“Pythagoras Winning Rate” to predict the winning rate by scoring points
To all who opened this article, welcome to the swamp of numbers. I, Shintaro, a self-proclaimed “amateur who takes the best stats in Japan,” would like to introduce and explain “advanced stats,” which is a step beyond the box score (basic stats) on the official website.
The stats to be explained this time are “Pythagoras winning rate”. I think it’s a stat that is not very familiar to basketball fans, but that’s right. This statistic originated in baseball (major league), and was created by Mr. George William James (commonly known as Bill James), who is also the creator of the term “sabermetrics (analysis method for analyzing baseball from data)”. A formula (and its result) that predicts a team’s winning percentage by
I think there are many people who think, “It has nothing to do with basketball,” but please wait a moment. In 1992, it was Daryl Morley (the president of the basketball operations department of the 76ers) who improved this number for basketball and proposed it in 1992. If you could sum up Morley in one word, he’s the one who statistically eliminated the midrange shooter from modern basketball.
Morley’s Pythagoras winning percentage is as follows:
Astute readers may have noticed that the NBA and the B League have very different scores, goals conceded, and the number of games played, so applying this as-is would result in a large difference. Therefore, this time, I would like to calculate using a coefficient of 10 that I have calculated independently based on past data.
Regarding the credibility of this coefficient, first of all, let’s take a look at the winning percentage forecast for the 21-22 season published on the website I manage. Although there was an error in the winning percentage, I was able to hit the CS participating team.
| 21-22 | Predicted standings at the end of 30 games | Expected win rate after 30 games | Actual standings at the end of the season | Actual win percentage at the end of the season | CS appearance |
| 1 person | Nagoya | 81.50% | Ryukyu | 87.50% | ☆ |
| 2 digits | Utsunomiya | 78.50% | Chiba | 77.80% | ☆ |
| 3 digits | Tokyo | 78.40% | Kawasaki | 76.40% | ☆ |
| 4 bit | Ryukyu | 77.30% | Tokyo | 73.60% | ☆ |
| 5 digits | Chiba | 76.90% | Shimane | 72.70% | ☆ |
| 6 digits | Kawasaki | 71.80% | Utsunomiya | 71.40% | ☆ |
| 7 digits | Shimane | 70.30% | Nagoya | 69.40% | ☆ |
| 8 bits | Akita | 65.40% | Akita | 57.40% | ☆ |
| 9 bits | Mikawa | 60.20% | Mikawa | 56.60% | |
| 10 digits | Hiroshima | 54.10% | Shibuya | 55.90% | |
| 11 bits | Shinshu | 52.20% | Shinshu | 51.90% | |
| 12 bits | Shibuya | 48.90% | Hiroshima | 50.90% | |
| 13 bits | Osaka | 43.20% | Gunma | 45.50% | |
| 14 bits | Hokkaido | 37.00% | Toyama | 40.70% | |
| 15 digits | Gunma | 36.30% | Yokohama | 38.60% | |
| 16 bits | Toyama | 33.10% | Hokkaido | 37.50% | |
| 17 bits | Yokohama | 32.80% | Osaka | 36.80% | |
| 18 bits | Shiga | 31.70% | Ibaraki | 29.60% | |
| 19 bits | Ibaraki | 20.00% | Kyoto | 24.60% | |
| 20 bits | Sanen | 20.00% | Shiga | 24.60% | |
| 21 bits | Niigata | 17.70% | Sanen | 17.20% | |
| 22 bits | Kyoto | 14.10% | Niigata | 13.50% |
The preamble has become long, but let’s predict the winning percentage for this season.
| 22-23 | After 30 games | After 30 games | CS appearance |
| Predicted ranking | expected win rate | Predict | |
| 1 person | Nagoya | 82.40% | ☆ |
| 2 digits | Chiba | 82.20% | ☆ |
| 3 digits | Tokyo | 75.90% | ☆ |
| 4 bit | Ryukyu | 74.00% | ☆ |
| 5 digits | Hiroshima | 73.60% | ☆ |
| 6 digits | Shimane | 72.30% | ☆ |
| 7 digits | Yokohama | 63.10% | ☆ |
| 8 bits | Shinshu | 61.50% | ☆ |
| 9 bits | Kawasaki | 59.10% | |
| 10 digits | Akita | 56.40% | |
| 11 bits | FE Nagoya | 55.30% | |
| 12 bits | Gunma | 52.60% | |
| 13 bits | Sanen | 47.60% | |
| 14 bits | Osaka | 44.50% | |
| 15 digits | Utsunomiya | 41.60% | |
| 16 bits | Kyoto | 40.50% | |
| 17 bits | Shibuya | 40.30% | |
| 18 bits | Ibaraki | 37.30% | |
| 19 bits | Mikawa | 36.70% | |
| 20 bits | Sendai | 32.50% | |
| 21 bits | Hokkaido | 27.30% | |
| 22 bits | Toyama | 26.80% | |
| 23 bits | Shiga | 14.90% | |
| 24 bits | Niigata | 10.90% |
As of January 19, Hiroshima, the 1st place in the West, and Shimane, the 2nd place, are wildcards. It was expected that Shinshu would participate in the second place in the middle district instead of the first place in the middle district.
Since this figure is calculated only by points scored and lost, in extreme terms, a team that loses in a close match and wins by a large margin tends to have a higher figure regardless of the actual win or loss. For that reason, Hiroshima has the most wins within 3 points in 5 games, which is the most in the league, and Shinshu has continued to win by double-digit points in the last 10 games.
Injuries to the roster are a big factor in the error (as is the case in the 21-22 season). The number of players who are registered on the injured list is increasing, so I would appreciate it if you could see it on “Tararaba”.
