Skøyteranking wereldranglijst

I have made a "digital book" about speed skating history, or rather the top skaters through history, all the way from Jaap Eden to Jordan Stolz. You can read more about the book here, and also find my e-mail adress if you want to buy the book.

Anyway, this is my version of the Dutch profile from 1893 til 2023. This figure shows how many Dutch skaters have been ranked among the top 40 in the world each year.

I have been following your tweets about the list and I have some serious questions about the validity of your methodology.

For example: You have Jorrit Bergsma ranked at 77 while Keiji Shirahata is 69th.

Shirahata has won;
World Championships (distance + allround): 3 silver, 2 bronze.
World Cup 5k/10k (overall): 1 silver, 1 bronze.
Olympic Games: 0 medals.

Bergsma has won (not counting mass start or team pursuit as Shirahata couldn't skate those):
World Championships (distance + allround): 3 gold, 8 silver.
World Cup 5k/10k (overall): 4 gold, 1 silver, 1 bronze.
Olympic Games: 1 gold, 1 silver, 1 bronze.

In total that's 4 silver and 3 bronze for Shirahata against 8 gold, 10 silver, and 2 bronze for Bergsma.
In addition to that Bergsma has been good for a longer time as he medaled at the world stage in a total of 9 seasons, while Shirahata only medaled in 4 seasons.
Lastly, Bergsma has 12 gold, 14 silver, & 6 bronze at world cup races on a classic distance while Shirahata has only 3 gold, 5 silver, & 9 bronze (which I'm not counting in this analysis, but still wanted to mention it).

I know why Shirahata is higher on your list, as he skated decent 1500 meters while Bergsma didn't. But ask yourself: Do the four top 10 places (6th, 6th, & 8th in the overall world cup and 7th at the WSD) that Shirahata skated on the 1500 really weigh up against the 8 golds and 6 silvers that Bergsma has more than him? Of course not, that's ridiculous.

No reasonable person would argue that Shirahata was a better skater than Bergsma, so if your methodology says that then you need to inspect your methodology to see why it produces an obviously incorrect result.

The only way Shirahata can come higher than Bergsma, is when you mainly count allround skating. But even when that's the case, it's still illogical that Shirahata is higher than Hedrick.

(Voor de duidelijkheid: Skøyteranking plaatst sinds een aantal dagen tweets over de top-100 van zijn all-time lijst.)

Thanks for the question, although it seems a bit strange to do this debate here and not on Twitter?

First of all: If you buy the book (please do!), you can see that Shirahata and Bergsma are actually very close on my all time list: Shirahata with 328 all time points, and Bergsma with 316.

The tables below show the rankings per year. I have transformed those into "all time points": Number 1 gets 100 pts, number 2 gets 80 points, etc. down to number 25 who gets 1 point.

Shirahata's top years (2-3-4-5-10) are somewhat better than Bergsma's (2-4-6-9-10), which gives him a better score on my all time liste. But Bergsma seems to be around even for the coming season, so his standing can improve.

...and here comes Hedrick, also with 316 all time points.

The reason I responded here is that I questioned your lists on Twitter before and never really got an answer except your explaining your methodology, which is not what I'm looking for. Also, 280 characters just isn't enough. Lastly, since this costs money to get people on here have a right to know what they're buying. I won't be buying it as I have no faith in your lists.

Here you've once again done the same. I can see the numbers that lead to Shirahata getting a higher score than Bergsma, but as I said that's not the problem. The problem is that you accept this outcome as fact as if your methodology is perfect. But if you get an outcome this wrong (you won't convince you actually believe that Shirahata was even close to how good Bergsma was/is) then you need to change your formula.

Another massive error that comes to mind in march 2023, after the WSD, you ranked Joy Beune above Ragne Wiklund on the 5k (https://www.skoyteranking.net/mar2023-kvinner) as you still counted a 7:31 Wiklund skated in Stavanger in 2017 as a 17-year old. I understand that on the longest distance data can be sparse, but this was just ridiculous. Wiklund has skated a 6:46, 6:52, 6:56, 6:56, 6:56, & 6:57 in the past two seasons (all her sub 7:00 times), got a silver medal at the WSD and 5th place on the Olympics. Beune only skated a 6:58 & 6:59 under 7 minutes and didn't even participate in the Olympics or WSD.

