This text initially appeared within the 2024 Olympic Preview version of SwimSwam Journal, courtesy of creator Daniel Takata. Subscribe to the SwimSwam Journal right here.
This summer time, the Olympic Video games will happen in Paris. And, as typical, there may be loads of expectation for excellent performances, legendary races and, after all, world information. Many will attempt to guess who the winners and medalists of the swimming occasions will probably be. As at all times, some favorites will verify favoritism and a few underdogs will seem. Clearly, it is extremely troublesome to foretell who they are going to be.
However we are able to use quantitative strategies to make one other kind of prediction. Primarily based on latest outcomes, what are the estimated occasions wanted to win the races in Paris? And which information are most certainly to be damaged?
It doesn’t take a genius to understand that there are occasions through which the world document is extremely unlikely (for instance, the boys’s 200 freestyle and the ladies’s 200 butterfly) and others through which the document is predicted, given the evolution lately (e.g., ladies’s 400 freestyle, a race that may function the final three world document holders).
However it’s potential to quantify this likelihood utilizing statistical strategies.
Estimating the advance in Olympic years
To get an thought of the occasions wanted to win the gold medal on the Olympic Video games, we have to hold one factor in thoughts. Traditionally, there’s a vital enchancment in athletes’ occasions in Olympic years. This assertion could sound apparent, and it actually appears to be the impression that many individuals have, however it’s mandatory to investigate the information. And the information exhibits precisely that.
Let’s take a look at the evolution of the boys’s 100 freestyle through the years. The graph under exhibits the typical time of the quickest 100 swimmers in annually since 1990.
The reducing sample is clear, as we’d count on. However we are able to observe one other fascinating facet. Each 4 years, it’s potential to note a dramatic drop within the common time, which isn’t a coincidence because it corresponds to the Olympic years. The one exception is the interval 2008-2009: The occasions lower from 2007 to 2008, and reduce much more from 2008 to 2009, which is uncommon, since 2009 was not an Olympic 12 months. After all, the hi-tech fits had been the trigger.
The identical sample is noticed in all different occasions included within the Olympic program. This must be thought-about when predicting swimming performances for 2024. And, to make this prediction, the statistical methodology considers the distribution of occasions of the 100 quickest swimmers annually, in every occasion.
We use a department of statistics referred to as Excessive Worth Concept (EVT), since we’re modeling the acute performances of athletes (quickest occasions on this planet). We took the occasions of the 100 quickest performers in annually, in every occasion since 1990, and used the EVT to find out its likelihood distribution. There’s a very highly effective mathematical theorem referred to as Generalized Pareto Distribution (GPD) that exhibits that this distribution is at all times the identical.
Analyzing the information
We are able to undertaking the form and the parameters of the distribution for the subsequent few years, contemplating the evolution of the occasions and, after all, contemplating an excellent better evolution in an Olympic 12 months. Now we are able to simulate from this distribution a potential state of affairs for the highest 100 quickest performers in 2024 and take the quickest time — that may be one single prediction for the quickest time on this planet in 2024, which normally corresponds to the Olympic successful time. By simulation, let’s say, ten thousand eventualities, we are able to compute the typical of all quantity ones in every state of affairs. That might be the ultimate estimate for the quickest time on this planet this 12 months.
By doing this, we are able to additionally compute an interval for the quickest time on this planet in 2024, let’s say, with 95% of confidence. Additionally, amongst all eventualities, we are able to depend on what number of of them the world document is damaged, and so we are able to estimate the likelihood of it.
The outcomes are within the following desk.
Ladies’s occasions
| Occasion | Anticipated successful time | Minimal | Most | WR likelihood |
| 50 freestyle | 23.65 | 23.54 | 23.77 | 39% |
| 100 freestyle | 51.75 | 51.56 | 51.94 | 44% |
| 200 freestyle | 1:52.97 | 1:52.47 | 1:53.49 | 42% |
| 400 freestyle | 3:55.70 | 3:55.19 | 3:56.22 | 41% |
| 800 freestyle | 8:06.80 | 8:05.78 | 8:07.84 | 9% |
| 1500 freestyle | 15:24.75 | 15:20.95 | 15:28.57 | 12% |
| 100 butterfly | 55.72 | 55.54 | 55.89 | 19% |
| 200 butterfly | 2:03.31 | 2:02.64 | 2:04.07 | 1% |
| 100 backstroke | 57.31 | 57.06 | 57.57 | 51% |
| 200 backstroke | 2:03.09 | 2:02.30 | 2:03.94 | 48% |
| 100 breaststroke | 1:04.43 | 1:04.25 | 1:04.59 | 12% |
| 200 breaststroke | 2:18.63 | 2:17.60 | 2:19.97 | 11% |
| 200 IM | 2:06.36 | 2:05.99 | 2:06.69 | 35% |
| 400 IM | 4:26.91 | 4:25.18 | 4:28.64 | 33% |
Males’s occasions
| Occasion | Anticipated successful time | Minimal | Most | WR likelihood |
| 50 freestyle | 21.14 | 21.04 | 21.27 | 15% |
| 100 freestyle | 46.86 | 46.66 | 47.05 | 36% |
| 200 freestyle | 1:43.76 | 1:43.40 | 1:44.11 | 1% |
| 400 freestyle | 3:39.73 | 3:38.55 | 3:41.32 | 44% |
| 800 freestyle | 7:35.12 | 7:33.91 | 7:36.25 | 1% |
| 1500 freestyle | 14:31.25 | 14:28.64 | 14:33.60 | 50% |
| 100 butterfly | 50.08 | 49.87 | 50.32 | 5% |
| 200 butterfly | 1:52.65 | 1:52.35 | 1:52.93 | 1% |
| 100 backstroke | 51.84 | 51.65 | 52.04 | 20% |
| 200 backstroke | 1:53.93 | 1:53.70 | 1:54.14 | 1% |
| 100 breaststroke | 57.69 | 57.49 | 57.89 | 1% |
| 200 breaststroke | 2:05.34 | 2:04.90 | 2:05.77 | 59% |
| 200 IM | 1:54.78 | 1:54.34 | 1:55.22 | 11% |
| 400 IM | 4:03.19 | 4:01.98 | 4:04.42 | 34% |
Because the predictions are primarily based on the evolution of swimmers lately, it’s pure that they’re at occasions equivalent to what we have now had in latest seasons. For instance, it has been a couple of years since any swimmer has threatened the ladies’s 100 breaststroke world document – since Lilly King set the world document in 1:04.13 in 2017, the quickest time on this planet has been half of a second slower, so the world document just isn’t anticipated to be damaged in 2024.
Alternatively, final 12 months Ahmed Hafnaoui and Bobby Finke got here very near the boys’s 1500 freestyle world document. Due to this fact, there’s a fairly excessive likelihood that long-standing Solar Yang’s world document from 2012 will lastly be damaged.
The outcomes additionally mirror what we noticed in 2023 when it comes to world information. There have been way more world information set in ladies’s occasions than in males’s occasions. Accordingly, we are able to see that the possibilities of world information being damaged, generally, are better amongst ladies.
It’s fascinating to watch the intervals for the occasions, however will they mirror the true leads to Paris?
