Ranking the Ninja Warriors, a statistical approach

[elshizzo]

After catching up on a slew of ninja warrior episodes, I decided to see how each of them would rank, so I came up with a statistical method to test it. I took the historical info for all the main competitors [from sasukepedia.wikia.com] and assigned numbers to each performance. Every time a competitor lost in the first round, they got 0 pts, losing in the 2nd round gets you 1 pt, 3rd round gets you 2 pts, losing in the final gets you 3 pts, and winning the whole thing gets you 4 pts. Obviously this isn't a perfect method, since the courses have different difficulty levels, but anyways, here's how it turned out.....

Total Points
1 Takeda Toshihiro 29
2 Nagano Makoto 28
3 Yamamoto Shingo 27
4 Yamada Katsumi 15
6 Iketani Naoki 14
5 Shiratori Bunpei 14
7 Takahashi Kenji 13
8 Li En Zhi 12
9 Okuyama Yoshiyuki 12
10 Akiyama Kazuhiko 11
11 Asaoka Hiroyuki 11
12 Nakata Daisuke 11
13 Nagasaki Shunsuke 10
14 Kane Kosugi 10
15 Jordan Jovtchev 9
16 Urushihara Yuuji 9
17 Yamada Kōji 8
18 Hashimoto Kōji 7
19 Kobayashi Masaaki 7
20 Kanno Hitoshi 6
21 Brian Orosco 6
22 Levi Meeuwenberg 6
23 Kobayashi Shinji 5
24 Shane Kosugi 5
25 Washimi Yūji 3
26 Paul Anthony Terek 2
27 Miyazaki Daisuke 2

Surprisingly [to me atleast] Nagano is not the winningest Ninja warrior ever, he's a very close 2nd place to Takeda [mainly because Takeda is more consistent]. Yamamoto is 3rd mainly because he's participated in the most contests.

To fix this imbalance caused by having an edge for being in more contests, I decided to take an average pt total per contest entered in

Avg Pts per Contest
1 Kane Kosugi 2
2 Nagasaki Shunsuke 1.67
3 Okuyama Yoshiyuki 1.5
4 Li En Zhi 1.5
5 Urushihara Yuuji 1.5
6 Kobayashi Masaaki 1.4
7 Nagano Makoto 1.4
8 Takeda Toshihiro 1.38
9 Jordan Jovtchev 1.29
10 Shiratori Bunpei 1.27
11 Nakata Daisuke 1.22
12 Levi Meeuwenberg 1.2
13 Brian Orosco 1.2
14 Hashimoto Kōji 1.17
15 Yamada Kōji 1.14
16 Asaoka Hiroyuki 1.1
17 Yamamoto Shingo 1.04
18 Kanno Hitoshi 1
19 Takahashi Kenji 1
20 Iketani Naoki 0.82
21 Shane Kosugi 0.71
22 Yamada Katsumi 0.63
23 Akiyama Kazuhiko 0.58
24 Paul Anthony Terek 0.5
25 Miyazaki Daisuke 0.5
26 Washimi Yūji 0.5
27 Kobayashi Shinji 0.5

Kane Kosugi ended up way ahead, which surprised me, except not that much since he participated in the early contests which were easier. In the 5 times he entered, he made the atleast the 3rd stage in all but one. Nagasaki also surprised me being 2nd. He competed in later contests which might make it even more impressive. He made atleast the 3rd stage in 4 of 6 contests. [15-18].

In order to take a composite list, a sort of final list, I multiplied the total pts by the average pts. Here is the ranking of the greatest Ninja Warriors ever [atleast by these metrics, not perfect i'll admit]

Composite Ranking
1 Takeda Toshihiro 40.05
2 Nagano Makoto 39.2
3 Yamamoto Shingo 28.04
4 Kane Kosugi 20
5 Li En Zhi 18
6 Okuyama Yoshiyuki 18
7 Shiratori Bunpei 17.82
8 Nagasaki Shunsuke 16.67
9 Urushihara Yuuji 13.5
10 Nakata Daisuke 13.44
11 Takahashi Kenji 13
12 Asaoka Hiroyuki 12.1
13 Jordan Jovtchev 11.57
14 Iketani Naoki 11.53
15 Kobayashi Masaaki 9.8
16 Yamada Katsumi 9.38
17 Yamada Kōji 9.14
18 Hashimoto Kōji 8.17
19 Brian Orosco 7.2
20 Levi Meeuwenberg 7.2
21 Akiyama Kazuhiko 6.37
22 Kanno Hitoshi 6
23 Shane Kosugi 3.57
24 Kobayashi Shinji 2.5
25 Washimi Yūji 1.5
26 Paul Anthony Terek 1
27 Miyazaki Daisuke 1

Also, let me know if I forgot anyone. I only considered contestants which have made the 3rd round atleast once and have been in atleast 4 contests.

