Poupou Escobar
Post Whore
- Messages
- 98,761
A few weeks ago I did some research for a debate that was occurring here, regarding favouritism in NRL finals matches, and I found the data so interesting that I went down the rabbit hole and saved it for its own thread when I finally got around to it. Today is that day.
I used a ten year data sample, because you have to stop somewhere, and ten years is a nice round number that was unlikely to have merkins accusing me of picking a sample to suit preconceived conclusions.
In those ten years (2016 to 2025) there have been 25 coaches overseeing at least one finals appearance. All coaches have a relatively small sample of finals games in ten years, with only one (Bellamy) coaching 20+ finals during the last decade. Twenty of the 25 coached a dozen finals games or less. This all means the data lacks statistical significance from the outset and can't be used to prove most narratives. However, the data is interesting, so like most rugbaleeg analysis, it should largely be used for entertainment purposes.
Firstly, let's get the dumbest of stats out of the way, being a comparison of total games and total wins (i.e. win rate) per coach:
It's tempting to assume this table reflects the ability of a coach to 'win big games' or other nonsense, but as previously stated, the samples are small; 80% of the coaches oversaw half an NRL season's worth of games or less.
So while per-coach stats are relatively meaningless, the data as a whole tells a few stories. Firstly, only 8 of the 25 coaches (less than a third) have a winning record, but this includes 3 who only coached 3 games. If you flipped a coin three times and got heads two or three times, only a f**king idiot would say the coin is rigged.
The data, at least this very simplistic set, also indicates that coaches aren't judged (by the clubs at least) based on their finals record. Seibold, Webster, Ciraldo and Fitzgibbon (all finals losers) are still employed, while Demetriou and Walters (with winning finals records) are not. It's understood by anyone with a basic understanding of probability that the result of a finals match, like any game, has a significant randomness component. There are obviously nonrandom factors, and some are preoccupied with proving the influence of coaching on results. But the biggest factor is the disparity in quality between the teams. This should be evident prior to kickoff, and indeed it is, through the betting odds, which will be the subject of my next post when I get around to it.
Thank you for your time.
I used a ten year data sample, because you have to stop somewhere, and ten years is a nice round number that was unlikely to have merkins accusing me of picking a sample to suit preconceived conclusions.
In those ten years (2016 to 2025) there have been 25 coaches overseeing at least one finals appearance. All coaches have a relatively small sample of finals games in ten years, with only one (Bellamy) coaching 20+ finals during the last decade. Twenty of the 25 coached a dozen finals games or less. This all means the data lacks statistical significance from the outset and can't be used to prove most narratives. However, the data is interesting, so like most rugbaleeg analysis, it should largely be used for entertainment purposes.
Firstly, let's get the dumbest of stats out of the way, being a comparison of total games and total wins (i.e. win rate) per coach:
| Coach | Finals since 2016 | Wins | Win rate |
|---|---|---|---|
| Maguire | 3 | 3 | 100% |
| Cleary | 19 | 16 | 84% |
| Demetriou | 3 | 2 | 67% |
| Walters | 3 | 2 | 67% |
| Bellamy | 27 | 17 | 63% |
| Flanagan | 7 | 4 | 57% |
| Green | 7 | 4 | 57% |
| Robinson | 19 | 10 | 53% |
| Griffin | 4 | 2 | 50% |
| McGregor | 2 | 1 | 50% |
| Payten | 4 | 2 | 50% |
| Bennett | 15 | 7 | 47% |
| Stuart | 14 | 6 | 43% |
| Arthur | 12 | 4 | 33% |
| Fitzgibbon | 9 | 3 | 33% |
| Hasler | 6 | 2 | 33% |
| Seibold | 6 | 2 | 33% |
| Webster | 4 | 1 | 25% |
| Ciraldo | 5 | 1 | 20% |
| O'Brien | 5 | 1 | 20% |
| Barrett | 1 | 0 | 0% |
| Henry | 1 | 0 | 0% |
| Holbrook | 1 | 0 | 0% |
| Kearney | 1 | 0 | 0% |
| Morris | 2 | 0 | 0% |
It's tempting to assume this table reflects the ability of a coach to 'win big games' or other nonsense, but as previously stated, the samples are small; 80% of the coaches oversaw half an NRL season's worth of games or less.
So while per-coach stats are relatively meaningless, the data as a whole tells a few stories. Firstly, only 8 of the 25 coaches (less than a third) have a winning record, but this includes 3 who only coached 3 games. If you flipped a coin three times and got heads two or three times, only a f**king idiot would say the coin is rigged.
The data, at least this very simplistic set, also indicates that coaches aren't judged (by the clubs at least) based on their finals record. Seibold, Webster, Ciraldo and Fitzgibbon (all finals losers) are still employed, while Demetriou and Walters (with winning finals records) are not. It's understood by anyone with a basic understanding of probability that the result of a finals match, like any game, has a significant randomness component. There are obviously nonrandom factors, and some are preoccupied with proving the influence of coaching on results. But the biggest factor is the disparity in quality between the teams. This should be evident prior to kickoff, and indeed it is, through the betting odds, which will be the subject of my next post when I get around to it.
Thank you for your time.
