USATF Road Running Technical Council
Road Race and Finish Line Management

[Previous] [Contents] [Next]


The final step is to reduce the mass of data gathered at the finish line and mauled over by the awards searchers to produce a list of runners, times, age/sex, hometowns, and other tidbits that accurately reflects the performances achieved by your runners on that day. If you mail out results to your runners and get the runner's time WRONG, he or she is left with a poor impression of your race.

Preparing the Data

The first step is to arrange your place/runner lists in finish order for EACH finish LINE separately. The manual recording system is set up this way. The place card only and place card/pull-tag systems require that the place cards be sorted into place order unless place boards have been used for the awards search. Pull-tag/spindle systems are in place order already but should be spread and taped with masking tape on the reverse WHILE STILL ON THE SPINDLE.

If you have more than one finish line, the data gathered for EACH finish line must be processed separately. Data from different finish lines should not be merged until the final step and AFTER select times have been used to insure that the assigned times are accu rate. In order to merge results from several finish lines, it is advised that times be carried to tenths of seconds if not to hundredths of seconds. Otherwise, you will find identical times from two different finish lines and no objective way of knowing which runner finished first.

Even if you do not employ computers on race day for timing, bar-code reading, awards search, etc., computers can greatly ease the task of preparing race results for races larger than a few hundred finishers. The first step is to enter the bib-number sequence into a computer file for each finish line separately. Usually this is simply a sequential list of bib-numbers.

When entering bib-number sequences into the computer file, ALL recorded finishers must be accounted for, official finishers and interlopers alike MUST be entered. You may wish to assign an "interloper" number or just enter them as "zero" or assign a negative bib- number.

Once this information is in the computer along with a registration roster, the computer can produce a place order list of bib-numbers with the corresponding name, age and sex for each place.

Using Select Times

If you are working up your results manually, the easiest way to use select times is simply to go through all the select times in finish order. Write the selected times on the recording sheet adjacent to the runner's number or directly on the place card or pull-tag for that runner. REMEMBER that these times are the MOST reliable times you have. If there are discrepancies, the discrepancies almost always occur elsewhere. You can count on these times as being correct. LEAVE THEM ALONE. Write them in INK.

If you are using a computer to help with results processing, take the place ordered bib-number sequence and enter the select times next to the corresponding bib-number. Again, these are the most reliable times you have. Write them in INK.

If you have more than one set of select times, it is convenient to enter one set of times in one color ink and the others in different colors of ink. This allows you to identify which times are from which team. Red ink should be reserved for corrections.

When two teams select the same runner, you can compare the times recorded. One team may be consistently "faster" or "slower" than another team. It helps to recognize how each team compares with the printing timer (primary timing system) when trying to identify which primary time to the selected time. This is why select time teams should be ON the finish line so that the times closely reflect the primary timing system.

If you are using a manual recording finish system and have spaced your recording sheets so that bib-numbers will match up with your printing timer tapes, simply lay the tape in the space provided and align the tape so it matches the selected times written next to the bib-numbers. Start from the beginning and, as long as the two lists remain synchronized, tape the time tape to the recording sheet in its proper position.

If you are using display boards or "spread" pull-tags on the spindle, you will need to identity those times which correspond to selected times on the printing timertapes (you may wish to enter the bib-number on the timer tape next to those times). Next, count the number of times BETWEEN the two selected times. Then, count the number of place cards or pull-tags BETWEEN those with the selected times written on them. IF they match, you may write in the times from the timer tape directly on the place cards or pull-tags for each runner lN~BETWEEN the selected times. Do this in PENCIL. Not only does this distinguish primary times from selected times, it allows you to correct errors more easily.

If you are using a computer system, the procedure is the same as for place cards or pull- tags but it is much easier to count the number of times in between those selected. Again, if the sequences match, enter the times from the primary system in pencil.

If the time/place sequence and the place/runner sequence do NOT match, there are two possibilities. The most COMMON problem is too many times. Extra times usually occur due to interlopers that escape before being recorded in the chutes. If you have TWO printing timer tapes, compare them.

If you need to eliminate an "extra" time, delete the FASTEST time in the block IN-BETWEEN the selected times. This assures that the assigned times are NOT FASTER than the runner actually ran. Table I gives an example of how this works.

TABLE I. Example of Using Selected Times. In this example, there are more primary times than runners recorded. Delete the FASTEST primary time in-between the selected times.

Primary     Selected   Bib       Assigned
Times       Times      Numbers   Time

44:34.2       44:34       # 637       44:35
44:35.6                   (delete this time)
44:36.1                   #1421       44:37
44:37.2                   # 259       44:38
44:37.9                   #-100*      44:38
44:39.5                   #1822       44:40
44:40.1                   # 690       44:41
44:40.4                   #1426       44:41
44:42.6       44:42       #1467       44:43
*interlopers assigned negative bib-numbers

If you have too few times, check to see if you have too many times in the preceeding or following blocks. This may occur if you have runners shifting positions in the chutes. This may also occur if your pull-tags gatherers collect two or more tags BEFORE spindling and have reversed them. In either case, there is little that can be done since you don't know which runners are in error. However, detection of errors of this type tells you where the weaknesses in your finish system lie. This is also why a redundant place/runner system is suggested when using the pull-tag/spindle method. If such a backup system (voice tape recorder) is available, check to see if the discrepancy can be resolved.

If you have too few times, a time needs to be "created." The "created" time should be equal to the selected time AT THE END of the block in which the error occurred. Again, this assures that the assigned times are NOT FASTER than the runner actually ran. Table II gives an example of this procedure.