[roy]

I believe a system in which includes older tournaments is flawed. It favors people(Far too heavily) who have competed for longer. Old results just shouldn't be relevant, if you want present rankings. That's not to diss on guys who had great results in the past, but can you really rank someone who hasn't competed in the last 7 or so tournaments?

For instance, who would you say is the stronger competitor: shingo or yuuji?

For this reason, I think era rankings(shin sasuke, etc) are better suited.

[elshizzo]

I believe a system in which includes older tournaments is flawed. It favors people(Far too heavily) who have competed for longer. Old results just shouldn't be relevant, if you want present rankings. That's not to diss on guys who had great results in the past, but can you really rank someone who hasn't competed in the last 7 or so tournaments?

For instance, who would you say is the stronger competitor: shingo or yuuji?

For this reason, I think era rankings(shin sasuke, etc) are better suited.

I gotta sleep but i'll take a look at different methods, maybe putting more weight on competitions that are more recent, maybe splitting it into sections

[supersheep]

Theres a formula somewhere on the board which has a breakdown for every tournament.

[scnoi1217]

Me and Yam came up with a system also, find that and the results in this thread:

sasukemaniac.proboards.com/index.cgi?action=display&board=talk120&thread=2581&page=3#62334

[elshizzo]

Me and Yam came up with a system also, find that and the results in this thread:
sasukemaniac.proboards.com/index.....81&page=3#62334

Wow, you guys went a lot deeper than I did [going obstacle by obstacle instead of by round]. I'm too lazy to go that deep.

I did notice that Yam ranked the courses by difficulty. To me, this seems arbitrary, since course difficulty is a matter of someone's opinion. I don't want my opinion to sway the results.

I think I will rerun it with a simple modifier which biases in a straight line for courses which are more recent [because I think it had gotten more difficult year after year]

[elshizzo]

Ok, I just reran my sheet with adjustments. To make a fair course difficult adjustment, I used the formula [difficulty modifier = sasuke#^0.3]. This makes sasuke 1 have a difficulty of 1, sasuke 26 a difficulty of 2.66 [and sasukes in between are linearly from 1 - 2.66]. Then, I took the average course difficulty for each competitor, and those are as follows...

Average Course Difficulty
Urushihara Yuuji 2.58
Hashimoto Kōji 2.58
Kanno Hitoshi 2.57
Brian Orosco 2.56
Okuyama Yoshiyuki 2.54
Miyazaki Daisuke 2.53
Levi Meeuwenberg 2.53
Li En Zhi 2.52
Paul Anthony Terek 2.47
Washimi Yūji 2.45
Takahashi Kenji 2.32
Nagasaki Shunsuke 2.32
Nagano Makoto 2.29
Kobayashi Shinji 2.28
Yamada Kōji 2.25
Jordan Jovtchev 2.25
Shiratori Bunpei 2.24
Takeda Toshihiro 2.21
Kobayashi Masaaki 2.2
Nakata Daisuke 2.14
Iketani Naoki 2.12
Yamamoto Shingo 2.08
Yamada Katsumi 2.06
Akiyama Kazuhiko 2.06
Asaoka Hiroyuki 1.86
Shane Kosugi 1.63
Kane Kosugi 1.58

Then, I took that number and multiplied it by the number in my original total points sheet, making it Total Points adjusted by course difficulty.

Total Points adjusted by course difficulty.
1 Takeda Toshihiro 64.07
2 Nagano Makoto 64.01
3 Yamamoto Shingo 56.27
4 Shiratori Bunpei 31.36
5 Yamada Katsumi 30.88
6 Okuyama Yoshiyuki 30.5
7 Li En Zhi 30.27
8 Takahashi Kenji 30.14
9 Iketani Naoki 29.62
10 Nakata Daisuke 23.52
11 Urushihara Yuuji 23.19
12 Nagasaki Shunsuke 23.16
13 Akiyama Kazuhiko 22.61
14 Asaoka Hiroyuki 20.51
15 Jordan Jovtchev 20.25
16 Hashimoto Kōji 18.04
17 Yamada Kōji 18.04
18 Kane Kosugi 15.77
19 Kobayashi Masaaki 15.43
20 Kanno Hitoshi 15.42
21 Brian Orosco 15.35
22 Levi Meeuwenberg 15.2
23 Kobayashi Shinji 11.41
24 Shane Kosugi 8.17
25 Washimi Yūji 7.34
26 Miyazaki Daisuke 5.07
27 Paul Anthony Terek 4.94

Then, i'll take that number and multiply it by the avg pts per course, to get a new composite ranking of greatest ninja warriors! I am satisfied with the list, as Nagano surpasses Takeda this time, and Kane Kosugi is further down. Also, I think Okuyama, Li, and Bunpei have earned a top spot.

