# Event Detection on NHL dataset

Source: [Event Detection on NHL dataset](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:45:07.197Z

Evaluation of Event Detection performance on the NHL dataset, reporting Precision, Recall, and F1 scores for specific in-game events (Faceoff, Shot, Advance, Play) and overall averages.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Event Detection |
| Dataset (including any stated split) | NHL dataset |
| Metric | Faceoff Precision |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 1 of 1 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Faceoff Precision](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Faceoff%20Precision>): Higher is better; unit: PERCENT.

- [Faceoff Recall](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Faceoff%20Recall>): Higher is better; unit: PERCENT.

- [Faceoff F1](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Faceoff%20F1>): Higher is better; unit: PERCENT.

- [Shot Precision](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Shot%20Precision>): Higher is better; unit: PERCENT.

- [Shot Recall](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Shot%20Recall>): Higher is better; unit: PERCENT.

- [Shot F1](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Shot%20F1>): Higher is better; unit: PERCENT.

- [Advance Precision](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Advance%20Precision>): Higher is better; unit: PERCENT.

- [Advance Recall](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Advance%20Recall>): Higher is better; unit: PERCENT.

- [Advance F1](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Advance%20F1>): Higher is better; unit: PERCENT.

- [Play Precision](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Play%20Precision>): Higher is better; unit: PERCENT.

- [Play Recall](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Play%20Recall>): Higher is better; unit: PERCENT.

- [Play F1](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Play%20F1>): Higher is better; unit: PERCENT.

- [Average Precision](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Average%20Precision>): Higher is better; unit: PERCENT.

- [Average Recall](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Average%20Recall>): Higher is better; unit: PERCENT.

- [Average F1](<https://www.sota2.com/research/sota/event-detection-on-nhl-dataset.md?metric=Average%20F1>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Faceoff Precision | Result date | Report paper |
| --- | --- | --- | --- |
| multi-tower temporal convolutional network | 0.7778 | Not recorded | 68fa675d1e209131cea15e13 |

### multi-tower temporal convolutional network

Faceoff Precision: 0.7778. Result ID: 698d31278b517eb4331fb3fc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa675d1e209131cea15e13.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa675d1e209131cea15e13.md>)

Recorded evaluation configuration: {"Sliding window stride":"1 second","Post processing":"true","Tolerance (delta)":"1 second","Weighted cross entropy loss":"true","Adam optimizer":"true","Initial learning rate":"0.001"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.