WAVA’s SERIOUSLY FLAWED TABLES, Scherf vs. Rupp 10K

The WAVA tables show Rupp’s 28:15.51 performance to be 96.8% and Scherf’s 32:51.20 performance to be a 92.2%. Rupp’s is 4.6% better. (100% = approximate world record level). Converting Scherf’s time to an age-18 male time (using WAVA tables) gives Scherf a time of 29:41, 85 seconds slower that Rupp.

But these WAVA adjustments and the WAVA tables are seriously flawed. The women’s WAVA tables are off by almost 4%. When corrective adjustments are made (see below) Scherf’s time is equal or even slightly better than Rupp’s time.

In comparing the WAVA men’s and women’s world standard times, the women’s time is slower by:

• 10.9% for the 5K

• 10.9% for the 10K

• 10.7% for the 15K

• 10.4% for the 20K

• 10.3% for the half-marathon (21.1K)

• PER WAVA: WOMEN AVERAGE 10.6% SLOWER FOR 5K-21K

The latest 1994 WAVA’s Tables apparently used data that came only from men’s races. Assumptions were made in 1994 about how fast the world’s top women could run. Unfortunately these assumptions were way off the mark. WAVA’s tables indicate that the world’s best women runners are only 10.3% to 10.9% slower than the world’s best men runners at distances from 5K to the half-marathon. The average difference is 10.6% slower for this 5K-21K range.

But current world records (http://www.usatf.org/statistics/records/) show that the world’s top female runners are really running 13.2% to 15.8% slower over 5K-21K, with the average difference being 14.4%, not the 10.6% WAVA tables are using.

If one wants to correctly compare a male performance to a female performance using the WAVA Tables, then one needs to add about 3.8% to the female’s WAVA performance number. For example, Scherf’s 92.2% performance should be adjusted to 96.0% before being compared to other male runners.

In comparing the current world record times for men and women, the women’s time is slower by:

• 14.2% for 5K (14:25 Abeylegesse ’04 & 12:37 Bekele ’04)

• 14.0% for 10K (30:01 Radcliff ’02 & 26:20 Bekele ’04)

• 13.2% for 15K (46:57 Meyer ’95 & 41:29 Limo ’01)

• 15.8% for 20K (65:11 De Reuck ’98 & 56:18 Tergat ’98)

• 14.7% half-marathon (67:59 Meyer ’98 & 59:17 Tergat ’98)

• PER WORLD RECORDS: WOMEN AVERAGE 14.4% SLOWER FOR 5K-21K

Only one of these world records is widely considered to be drug enhanced, the unbelievable 10K 29:32 time by Wang Junxia of China in 1993. So I used the world’s second best time (30:01 by Paula Radcliff in 2002) as the 10K world record. Note, if other women’s records above are drug enhanced, then correcting for that would result in an increase the above percentage numbers. For men, such corrections would reduce the above percentages. In most cases (Junxia’s 10K is the exception), there is not much difference between the current world record time, and the next few best times under it. Certainly not enough of a difference to make a noticeable difference in my analysis here.

RUPP’s VS. SCHERF’s RECENT 10K PERFORMANCES:

Scherf’s adjusted WAVA performance (from 92.2 % to 96.0% to correct for WAVA’s bad data) is very close to Rupp’s 96.8% performance. It is only .8% under Rupp’s performance (or 13.5 seconds slower). Rupp just turned 19 (May 8th), but Scherf won’t turn 19 until Sept.18th. Adjusting for that small age difference, this .8% becomes .6% which is only a 10 second difference between them, with Rupp slightly faster.

This analysis agrees very closely to results obtained using the much more sophisticated Mercier/IAAF points system (http://myweb.lmu.edu/jmureika/track/Mercier/). Mercier’s senior/open tables put Scherf's run equal to 28:22, and the junior tables put it at 28:30. The totally different Mercier method indicates Rupp’s time is somewhere between 7 & 15 seconds faster than Scherf’s.