Composite Ranking modified by difficulty
1 Nagano Makoto 89.61
2 Takeda Toshihiro 88.48
3 Yamamoto Shingo 58.44
4 Okuyama Yoshiyuki 45.76
5 Li En Zhi 45.41
6 Shiratori Bunpei 39.92
7 Nagasaki Shunsuke 38.6
8 Urushihara Yuuji 34.79
9 Kane Kosugi 31.55
10 Takahashi Kenji 30.14
11 Nakata Daisuke 28.75
12 Jordan Jovtchev 26.03
13 Iketani Naoki 24.39
14 Asaoka Hiroyuki 22.56
15 Kobayashi Masaaki 21.61
16 Hashimoto Kōji 21.04
17 Yamada Kōji 20.61
18 Yamada Katsumi 19.3
19 Brian Orosco 18.42
20 Levi Meeuwenberg 18.24
21 Kanno Hitoshi 15.42
22 Akiyama Kazuhiko 13.09
23 Shane Kosugi 5.84
24 Kobayashi Shinji 5.71
25 Washimi Yūji 3.67
26 Miyazaki Daisuke 2.53
27 Paul Anthony Terek 2.47

[elshizzo]

Also, if there's any interest, here is the Average number of pts per course adjusted by course difficulty. [minimum 4 courses entered]

1 Urushihara Yuuji 3.87
2 Nagasaki Shunsuke 3.86
3 Okuyama Yoshiyuki 3.81
4 Li En Zhi 3.78
5 Nagano Makoto 3.2
6 Kane Kosugi 3.15
7 Kobayashi Masaaki 3.09
8 Brian Orosco 3.07
9 Takeda Toshihiro 3.05
10 Levi Meeuwenberg 3.04
11 Hashimoto Kōji 3.01
12 Jordan Jovtchev 2.89
13 Shiratori Bunpei 2.85
14 Nakata Daisuke 2.61
15 Yamada Kōji 2.58
16 Kanno Hitoshi 2.57
17 Takahashi Kenji 2.32
18 Yamamoto Shingo 2.16
19 Asaoka Hiroyuki 2.05
20 Iketani Naoki 1.74
21 Yamada Katsumi 1.29
22 Miyazaki Daisuke 1.27
23 Paul Anthony Terek 1.24
24 Washimi Yūji 1.22
25 Akiyama Kazuhiko 1.19
26 Shane Kosugi 1.17
27 Kobayashi Shinji 1.14

[supersheep]

I have computer simulations which can to an extent accurately the difficulty of each course. Of course the human element is still there so the results are variable, but over time the statistics catch up.

[YoDaUO]

I still think for this "ranking thing" each obstacle should have a certain # of points.

EX :

Quintuple Step : 1
Jumping Spider : 3
UCH : 10

[roy]

''If you'd like to try it per obstacle- lets say Nagano gets 125 pts(arbitrary) for reaching the G-Rope in 23, but since he did not reach the final in 25, he loses those points. However, he did make it to the jumping spider in 24(lets just say 4 pts there-) and the circle slider in 25(-and 7 here). So he loses the 100, since two tournaments have passed since his finals, and then picks up 11 points for his first stage performances. So then lets say he makes the 3rd stage in 26, fails the UCH(lets just say this is 65 pts), then his 4 pts from S24 falls off, and he has a new point total of 72(Circle Slider reach-7+UCH reach-65).

According to a system like this, if the points are worked out correctly, then Yuji in all-likelyhood is ranked 1, while guys like Kenji Takahashi and Li En Zhi could be top 5. Takeda would be up there too, with a deep third stage run in 24, and 2nd stage salmon ladder reach in 25.''

Here's a system I built a few months ago. Obviously my prediction about nagano in 26 is wrong, but you get the point XD

it's focused on a per obstacle/bi-annual basis.

Oh and if you're going to try to keep that formula, but use it for present(not all time) ranking, why not drop people from the final list who haven't competed in x number of tournaments?

[Badalight]

From the formula in the first post, wouldn't Akira Omori be at the top despite only getting to the final stage in the early Sasukes? I don't see how that can compare to people like Koji or Lee considering how horrible he did when he returned.