TABLE II. Example of Using Select Times. In this example, there are fewer times than recorded runners. An extra time needs to be "created" which is the same as the selected time at the END of the block.

Primary     Selected   Bib       Assigned
Times       Times      Numbers   Time

44:34.2       44:34        # 637       44:35
44:35.6                    #1201       44:36
44:36.1                    #1421       44:37
44:37.2                    # 259       44:38
44:37.9                    #-100       44:38
44:39.5                    #1822       44:40
44:40.1                    # 690       44:41
44:40.4                    #1426       44:41
(time created)             # 94        44:43
44:42.6       44:42        #1467       44:43

If you have a single finish line OR if you have separate finish lines for men and women, you are ready to produce final race results. If you are producing results manually, simply transcribe place/time/ name/age/sex information in sequence, deleting interlopers.

If you are using a computer, create a sequential file of times from your bib-number sequence work-sheets. Enter times for ALL finishers but program your computer to delete interlopers. Check to make sure the bib-number and time sequences match properly. If you don't have the same number of bib-numbers and times, something is wrong.

If you have several finish lines, you are now ready to merge the results from each finish line to produce an overall list. If your times have been entered to tenths or hundredths of seconds, you should have relatively few "ties." Otherwise, you will encounter "ties." If you intend to "break" ties, use a tie-breaking scheme which is nondiscriminatory, i.e., don't break ties by age, sex, or alphabetically. Give each runner an equal chance. You can use a random number generator or simply use finish line sequence, reversing order each time a choice is to be made.

The Select Only Mode

Although a select-only mode is NOT recommended it is sometimes used to handle high finishing rates. In this case, not all runners receive a "real" time, i.e., times need to be "created" in-between the "real" times.

There are three ways to create in-between or "interpolated" times. The simplest way is simply to assign ALL runners not explicitly timed, the NEXT SLOWER real time, in accordance with the way records are kept. This insures that no runner is assigned a time FASTER than they actually ran. This produces gaps in the continuity of the times reported but does serve the function of identifying real and interpolated times.

The most common method is to assume a linear distribution of times between the real times. Thus, if you have ten runners between real times 5 seconds apart, e.g., 46:10 and 46:15, times should be spaced 5/(10+1) seconds apart or 0.455 seconds. The real times here are assumed to be read from a running watch, hence the 46:10 should be reported as a 46:11. Including the second real time in the five second range gives 11 runners to spread out uniformly over 5 seconds. This is where the formula comes from. Table Ill shows how this works. Note that the "real" times are flagged with an * to conform with record-keeping requirements.

TABLE Ill. Example of Producing Interpolated Times in the Select Only Method.

Real	       Calculated    Race
Times         Times         Results

46:10*          46:11.00         46:11*
                46:11.45         46:12
                46:11.91         46:12
                46:12.36         46:13
                46:12.82         46:13
                46:13.27         46:14
                46:13.73         46:14
                46:14.18         46:15
                46:14.64         46:15
                46:15.09         46:16
                46:15.55         46:16
46:15*          46:16.00         46:16*

The third way is rather uncommon but is intended to produce a "more realistic" distribution of times. It does nothing to improve the accuracy of the assigned times. Note that the time gap (inter-arrival time) as assigned above between successive runners arriving at the finish line is the same between 46:11 and 46:16. The third method is to assume the inter- arrival times are RANDOMLY distributed. This is not true either since a considerable amount of non-random "clustering" is evident, even at high finishing rates. However, it does look more realistic.

The procedure is to select random numbers (pseudo-random numbers when using a computer), one for each time to be interpolated plus one for the next real time. Add these together and divide them into the time difference between (adjacent) real times. When you multiply this by each random number, you obtain a "random" inter-arrival time between adjacent runners.

Results Format

Your race results should contain all the information that is needed by the various persons that will use your results. Some groups have several results formats depending on who the results are being sent to. Newspapers are interested in the winners and to a lesser extent on complete lists of finishers.

At the other end of the scale, the National Running Data Center which maintains national rankings and records, is interested in EVERYTHiNG. Not only does the NRDC want COMPLETE results, it also NEEDS age/sex information on each runner and would also appreciate receiving date of birth and address information. Note that address lists submitted are kept confidential and not released without authorization.

The race result format drastically affects the time required to process your race results for national rankings, especially for large races. A sequential listing for a io,ooo person race may take 6 hours to process. The same results, broken down by sex and into ten year age groups, i.e., all 40-49 men are listed sequentially in their own list, etc. may take only one hour.

Results should be in COLUMNAR format. The condensed format used by newspapers is exceedingly difficult to scan for specific information. Likewise, if the information submitted is in TWO separate lists, e.g., place/time/name AND bib-number/name/ age/sex/address separately, it should be possible to quickly cross-reference the lists. This is accomplished by providing an alphabetical entry roster OR, if the entry roster is in bib-number order, by giving bib-numbers with the race result list. Otherwise, cross- referencing is far too time consuming.

The order of information is less important. The most convenient form for processing would match the format used in the national ranking lists, e.g.,

time name age town st date of birth
43:59 Joan Smith 41 F Anytown AL 09/24/44

Your race results should be headed by the DATE of the race, the name of the race, distance, and the town the race was held in. If the race was held on a certified course, the certification code number should be given, e.g., "Race was held on certified course AL-85099- PR." This not only shows the course was certified but it identifies the particular course used.

The form used for completing the documentation needed to process race results for national records and/or rankings is illustrated. This helps you organize your official times, synchronize your primary timing device, and lets us know what was done at your race.

Go to Next Section: Scoring Races by Computer
Previous Section: Award Search and Posting Results
Back to: Table of Contents