There is a British runner named Jessica Andrews and last Saturday she ran an eighty three second PB to become the surprise winner of the British Olympic 10,000m trials and qualify for Rio later this summer…
…But what if her name was Džéssika Andreyeva?
Or Dzhes Arzamasava?
Would we feel differently about an 83 second improvement over 10,000m if we had heard that it happened in the Luzhniki Stadium?
At the Russian track and field Olympic trials?
By an athlete that was not even expected to make the Russian Olympic team?
There would be a strong level of suspicion attributed to such an improvement if just one aspect of this performance was changed – the athlete’s nationality.
This is not intended to accuse or offered as proof of doping in a specific athlete, but rather as a way to highlight the nationalistic double-standards that, not for the first time, influence the anti-doping conversation. We know, thanks to the investigative journalists who have exposed it, that Russian sport is rotten with doping, at a state level, and this fits our narrative and preconceptions of doping.
However, in the aftermath of the performance on Saturday night, questions were (rightly, given the current climate of mistrust) asked of Andrews, and I encountered a great deal of obfuscation and misinformation, which only served to make me more committed to asking questions. So I’m writing this blog to get to the bottom of the following questions: How big was Andrews’ improvement really, and are some of the justifications and explanations for her improvement valid?
The point is that when performance is offered as proof or reasons for suspicion of Russians, as it often is, then the very same standard should be applied to any athlete. I want to look at that improvement, according to actual data, not self reported tales of time away.
So where do I start? Firstly, Jessica Andrews was not expected to feature in the Highgate Harriers Night of the 10,000m PBs, let alone win, and this is evidenced by the fact that she was not even mentioned on the England Athletics website’s list of athletes to watch out for in the race:
“The women’s race will be headlined by European 10,000m champion Jo Pavey (Coach: Gavin Pavey; Club: Exeter) as the 42 year old bids to qualify for her fifth Olympic Games. Pavey will face a strong test however, as Kate Avery (Coach: Tony Simmons, Club: Shildon) and Beth Potter (Coach: Mick Woods, Club: Shaftesbury Barnet) have already achieved the qualifying mark for the Olympic Games with times of 31:41.44 and 32:03.45.
Further names to look out for will be Rhona Auckland (Coach: Joyce Hogg, Club: Banchory Stonehaven), Jessica Coulson (Mick Woods; Stockport) and Lily Partridge (Coach: self, Club: Aldershot Farnham & District), who are all within sight of 32:15, and Charlotte Arter (Coach: James Thie, Club: Cardiff), who has been in excellent form over the winter.”
Yet her improvement in this race was so great that she beat all of the aforementioned athletes, won the race and finished 17 seconds inside the qualifying time for Rio.
I have seen it argued repeatedly on Twitter that this 83 second improvement should not be considered suspicious as the time (31:58.00) is not overly impressive in the grand scheme of things, e.g. at an Olympic final. However I think that looking at Andrews’ time in the context of what other athletes can run is to entirely miss the point, for two reasons. First, because the important point here is the relative personal improvement, and second, because of the recent revelations which have exposed state sponsored doping in Russia and bribes to cover up positive tests in Kenya. These revelations, along with the ever growing list of countries that are non-compliant with WADA code mean that we cannot safely state what a suspicious or non-suspicious time would be. In fact, whether a time should be considered suspicious or not has nothing to do with the time itself and everything to with the ability, genetics, physiology and improvement of the athlete involved.
Therefore we need to look at Jessica Andrews’ time within the context of her own progression. Athletics is mainly an individual pursuit, and so it is much easier to judge an athlete’s improvement than it would be in football/ rugby/ basketball etc. Jessica Andrews’ drop from 33:21.53 in April 2016 to 31:58.00 in May 2016 indicates an improvement of 4.14%. The 10,000m time is significantly stronger, comparatively, than her outdoor 3,000m or 5,000m PBs (9:38.69 and 16:19.66 respectively). Andrews actually ran faster than both her 3,000m and 5,000m PBs during her 10,000m race on May 21st.
She has raced every year for the past eight years, which is inconsistent with her reported complete break (reported by her best friend to me on twitter as “pretty much taking a year out”) from track running. Power of 10 lists performances from her in late November 2014, April, June and July of 2015 and early January 2016. Therefore it cannot be said that she took a long break, no longer than a bad injury layoff. It is true however that she has only ever run two 10,000m track races so I am also going to look at her time in comparison to her results over other distances on the track, which she has raced more frequently.
If she had run her 3,000m track PB (set in 2014) for the entire 10,000m race she would have run 32:06, and,
If she had run her 5,000m (also set in 2014) track PB for the entire 10,000m race she would have run 32:38.
