Recognizing Events in Spatiotemporal Soccer Data

Round 1

Reviewer 1 Report

I evaluated the previous version of manuscript.

There are many changes in content compared to that time(previous version), and I think it is an interesting topic.

Therefore, I recommend that this manuscript to accept after checking the overall grammar(It would be okay to get professional grammar correction).

Author Response

Thank you for the feedback!

We have asked Dr. Allan Nicholas, a Senior Associate Professor at the Center of Language Research of the University of Aizu, to proofread the manuscript.

Reviewer 2 Report

1. Line 3 in abstract: Sentence beginning with “However, …” Consider rewording. It was difficult to read and understand.
2. Introduction: The phrase “rules-based algorithm” is used in the abstract and introduction. An example of a “rule” early on would help a reader’s early comprehension of the approach, such as the speed and direction of a ball as well as the position of other players relative to the ball contribute in the algorithm to determine whether or not a ball is in a player’s possession, passed, or shot.
3. Equations 1 and 2 and Figure 1: The notation is not defined and there is no labelling or description for the lines on the figure.
4. Listing 1: an explanation in words of the listing would only take a sentence and would speed up a reader’s comprehension.
5. A data-heavy algorithm such as this is prone to overfitting. Overfitting would make predictions on data not included in the sample less reliable, although likely still usable. To evaluate if overfitting has a large effect in this model, please designate some data or possesions as holdout data and then evaluate the trained model on the holdout data and evaluate precision and recall. Ideally, the precision and recall will not fall much, although it should be suspected that it will decrease some.
6. Have all three methods (Richy, Morra, and yours) ever been used on the same data set? If so, would that be worth including in the model? If not, should that happen here in this paper? As it stands, the conclusion mentions that the reasoning for the precision and recall being higher is due to the other papers not focusing on those criteria. If the authors could be more specific about why their approach achieved a higher precision than the others in the conclusion that may be sufficient. Is it simply the data quality or is there something fundamental about the approach that affects those numbers. I believe this information is obtainable through careful reading of section 2 and looking more closely at the other papers, but explicit statements would be helpful.

Author Response

Thank you very much for the valuable feedback!

We have prepared a point-by-point response to each of your suggestions. Please see the attachment.

For the English language editing, we asked our colleague Dr. Allan Nicholas to copyedit the manuscript. He is a native English speaker and a Senior Associate Professor at the Center of Language Research. {We've added changes to the final version.

Author Response File: Author Response.pdf

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The authors propose a rule-based algorithm for identifying several basic types of events in soccer. They provide a simple procedure that can be used for practical soccer data analysis tasks, and also serve as a baseline model for the algorithms based on more advanced approaches. The results are interesting, but should add more various simulations. Therefore, I recommend major revision the present manuscript for publication in Applied Sciences.

Author Response

Thank you for the feedback.

> The results are interesting, but should add more various simulations.

We have used all data available to us in the present simulations. In addition to precision and recall, we have added F1 score. These measurements are consistent with those found in the related literature.

Please see our revisions in the updated manuscript.

Reviewer 2 Report

I agree with the paper. The contribution is writtenw ell. The topic and appliatin into sport is interested and nice. I recommend the publication.

Author Response

Thank you for the feedback.

Reviewer 3 Report

In general terms, the document is well-written. However, it seems that authors did not communicate effectively what their aims were and consequently, what they wanted to achieve. It is not clear, what are the gaps in the knowledge that the authors wanted to fill? Furthermore, hence, what are the contributions of the paper? What is different and new in the authors’ approach? In that regard, the authors should clarify what the research questions were and how their methodological approach is adequate and novel that makes a substantial contribution.

The introduction is feeble and should be extended. An effective introduction has to have the following structure:

• Introduction to the topic
• Main problem
• Possible solution using XXX
• It is new! But several works have been developed in the field... "a brief" literature review
• Advantages
• Work that was done
• what do we try to prove...
• Data used
• Schema used
• Structure of the paper

Results have to be rewritten. Currently, what they wrote is very superficial. It is imperative to present the results in the same order as the methodological steps. You must be sure that datasets, graphics, and tables are properly described and answer to a part to the proposed method. You have to include how your method performs in both training and validation datasets. Moreover, if you claim that your approach is better than others, you have to test those kinds of affirmations using adequate statistical testing. It is not valid, to claim better performance based on plots or only numeric differences.

Authors have to organise their Conclusions newly. In this section, authors have to answer to the research questions and highlight the constraints that arose in this article and indicate what the new open lines that this manuscript pointed it out are.

Author Response

Thank you for the feedback. 

> the authors should clarify what the research questions were and how their methodological approach is adequate and novel that makes a substantial contribution.

We have extended introduction, discussion, and conclusion sections.

> The introduction is feeble and should be extended. 

We have added the structure of the paper to the introduction. 

> if you claim that your approach is better than others, you have to test those kinds of affirmations using adequate statistical testing.

In addition to precision and recall, we have added F1 score. These measurements are consistent with those found in the related literature.

> Authors have to organise their Conclusions newly. 

We have improved the discussions and conclusions sections.

Please see our revisions in the updated manuscript. We highlighted the changes and the parts of the manuscript addressing the rest of the concerns. 

Round 2

Reviewer 1 Report

The overall modifications have been good. Therefore, I recommend that this manuscript to accept.

Author Response

Thank you very much for your comments.

Reviewer 3 Report

I am very disappointed with the way of how the authors managed my previous comments. The changes were very superficial and did not tackle any of my concerns. The major revisions cannot be answered adding a single paragraph with the structure of the paper in the Introduction, a new column in a table, or a single sentence in the Conclusions. The major revisions have to be answered clearly and carefully to each of the comments of the reviewer. These kinds of answers are very disrespectful with the time that researchers devote to understand and give scientific advice.

Author Response

Thank you for your opinion. We have incorporated some changes into the new version of the paper, and the rest is addressed in our "rebuttal" document sent to the editorial office.