Over the course of the past several years, we have had several requests from different runners to produce some sort of ranking system for road racing. Additionally, the best performance list in our track and field section is our single most visited area on the web site. These two facts combined with the compulsion of the editors of RunWV to score, tabulate, analyze, rank and predict anything and everything led us to decide to begin to examine the feasibility and need for road race rankings.

Need was looked at in terms of "If we build it, will they come?" The recent success of the Clarksburg Area Race Series and the Morgantown Area Grand Prix indicated that there were hundreds of runners in those areas alone that desired to participate in a series and see how they would rank against their peers over a series of races. This, coupled with the fact that runners are generally measurement oriented as a group (with our PR’s, mileage logs, and general fascination with time and distance), led us to believe that there was a need for West Virginia Road Runner Rankings.

The next step was to choose a scoring system. The system needed to be objective, measurable and manageable. It also needed to have a starting point and a basis for continuing maintenance to a runner's ranking and score. After much research and discussion, we decided that the best system to rank runners would be one that would allow them to first be ranked according to the times they were able to run and once that was established, to change those rankings based on how the runners fared against one another.

We wanted to be able to include all road races and rank them so we needed to come up with a system that could compare a 5K time to a 20K time, etc. After exploring several different methods, we choose a method called the Performance Factor. This method was developed in the late 1970’s and basically makes you able to compare a 10K to a 5k, etc. This is done by the following formula:

Time 2 = Time 1 x (Distance 2 / Distance 1) to the 1.07^th power.

The formula was developed by Pete Riegel and was first published in Runner's World, August 1977, in an article in that issue entitled "Time Predicting." The formula was refined for other sports (swimming, bicycling, walking,) in an article "Athletic Records and Human Endurance," also written by Pete Riegel, which appeared in American Scientist, May-June 1981.

After testing the formula by comparing times for runners we believe to be in great shape for any distance (from the Marathon to the Mile), we found this to be a very good indicator of equivalent times for different distances. Since the majority of races (over 50%) are 5K’s, we found it easiest to base our scoring system on the 5K and reduce/increase all distances and times to their 5K equivalent. For instance the equivalent 10K time for a 19:10 5K (the minimum Men’s scoring time) is 40:12. This means that a 40:12 10k is equivalent in performance to a 19:10 5k. This can be done for any distance will yield a good approximation of how any race would be scored in 5k terms.

The next step was to decide a way to assign base points for the 1998 year to give us something to go from for 1999.

The following system was decided on for men and women. (Women’s times)

15:00 (18:00) = 500

15:15 (18:15) = 470

15:30 (18:30) = 440

15:45 (18:45) = 410

16:00 (19:00) = 380

16:15 (19:15) = 350

16:30 (19:30) = 320

16:45 (19:45) = 290

17:00 (20:00) = 260

17:15 (20:15) = 230

17:30 (20:30) = 200

17:45 (20:45) = 170

18:00 (21:00) = 140

18:15 (21:15) = 110

18:30 (21:30) = 80

18:45 (21:45) = 50

19:00 (22:00) = 20

19:10 (22:10) = 0

Basically, one second is equal to two points. If you ran a 14:45, you would score 530 points. Using 500 points and 15:00 (18:00 for women) as starting points enabled us to set up a system that would qualify enough men and women to provide us with rankings that would include 50-100 men and at least 25 women. Using a system that would score every runner that runs a race would make recording keeping nearly impossible (without a staff of hundreds). The number of runners to run at least one race that met the minimum standards last year was nearly 700. Considering the fact that less than 10% of the runners might score in a given race, you could have expected a database of well over 7,000 runners. Comparing each runner against each other in a race of 250 people (roughly 250x250 or 62,500 individual calculations) would have taken days to tabulate and record much less a race that has nearly 1,000 participants.

Although a women running an 18:00 is likely not as good as a man running a 15:00, we decided to use the 3 minute step up for women for two reasons

1. The even 3 minute increment makes converting the points easier
2. We needed to go down at 22:10 to include at least 25 women in our rankings.

Due to the fact that a 18:00 female is not equivalent to a 15:00 male, it is important not to compare men and women when trying to determine who is the best runner in the state when taking gender into account.

To establish your base, you must run 4 times under the minimum scoring time (19:10 5k equivalent for men and 22:10 5k equivalent for women) within one year. For the initial 1998 year everyone with 4 or more times were ranked. For each runner, the higher and lower times were dropped and the median 2,3 or 4 times (depending on the number of scoring races they had ran) were added together and averaged. This creates your "Base Score" and your score for the initial rankings.

Once your base score was established, it became necessary to be able to compare the runners on a head to head basis. Additions and subtractions to your base score will depend on how well you do against other ranked runners within the system. Ranked runners will carry a score into each race. For a particular race, they will be scored against other ranked runners (men vs. men and women vs. women) participating in the same race. If you were supposed to beat a particular runner by 30 points and only beat him/her by 10 points, you will lose points and he/she will gain points. You will be compared to each runner and your base score will change by the average points won/lost per ranked runner participating in that race.

For instance, in race 1 on January 1^st, 1999.

Runner A has a Base Score of 190 points.

Runner B has a Base Score 78 points.

