More unreasonable comparisons in sports

Recently, I have seen two disputable men/women comparisons: Djokovic “overtaking” Graf in weeks-at-number-one (tennis) and Shiffrin “overtaking” Stenmark* in world-cup races won (alpine skiing).

*This at least in Sweden. I have not paid attention to what is said internationally, and the issue has likely been given more weight in Sweden (both Stenmark and I are Swedes). However, I note that several instances of a Shiffrin-win closing the distance, one or two near misses, the original tie, and, today, the actual surpassing of Stenmark’s mark have all been in the news, which seems highly excessive—especially, as it was less a matter of “whether” and more of “when”.

A comparison of this type is of dubious value even when fixed at one sex, and is usually best viewed as a curiosum.* When we make a comparison that includes both sexes, unless the sport or whatnot is not segregated by sex to begin with, we are basically wasting our time, as the respective competition is too different (cf. excursion). For instance, the headline that informed me that Djokovic had overtaken Graf said that he was now “better”** than Graf, but by what standard? Certainly, even a sub-par Djokovic would have outclassed even a prime Graf in a fair competition match, had such a match been arranged by time travel,*** and there can be no doubt that he has been the better in a non-segregated comparison since long before overtaking her on this dubious criterion. On the other hand, if we make a segregated comparison, the currently very small difference in weeks is negligible in light of the many other factors that play in, e.g. that Graf was younger when she retired than the (still active) Djokovic is today/was at the time of overtaking her.

*Cf. some of the many earlier texts dealing with tennis, GOATs, and/or Djokovic, in particular Tennis, numbers, and reasoning: Part II. In tennis, weeks-at-number-one might be an exception to the “curiosum”, as it is a much better heuristic for comparisons than “majors won” or, in alpine skiing, “world-cup races won”; however, it is still just a heuristic, and more holistic comparisons are better—and it has no true value when comparing a male and a female player.

**Or something to a very similar effect. I did not keep references at the time.

***Graf arguably had the best year of her career in 1988, while Djokovic was born in 1987. Had Graf extended her career longer than she did, we might conceivably have had a 30-something Graf playing a teenaged Djokovic.

Comparing Shiffrin and Stenmark borders on the ridiculous, even the sex issue aside: there are roughly four decades between them, the sport has changed considerably (contrast e.g. 1970s skis with modern skis), the number of races per season has not been constant, and Super-G was not even a part of the world cup during Stenmark’s prime (while Shiffrin has won at least some Super-G races; the lack of Super-G might also have increased the competition in the other disciplines).* We are now at a point where a comparison between e.g. Djokovic and Tiger Woods makes as much (little) sense.

*I will leave research of details to those with a large enough interest.

For that matter, while Stenmark long had a hard-to-dispute claim as the (male) GOAT of alpine skiing, chances are that he already has been surpassed by Marcel Hirscher in a more holistic comparison, which pushes the “curiosum” angle further. If we were to make a male–female comparison, would not a comparison between Shiffrin and Hirscher make more sense? (But note the complication caused by their career overlap. Cf. excursion.)

Excursion on issues with inter-sex comparisons:
The most obvious issue is that men and women face different competitors, which, even the greater physical abilities of men in most sports aside, makes the comparison flawed. Firstly, a different set of competitors inherently makes a comparison hard, as with e.g. two male tennis players competing in different eras. Secondly, if the objects of comparison are contemporaries, the comparison can be near impossible.* Consider e.g. the 400m hurdles and the recent successes of respectively Warholm (man, gold medals in multiple recent championships, unbelievable world records) and McLaughlin (woman, gold medals in multiple recent championships, unbelievable world records)—clearly, at least one of them would look far less impressive had they been in direct competition, and their relative stature is contingent on being in separate competitions.** (Ditto, but with less overlap, the aforementioned Shiffrin and Hirscher.)

*Unless they are of the same sex, in which case the comparison becomes more relevant, as they are in direct competition with each other.

**Imagine, similarly, that the world of athletics was split into countries where only the 370m hurdles resp. the 430m hurdles were run. We might now have two men looking unbeatable at the respective distance, at the same time, and with the knowledge that at least one would be diminished if the distances were unified to the 400m hurdles. Also note the old split into amateur and professional tennis and how that might have affected the success of various players. (As briefly discussed in one of the older texts.)

However, other issues exist, e.g. that there might be more men (women) than women (men) competing in a certain sport, which makes the level of competition higher;* that the number of competitions available might differ;** and that the rules might differ,*** making the events less comparable.

*Even had men and women been otherwise interchangeable. To revisit an earlier footnote, say that there were twice as many running the 370m hurdles compared to the 430m hurdles, and note how much harder it would be to be competitive in the former than the latter (all other factors equal).

**For instance, the German soccer Bundesliga for men has 18 teams and 34 games per team, while the women’s has something considerably lower (maybe, 12 and 22). Does it really make sense to compare the number of goals made by some male player with those of some female player? (Note that such issues might have an effect on more reasonable seeming measures, e.g. goals per game, too. For instance, fewer games mean a greater risk that statistical fluctuations are important; for instance, fewer teams in the league can imply that the weakest teams are relatively stronger, which reduces the possibility of scoring cheap goals.)

***For instance, tennis majors are played in best-of-five for men but best-of-three for women; for instance, male and female boxers have different rules for where punches are forbidden.