Could be worse, as Arianna Pruisscher, who skated only four 5k's in her lifetime (two in 2016 as a junior and two in 2022) and never under 7:04, is somehow ranked 7th while Beune and Wiklund are ranked 10th and 11th respectively.

I'm not interested in why exactly these rankings roll out of your formula. No, I want to know why you think your methodology creates an accurate list while it has such weird rankings.

(Sorry voor al het Engels op het forum maar het is belangrijk dat dit goed overkomt).

Ik heb er zeker wel respect voor dat het erg moeilijk is om zo'n lijst te maken, zeker op de lange afstand. Mijn probleem is nu dat er geld voor gevraagd wordt, en dan mag de lijst best wat kritischer bekeken worden.

Het is ook hoe vreemd de uitkomt is. Of je nou Romme boven Kramer hebt op de 10k of andersom maakt niet zoveel uit, voor beide kun je goeie argumenten maken. En sommige mensen waarderen goed zijn op meerdere afstanden meer dan domineren op één afstand en anderen andersom. Daar kun je leuke discussies over hebben. Maar op het moment dat Pruisscher boven Wiklund komt op de 5 kilometer heb je geen been om op de staan.

Thanks again for your response! As already mentioned: There is no such thing as a perfect ranking system. (I have been working on this for almost 13 years now, so I should know...!) The long distances are particularly difficult, as there are few races, but also because there are few skaters from each country in the most important competitions. For example: Would it be fair to say that Romme was not among the top 32 5000 m skaters in 2002 - just because he was not at his best in the qualification race.?

As from the 2024 season, I am making an adjustment in the ranking formula for the 5000 m L and 10 000 m M, so I hope this will make those rankings a bit more relevant in the future. But the main focus have always been the total top 40 rankings, and those are also the main feature of the digital book I was talking about. (See the link for a detailed list of contents).

About the price of 10 euros: As I have been working a lot of hours to make the book, it seems fair that it should have a price also. But please note that all sales income goes to a good cause: namely the 2024 Norwegian junior champions!

Yes, so the 2017 ranking does not necessarily equal "who did best in 2017". But Bergsma's problem is that he only have those two distances to lean on. Still, he is quite close to the top:

Ja dat is bij de overall lijsten het grootste probleem. De 45e plek van Beckert op de 1500 meter zou nooit genoeg moeten zijn om met een 8 & 5 op de lange afstanden boven iemand met 2 & 2 op de lange afstanden te komen, maar dat soort dingen zie je constant terug hier.

Actually, the biggest problem here and now seems to be commenting without being fully familiar with how the point system works.

Ik heb gelezen op je website hoe het werkt met de ELO punten, en het is me niet duidelijk waarom Beckert boven Bergsma zou komen. Ze krijgen allebei de bonus voor op meerdere afstanden goed zijn, maar ik snap niet hoe de 1500 meter van Beckert genoeg punten verzamelt om Bergsma voorbij te gaan.

"Skøyteranking wereldranglijst" of zoiets. Goed idee verder ja.

Beckert-Bergsma: For all seasons but 2023, I have added socalled allround and sprint points, rewarding skaters who take part in the major samalog events. This means that Patrick Beckert (and also several other skaters) had an ELO-ranking in allround that doesn't show in the table above. So it is not primarily his 1500 m ranking that makes the difference compared to Bergsma.

As from the 2023 season, I had to admit that the samalog championships are getting so few that they should not be used in my rankings anymore. But I still have a bonus given to skaters with good results on more than one distance. And the bonus increases from 2 to 3 distances etc.

I understand that the rankings must have been fun to create , but such a project seems too dependant on the subjective weighting points. You can manipulate the outcome and so on. The skaters themselves had no clue that they were taking part in such competitions.

What would have been more interesting, were if such an "ELO"-concept were introduced to all the skaters knowing with huge prizes at stake. I.e. a World Ranking comparable to ATP in tennis. Difficult to implement of course.

The countdown was finished today, and the "winner" was.... (read yourself).

The digital book mentioned above is still available. Feel free to contact me about it.