "Group 0 Prop" are not the same. To keep these notes concise, only the "Group 1" values will be considered here. Group proportions for groups 0 to 8 are used to generate times in the full list of results sorted by crew number, available here as an
<ASCII> text file.
The following symbols are used to help search for information in pages:
- # indicates the number of a crew
- @ indicates the position of the crew in the race as far as one of the timing points
- $ indicates the sequence number for the crew at a location
- % indicates the position of a crew in a section of the race between two adjacent timing points.
If you wish to download a plain text version a page of results without notes for use in a text editor,
then click on the appropriate <ASCII> tag.
Actual split times for Group 1 crews for 2006 Race : Prepared 31-Mar-06 10:42:43
Start Barnes ARA Finish
@001 #001 03:13.87 07:06.94 07:07.69 17:28.50
@002 #003 03:16.80 07:08.21 07:06.68 17:31.69
@003 #004 03:15.79 07:10.49 07:12.82 17:39.10
@004 #002 03:17.68 07:11.48 07:12.49 17:41.65
@005 #009 03:19.97 07:15.17 07:16.36 17:51.50
@006 #018 03:21.68 07:14.43 07:20.22 17:56.33
Total 19:45.79 43:06.72 43:16.26
Group 1 Prop 0.186 0.406 0.408
Group 0 Prop 0.185 0.403 0.412
Expected Split Times
Using the proportions calculated above, 'Expected' times for each crew in Group 1 have been calculated. These are the times that each crew would have achieved if they had raced with the same profile of speeds as the 'average' crew in the group. An example will illustrate the method:
Suppose the average proportions for all crews in the group of times spent over the three stages of the race had been 0.2, 0.3 and 0.5. A slow (!) crew with average speed profile that took a total of exactly 30 minutes for the race would have spent 20% of that time between the Start and Barnes - ie 6 minutes, 30% between Barnes and the ARA (9 minutes) and 50% (15 minutes) between ARA and the finish.
Expected split times for Group 1 crews for 2006 Race : Prepared 31-Mar-06 10:42:43
Start Barnes ARA Finish
@001 #001 03:15.21 07:05.85 07:07.42 17:28.50
@002 #003 03:15.81 07:07.15 07:08.72 17:31.69
@003 #004 03:17.19 07:10.16 07:11.74 17:39.10
@004 #002 03:17.66 07:11.19 07:12.78 17:41.65
@005 #009 03:19.50 07:15.19 07:16.80 17:51.50
@006 #018 03:20.39 07:17.15 07:18.77 17:56.33
Deviations of Actual Split Times from Expected Split Times
In practice, crews do not all have the same profile of speeds. To continue the illustration above with an extreme case, suppose the slow crew above actually recorded proportions of 0.1, 0.3 and 0.6 for the three sections of the race, and took 30 minutes for the whole race. This would mean that they took 10% of 40 minutes (= 4 minutes) between the Start and Barnes, 30% between Barnes and the ARA (9 minutes) and 60% (18 minutes) between ARA and the Finish.
The deviations from expected times for this crew would be -3 minutes, 0 minutes and +3 minutes. Note that the total deviation for the crew is zero.
There could be a story here : "This crew was a middle ranking crew, and they went quite well as far as Barnes. The had a crash at the Band Stand, and it took them over 5 minutes to get going again. More bad luck hit them at Harrods, where they were holed and sank. The crew swam with the boat to the shore, and emptied the water out, then used some sticky tape to patch the hole, before getting back in to complete the course."
In practice, actual deviations from expect times are smaller than those in the illustration. Actual figures are illustrated in the next table which shows the differences between times in the first two tables.
The deviation is calculated as (actual time) - (expected time)
A positive deviation indicates that the crew has recorded a slow time, and a negative that it has recorded a fast time. Slow and Fast indicate the crew's speed profile when compared with the profile for the average crew within the group. Absolute speeds are NOT being compared.