The 10K records that Rupp and Scherf broke were also nearly identical in difficulty. Per T&F News, Rudy Chapa's 1976 10K junior record would have been the 19th best time in the US last year and the 43rd best time in the world this year. In contrast, Mary Shea’s 1979 10K junior record of 32:52.5 would have been the 11th best time in the US last year and the 47th best time in the world this year.

OTHER FACTORS TO CONSIDER:

Rupp ran in ideal conditions and likely planned and tailored his workouts for this one big race of his. He had a rabbit for the first 17 laps of his race and had 3,000 people cheering him on. He also had lots of training advantages (Altitude tent, underwater treadmill, etc.) under the Nike Oregon Project, and could train and rest like a full time professional runner. He didn’t enter college until this past month.

Contrast this to Scherf, who ran in windy 10-20mph conditions for the first 20 laps of her race (winds quickly subsided after 20th lap). She had no one to follow for any part of her race (she was the rabbit), and she had to pass other slow runners 21 times during her race (5 once, 1 twice, 2 three times, and 2 four times). Rupp had to pass no one. She only had about 20-30 teammates cheering her on and they were largely drowned out by about 300 nearby Columbia Univ. students loudly cheering Bierbaum on. Being a full time student at Harvard certainly limits Scherf’s training/relaxation time. She can’t train like a pro. Harvard’s demanding academic workload possibly adds some stress that negatively impacts her running to a small degree. Scherf also didn’t know if she would have any strong competition in her race until the afternoon before her race. And with strong winds and rain being forecast for several days before this race, she knew there was a good chance that she would not run this 10K and instead go for the record the following week at the ECAC conference meet. With all this uncertainty, she barely tapered her workouts and only ran easy for a few days before the race.

CONCLUSION:

Considering these “other” factors, one could easily conclude that Scherf’s performance was more impressive than Rupp’s, although Rupp’s performance was certainly impressive as well. Their performances are so close that in fairness to both these outstanding runners, we should consider their performances essentially equal. A tie! USATF should have honored them both by naming them “co-athletes of the week”. But this didn’t happen. Perhaps because Rupp has a big media team behind him (Univ. of Oregon and Salazar’s Nike Oregon Project) while Scherf has zero media support. Harvard hasn’t yet put up a “player bio” for Scherf on its athletic website. Here’s their write-up on Scherf’s 10K (http://gocrimson.collegesports.com/sports/w-track/harv-w-track-body.html).

WHAT’S NEXT?

Scherf may run at Regionals (5K), and will run at NCAAs (10K) and most likely in the USA Championship two weeks later. I suspect Rupp will do likewise. It will be interesting to see how they compare under more similar conditions. Scherf needs to get more people cheering for her, as Rupp has found that to be quite helpful. Hopefully the NCAA’s 10K races don’t turn out to be strategic and slow.

And comments here? Idea for new discussion topic: Should the NCAA use rabbits for its championship distance races? Does anyone know if this has been seriously considered by the NCAA?

I agree with every part of your analysis. One thing I want to point out, though, is that Scherf did CHOOSE Harvard. No one made her put herself in sub-par conditions.

Also, we don't know how hard she works at school. I am still convinced that Ivy league schools aren't harder than any others, once you get in.

I have no idea why you're obsessed with this but your thumb is on the scales in more than one instance. In addition to disregarding the actual women's 10000m record, you inexplicably ignore the IAAF listed half marathon WR of 1:06:44 for Meyer ('99). That produces a 12% margin (as does the actual women's 10000m WR) and Meyer, while a fine talent, hardly defines the limit of female abilities. Radcliffe's marathon WR, of course, produces an 8.5% margin. It's absurd to believe that the female/male ability differential varies from 14% at the half marathon to 8.5% at the full.

the WAVA tables are being/have been updated. Use Mercier when comparing similar age times for open/juniors. WAVA is all we have for comparing age-group times at the master's level.

the WAVA tables are being/have been updated.