Therefore she ran faster than her 3,000m and 5,000m track PBs on route to a massive 10,000m PB.
Runner’s World prediction calculator (which can be found here: http://www.runnersworld.co.uk/general/rws-race-time-predictor/1681.html) estimates times of 34:30 and 34:01 for 10,000m using Jessica’s respective 3,000m and 5,000m PBs. If we take Jessica’s fastest ever 3,000m time (a 9:22.69 from an indoor competition in February 2016), that equates to roughly a 33:33 10,000m time. The “Running Ahead” website (https://www.runningahead.com/tools/calculators/race) allows users to insert two race times to predict a 10,000m time. Using her lifetime PBs – the indoor 3,000m and outdoor 5,000m – the estimated 10,000m time is 34:39. Clearly these times are not remotely near the 31:58.00 10,000m time she ran on Saturday. This marks the 31:58.00 out as a complete outlier in the context of her other track races.
10km is a distance that Jessica Andrews has raced much more frequently on the road, with her fastest time being 33:33 in Alcaniz in April 2016 (she has also recorded marks of 33:51 in 2015, and 33:55 and 34:17 in 2014). I cannot find any scientific studies comparing road to track times (some athletes never run faster on the track than road, others are faster on the track, I acknowledge this), but the IAAF offers tables (found here: http://www.iaaf.org/about-iaaf/documents/technical) for converting road performances to track performances which are helpful in this context. According to their scoring tables, her best road 10km time of 33:33 (1077 points) is the equivalent of a 33:12 over 10,000m on the track. To look at it from the other perspective, her 31:58.00 is worth 1141 points, equivalent to a 32:20 on the road.
There are other more generous conversions, including:
- 15-20 seconds slower on road for 5km and 40-50 seconds slower for 10km
- 3% slower for all distances on road compared to track
- 3.3% slower for all distances on road compared to track
Using these calculations Andrews’ time equates to 32:43 (using method 1), 32:33 (using method 2) and 32:26 (using method 3).
These road to track conversions bring us closer to the 31:58.00 that Andrews ran, which makes her performance improvement smaller (a converted 33:12 to 31:58 represents a 3.7% improvement, for instance) However it is still a major leap in performance in a very short space of time – especially for someone who has not shown much track pedigree in the past.
We don’t have to look too far into athletics history to find a similarly impressive performance in London. Russian runners have not been allowed to compete in recent months due to the well publicised revelations about state sponsored doping so instead let’s look at Volha Mazuronak (BLR), who ran 2:23:54 during this year’s London Marathon. This performance sparked debate from many sources, including Mara Yamamuchi who labelled the run “astonishing” and noted that she was “unable to understand her performance.” She pointed out that she “wasn’t accusing her of anything” but asked “If anyone with a good knowledge of women’s marathon running can make sense of this performance please let me know.” Mazuronak improved from 2:25:36 in London 2015 to 2:23:54 in the same marathon in 2016. This equates to a 1.17% improvement (much less than Andrews’ 4.14%, or the 3.7% improvement if you use the IAAF’s road conversion formula for Andrews’ best road 10km performance), yet Mazuronak was greeted with much more suspicion.
Also, many people were critical of the fact that Mazuronak broke her half marathon PB during the London marathon, pointing out how implausible it is to set a PB for a shorter distance race on route to a PB in the longer race. Which is identical to Andrews, yet similar complaints were not made when she broke her outdoor 3,000m and 5,000m PBs during her 31:58.00 run.
Mazuronak is thus guilty by nationality first, performance second. Or is it the other way around? Perhaps, most accurately, her performance is judged differently because of her nationality. That’s called profiling. Might this be fair given the ‘form’ shown by eastern Bloc nations? Perhaps. But it’s certainly not fair to imply that Athlete A’s performance is unbelievable, when Athlete B’s, even more profound an improvement, is simply dismissed because of her nationality and (untrue) explanations about time away from running.
I cannot say Jessica Andrews is doping. Similarly, I cannot say that she is clean. In the same way that I cannot speak with certainty either way about Volha Mazuronak. My point is simply that we cannot give certain athletes a free pass based on the country of their birth. Carl Sagan said that “extraordinary claims require extraordinary proof” and I feel that the same should be demanded of extraordinary performances, no matter who produces them.
I don’t think an assertion that one is now “in love with running” should be blindly accepted as the reason for between a 3.7% and 4.14% performance increase – the figures garnered from comparisons to her other track performances over a variety of distances and to her adjusted road performances. If we want clean sport we have to open our eyes, use rational and critical thought, engage our mouths and demand it – the time has long passed for believing everything we see without question or giving anyone the benefit of the doubt because of the nationality listed on their passport.