what kind of software you use for analytics

Home for all your discussion of basketball statistical analysis.
Post Reply
Snedecor
Posts: 14
Joined: Tue Aug 07, 2018 11:15 am

what kind of software you use for analytics

Post by Snedecor » Tue Aug 07, 2018 11:44 am

Hi,

I understand there are some analysis need to be done with special software and others metrics just need an spreadsheet.

Here is the question, what software are you using? I'm 'new' on these, and I would like to be more efficient in my results.

I used R and spreadsheet, get the data by box after match ends and start analyzing, in Europe we don't have by-by-play so some metrics are difficult to get, we don't have sportvu or synergy.

Any advice? Any tricks?

Thanks in advance

eminence
Posts: 103
Joined: Sun Sep 10, 2017 8:20 pm

Re: what kind of software you use for analytics

Post by eminence » Tue Aug 07, 2018 2:14 pm

I guess I usually use matlab, but you can use pretty much any language when it comes to basketball, it's not too complex/specialized. Player tracking stuff might be an exception, I've never tried to do that myself.

Snedecor
Posts: 14
Joined: Tue Aug 07, 2018 11:15 am

Re: what kind of software you use for analytics

Post by Snedecor » Tue Aug 07, 2018 4:00 pm

For tracking players, there's any software for free or open source good to use?

There's not enough real time to watch 30-40 games per weekend or even in a week in best cases, so is a tedious task......

DSMok1
Posts: 859
Joined: Thu Apr 14, 2011 11:18 pm
Location: Maine
Contact:

Re: what kind of software you use for analytics

Post by DSMok1 » Tue Aug 07, 2018 4:54 pm

I use R and Excel for my needs. I'm not doing crazy complex math or dealing with giant datasets typically.
Developer of Box Plus/Minus
APBRmetrics Forum Administrator
GodismyJudgeOK.com/DStats/
Twitter.com/DSMok1

shadow
Posts: 207
Joined: Fri May 29, 2015 3:38 am

Re: what kind of software you use for analytics

Post by shadow » Tue Aug 07, 2018 7:44 pm

I use Python for both scraping and regression analysis.

Rd11490
Posts: 111
Joined: Mon Sep 29, 2014 4:54 am

Re: what kind of software you use for analytics

Post by Rd11490 » Wed Aug 08, 2018 12:58 pm

I have a mix of python and scala for my data processing and analysis, and I store my data in a mysql database

patrickj
Posts: 2
Joined: Sun Sep 02, 2018 10:54 pm

Re: what kind of software you use for analytics

Post by patrickj » Thu Sep 27, 2018 10:36 pm

R and Stata.

njworange
Posts: 5
Joined: Mon Nov 02, 2015 12:54 pm

Re: what kind of software you use for analytics

Post by njworange » Sat Oct 06, 2018 10:31 pm

I use Excel. It is enough for me because I don't have any tracking data and play-by-play in South Korea. However I hope I can analyze/calculate more metric with more language and raw data.

pdevos
Posts: 22
Joined: Fri Jun 26, 2015 6:18 pm
Contact:

Re: what kind of software you use for analytics

Post by pdevos » Sat Mar 23, 2019 4:30 pm

To acquire (scrape, pull API) data, I used Python with libraries Requests and BeautifulSoup. I may use Pandas here as well to formulate the data as tabular data (e.g. CSVs).

To "munge" and clean the data I use primarily Pandas again and a SQL Database, I prefer Postgres (aka "PostgreSQL") where I use SQLalchemy & psycopg2 to connect to that Database. I will create tables (I have a background in databases) that are easy to join later for what I need, e.g. teams, players, coaches, advanced stats for each, etc.

I really like Jupyter Notebooks when I'm doing "data exploration" for data visualization and other aggregations. The data viz libraries I use most are Matplotlib, Seaborn, and Plotly for the most part.

To utilize regression, clustering, and classification algorithms, I primarily use Scikit-Learn, which enables you do do about 3-4 lines of code (once you've transformed your data into a tabular format that works with Scikit-Learn).

R has similar features. The basic descriptive statistics packages and models are "quicker", but I've found it's syntax is a little wonky for me and I enjoy Python for much greater versatility and job prospects, ergo, I use Python more than R.

For quick and dirty discovery analysis, I might still throw things into Excel if it's just a bit of "copy pasta" vs scraping.

If wanting to learn Python and Data Analysis in Python I'd recommend Jake VanderPlas' Data Science Handbook (https://jakevdp.github.io/PythonDataScienceHandbook/) and Chriss Moffitt's Practical Business Python blogs where he really helps Excel users make the transition to Python (https://pbpython.com/)

Post Reply