I have the road race tables, and they seem consistent with current records as listed on usatf.org

elflord wrote:

the WAVA tables are being/have been updated.

I have the road race tables, and they seem consistent with current records as listed on usatf.org

RunScore has an Age Grade function built in based on WAVA (now WMA, technically) tables - I believe the updated ones. Nonetheless, when you are setting the parameters, you are advised to put in a female "correction factor" of around 90 per cent to make things even. I'll ask Alan Jones how he came up w/ that figure.

God help us.

Run37 wrote: Idea for new discussion topic: Should the NCAA use rabbits for its championship distance races? Does anyone know if this has been seriously considered by the NCAA?

Stupid, stupid, stupid suggestion! Why on earth would there be any call for rabbits in a championship race?

I think I once used some of the above rational to "prove" to myself why a race I ran was much better than the actual time. Not sure of the way the WAVA tables are constructed but there are a lot of valid ways to skin the statistical cat when comparing two populations - you obviously believe that the WR's are equal in terms Std dev away from the norm. there is a whole lot more evidence that that is not the case (more comp male runners for starters, longer evolution in the sport, higher average training volume over a much greater number of people I could go on). You find me a study that shows the difference in times between two substantial populations of men and womem (pop that mirror the world pop) who are trained the same way, with the same intensity, and then timed at various distances - and that study shows that there is a 14.4% difference between men and women - then I'll buy it. Till then nearly all significant observations suggest that the Men's WR's should be superior to the women's WR's and certainly there should be no doubt that the men's lists are far deeper making comparative postions misleading at best.

One observation: last year, when Rupp broke Lindgren's record (by a fairly wide margin), Chock broke the girl's (albeit soft)5000 record, by an even wider margin. Now Rupp breaks the 10K AJR (technically), and his effort is mirrored the very same weekend by another girl, Scherf. Not sure what it means, but it will be intersting to see what develops on both sides of the gender line the next few years. I will side with the analysis that Rupp's effort is definitely more noteworthy, considering the competititve milleau, and I also suspect that Rupp will outlast both Chock and Scherf in the long term of development.

Today many road races list one’s “WAVA performance level” alongside one’s finish time so that runners can see who really had the best race performance after adjusting for age and gender. The word needs to get out to female runners that they are performing a lot better against their male competitors than what their WAVA performance level indicates. For the last four years WAVA folks having been saying they are about to release updated performance tables. We’re still waiting. Apparently the women’s tables may continue to be based on men’s running times. WAVA’s shortcomings need to be publicized so there is pressure on the WAVA people to make some long overdue corrections.

Sorry, I accidentally overlooked Meyer’s “pending” half-marathon world record of 66:44 and used her certified world record of 67:59. I agree, the male/female differential is 12.6% for the half marathon, not the 14.7% I stated. That changes my calculated average differential for 5K-21K to 14.0%, and indicates the WAVA tables for women are off by about 3.4%, not 3.8%. That’s still a huge error. So Rupp’s performance (using my WAVA corrections) is perhaps 17 seconds better than Scherf’s, not the 10 seconds initially stated. That’s still in close agreement with the Mercier adjustment.

It’s not absurd the male/female differential drops from 12.6% for the half marathon to 8.4% for the marathon. That’s what the world records now show. The marathon is a far different race than half/20K/15K races where strength/speed is more of a factor, and where men have more of an advantage. In ultra-distance running events, 100K and over, the top women occasionally beat the top men. The male/female differential at these ultra distances may be close to zero.

I ignored Wang Junxia’s 10K world record because it is the one world distance record that is totally absurd. Wang broke the former 10K record by 42 seconds!! Looking at the progression of world record times across a variety of distances, Wang’s 29:32 world record is so out of line that I felt compelled to ignore it. She broke the 1500, 3000 & 10000 world records within a span of 6 days and then conveniently stopped competing outside of China where she could be drug tested.

Championship distance races should be won by runners who can run that distance the fastest. To me "racing" is seeing who can get to the finish line in the shortest amount of time.