Deviations of actual from expected split times for Group 1 crews for 2006 Race : Prepared 31-Mar-06 10:42:43
Start Barnes ARA Finish
@001 #001 -1.34 1.09 0.27 17:28.50
@002 #003 0.99 1.06 -2.04 17:31.69
@003 #004 -1.40 0.33 1.08 17:39.10
@004 #002 0.02 0.29 -0.29 17:41.65
@005 #009 0.47 -0.02 -0.44 17:51.50
@006 #018 1.29 -2.72 1.45 17:56.33
Proportions and deviations compared with ALL crews in the race
These are not as useful as the Group Proportions described above, but they do illustrate a difference in speed profiles between slow and fast crews. In the numerical order list, the Group deviations are shown on the left, and the overall deviations on the right.
Reasons for and uses of deviations
The timekeepers originally produced tables of deviations to help check the validity of times. It was only later that tables were adapted for publication because they contain information of interest to crews.
Most deviations are small. If a crew has a good patch and gains a length over in a section of the race, there will be a corresponding deviation of about -3 or -4 seconds. It is important to understand that a gain by a length in one section by a crew is balanced by a corresponding loss elsewhere in the race, because the total gain for each crew is zero. Small deviations may be of interest to crews, but are accepted by time-keepers without special investigation.
However, there are some deviations which timekeepers will have investigated.
Any large deviation for a particular crew not also experienced by a neighbouring crew is likely to have been caused by a problem on the water for that crew. Such a crew was number 043 in the 1999 race.
#043 @040 03:17.64 @073 11:06.61 @336 20:36.23
%040 03:17.64 %093 07:48.97 %389 09:29.62 592
-25.94 -25.92 -45.93 -45.48 71.88 71.41
$041 12:40:05.01 $041 12:43:22.65 $042 12:51:11.62 $052 13:00:41.24 12
The boat became waterlogged after Hammersmith, and went very slowly for the last part of the course. Group Deviations were -24.94, - 45.93, +71.88.
Positive and negative deviations of about the same absolute size for a crew
This may be a timekeeper's error. Crew 264 in the 1999 race had times recorded as here:
#264 @337 03:44.81 @198 @292 20:05.97
%337 03:44.81 %105 07:51.00 %358 08:30.16 111*
5.81 6.72 -30.44 -30.85 24.64 24.14
$258 13:29:13.11 $258 13:32:57.92 $257 13:40:48.92 $257 13:49:19.08 2
39:13.11 42:57.92 50:48.92 59:19.08
The extra (fourth) line of figures shows the watch times for the crew at each location (39.13.11 at the start). These watch times are converted to GMT for all calcluations by adding the appropriate watch start time, giving the GMT time in the line above.
The 5.81,-30.44 and 24.64 deviations caught the timekeepers' attention. The small value 5.81 indicates that the crew went at 'normal' speed as far as Barnes. The other two figures indicate a suspected timing error at the ARA. There was no firm evidence of a problem on the water, and the crew was awarded NTT for its time at the ARA. This was the only such error detected by timekeepers in 1999.
The final times as published include the following:
#264 @337 03:44.80 @198 NTT @292 20:05.97
%337 03:44.80 %105 NTT %358 08:30.16 111
5.81 6.72 -30.44 -30.85 24.64 24.14
$258 13:29:13.11 $258 13:32:57.91 $257 NTT $257 13:49:19.08 2
The deviations -30.44, -30.85, 24.64 and 24.14, which are affected by the NTT are unreliable.
Similar large deviations of opposite signs for two crews which pass a timing point fairly close together.
Such deviations may indicate that two crews have been wrongly identified by the timekeepers. If, for instance, one crew is given a time 5 seconds too quick at Barnes and another 5 seconds too slow because the two crew numbers have been interchanged on timing sheets, this will show up in the check lists used by the timekeepers. In cases of doubt, backup timing and crew number records are checked. NTT is given by the timekeepers if they have any doubt about the validiy of any time recorded.