Therefore, according to rankings, Runner A should run 112 points better or 56 seconds faster. (2 points per second)

The race is completed and Runner A runs a 17:41 178 Points

Runner B runs a 18:17. 102 Points

Therefore, Runner A wins by 76 Points.

Runner A was supposed to win by 112 points and only won by 76 points. We will now calculate the point exchange for the race.

The points differential was 36 points (112-76). To calculate the final addition/subtraction to each runner, divide the points difference by four. 36/4=9. 9 Points would be subtracted from Runner A’s score and 9 points added to Runner B’s score. The divisor of 4 is used since it will allow runners that are originally ranked differently to be ranked evenly at the end of a 4-race stretch in which they ran exactly the same time (approximation).

If Runners A and B ran against additional ranked runners in this race, your points differential would be calculated for each ranked opponent, totaled and then divided by the number of ranked runners they competed against.

You are only compared to other ranked runners within the same race and how non-ranked runners finished will not affect the scoring system.

Your score would not change until after the race. (I.E. your current score would be compared to all ranked runners and would not change from runner to runner within the same race.) At the end of the race, you would take into account how you fared against all other ranked runners and that would dictate your new current score for use in your next race.

You might ask the question "Why not just take a moving average of the last 4 races to establish the score". The answer to this is that we wanted a way to compare runners that would not be reliant on times only. As you know, all race courses are not created equally. Some are flat, some involve running up mountains and some are even straight downhill. Unlike track in which 800 meters in Charleston is the same distance and topography (i.e. flat surface) as 800 meters in Martinsburg, a 5k road race can be as different as night a day by just going from one spot in a town to another spot in the same town. This means we can’t simply take the 10 best 5K times in the state and call them the best 5K runners. The only way to rank the best runners is to allow them to compete head to head. To do this, we needed a starting point for scoring purposes so we took an average of the runners (2 to 4) median 5k equivalent times. This removed most of the downhill and tough courses and helped put them on an even basis to start the scoring. Although Runner A may never run against Runner B, both might run against Runners C,D and E. By scoring how each of them fare against ranked Runners C,D and E, we can then conclude how they might fare against each other. No system is perfect or can cover every circumstance, but we think we have developed something that should allow for a fair and objective measurement of road running in the state.

You can be removed from the ranked list for one of two reasons.

1. You must run at least 4 "Scoring" races during a 12 month period. A scoring race is one in which you meet the minimum criteria of 19:10 for men and 22:10 for women. If you fail to do this, you will be removed from the list. You can be reinstated with a new base score once you again run 4 scoring races within a 12 month period.
2. If you "Current Score" falls below 0, you will be removed from the list. This can happen if you run poorly for several races and lose points according to the scoring system described above. You will be reinstated to the list upon completing another scoring race. You will then be treated as a new runner and be assigned a new base score. To do this, you still must have 4 scoring races within the most recent 12 month period.

You will be added to the list once you have completed 4 scoring runs within a 12 month period. The highest and lowest scoring races will be dropped and the average of the remaining two will be your new base score and will determine where you enter the rankings.

1. Only 1 of your scoring races that are used to establish your base is permitted to be in a race that is less than 5K in length. This is done to assure that someone won’t run only short events to assure themselves of a great ranking. We want to rank the best road runners, not a track meet.
2. The maximum points you may gain or lose each race is 25 points. This is done to protect the ranked runner that may be running while sick (hey, they already paid the race fee) or one that is jogging with a friend, etc. We also don’t want other ranked runners to benefit too greatly from this situation.
3. The maximum points you main gain or lose against an individual runner is 50 points a race. This means that the raw difference was 200 points (50=200 divided by the 4 as explained earlier). 200 points is equivalent to 100 seconds which means that a runner either wins or was beaten by 1:40 more than they were supposed to against a single runner. Anything greater than 50 points is an aberration and would skew results if not capped at 50 points. This rule can be differentiated from Additional Rule 2 by the fact that this cap is for a single runner and Additional Rule 2 is the average win/loss to all runners you faced. You may net 50 points against a single runner, but your score can never increase or decrease by greater than 25 points for a single race.
4. If your last 4 races would have established a base that would be 60 points or higher than your current score (by average and median methods), you will be awarded a new base score based on the "Breakthrough Provision". This is put in place for runners that, after a bout of serious training, come in and demolish their old records. If several of these breakthrough runs are done in races with only a few (or no) ranked runners, your score will not rise quickly enough to match your new level of accomplishment. (60 points is equivalent to 30 seconds which
5. Males and Females will be scored separately

The only thing that you must do is to make sure we get race results for any race that you are involved in. We will post the entire results of every race we receive and then will score the ranked runners based on the previously described system. If we don’t receive the race results in a timely manner, we won’t be able to score the runners. We can’t go back and score a race from April 15^th when we have already tabulated results for several races in late April and early June. To go back and use April 15^th would dictate that we would have to recalculate all of the other races we had already scored. Therefore, it is important that we receive the results from a race as soon as possible. We have several methods of delivery available:

Rankings will generally be updated on a monthly basis and will usually occur within two weeks of the last day of the previous month. However, we will be releasing updated rankings every couple of weeks until we are caught up to the current month.

Please email us if you have any questions or comments about the system.