Slow strategic races are often won by slower distance runners who have great kicks, but who can’t run really fast times. Sometimes distance races go out so slow (no one wants to lead) that spectators start booing. To me that's not racing. It's embarrassing to watch. Prefontaine raced the entire distance of his races and his races were exciting to watch. I'm a fan of Pre's famous quote "To give anything less than your best is to sacrifice the Gift. But today, in closely competitive championship races, nearly everyone is scared to take the lead early on, and runners only give their best during the last lap or two.

This year's Indoor NCAA Championship 3K race was run at a slower pace than the 5K race -- except for the last two laps. This sport is going to lose fans if the first 8 minutes of a 9 minute race involve the entire field of runners jogging slowly around the track in a tight pack that only starts racing (running fast) the last lap or two. Perhaps races like this could be set up similiar to the Olympic marathon. Let the runners run around on the roads at a leisurely pace. Then with 400 meters remaining, they can be let back on the track to begin their race to the finish.

Perhaps adding a rabbit to these races will get the runners racing the entire distance.

Congratulations on your efforts to explain WAVA, I wish I had as much time to spend on something productive. But, I can tell this is your passion, so good for you.

I've always been suspect of basing race results/awards/prize money on the WAVA tables. While they are interesting to look at, I believe those that cross the finish line first should be rewarded. To me using the WAVA to award results is like having gymnastics judges grade who they thought ran the hardest.

I'm originally from Kansas and I know that there and in OK they base a lot of their awards on the WAVA standards. I know I'd be upset if I won a race and didn't win the prize. Thank God for TX, where we haven't subcombed to the WAVA madness!

have you been to one? like any college it's as hard as you want to make it. it just so happens that most students end up working a ton more on the average than others from other colleges.

I agree, I agree, I agree.

Everyone wants to run in the wind shadow of the others, because if it is a close race, the one in front expends the most energy. The races are more like: go ahead, you first; no you; age before beauty; no I insist, you go first.... and so on.

Adding a rabbit to break the wind evens things up for everyone. Nobody is placed at a disadvantage because they all have a wind shadow to run in, including the "leader".

god this post is hilarious...

Run37 wrote:

Championship distance races should be won by runners who can run that distance the fastest. To me "racing" is seeing who can get to the finish line in the shortest amount of time.

yeah that's nice to see a lot. but racing is also about tactics.

But today, in closely competitive championship races, nearly everyone is scared to take the lead early on, and runners only give their best during the last lap or two.

today???? tactical races have been going on since track and field was started.

track is going to lose fans if the races go slow? give me a break. so i bet ElG's 2004 olympic win in the 1500m and 5000m really lost a lot of fans. I can assure you that race is immensely more popular to watch than one of his timetrialed affairs.

Run37 wrote:

Championship distance races should be won by runners who can run that distance the fastest. To me "racing" is seeing who can get to the finish line in the shortest amount of time.

Couldn't disagree more. It's a "race" not a "time trial," which is what you're describing.

Racing involves tactics, strategy, mind games, etc. Unfortunately an increasing majority of track fans embrace your outlook, and only judge a race by the watch. Since only a tiny fraction of races can produce records of any kind, these fans walk away disappointed in the "slow" times, and don't come back. IMO, the emphasis on times is what's killed a lot of interest.

Your idea would be like basketball outlawing any kind of defense - "the winning team should be the one who can shoot the most (uncontested) baskets," or baseball only allowing batting practice pitchers in games, "the team that hits the most home runs deserves to win." It's the nuances of a sport that make it interesting - it's too bad more people can't appreciate them.

Hey, Lets get together and celebrate a rising star in mens distance running. Galen Rupp and his coache\'s acocmplishments are nothing short of amazing. This is a wonderful young man who has done nothing but run fast and been loyal to his school and friends. I think you all need to get a life.

someone out here wrote:

I am still convinced that Ivy league schools aren't harder than any others, once you get in.

As for Harvard, Yale and Princeton you'd be wrong. As for the others, well, you're probably right.