This repository contains MFL data gathered via ffscrapr and stored as Apache Arrow’s Parquet files.
Usage instructions currently written out in R - if you use a different programming language, I would gladly take a PR with parallel instructions!
There are currently two included datasets: SafeLeagues data (2019-2020) and ScottFishBowl data (2016-2020).
Within each of these, you will find subfolders:
league: contains a table summarising features about the leagues within the dataset such as scoring_flags, best_ball, idp, qb_type, player_copies, league_size.draft: contains the most recent draft for that league yearstandings: contains season-finish data, including H2H/All-Play win records, points for, potential points etcstarters: contains the starters for each week of the regular seasontransactions: contains all public transaction data (adds, drops, trades, blind bidding etc)
You will need to clone or download this repository to your machine in order to use this data (as of this writing). You can do this via the buttons above, or by running this command in a terminal.
git clone https://www.github.com/dynastyprocess/data-mfl_public.git
The Arrow R package can be installed with:
install.packages("arrow")
# or, if that gives you problems, I usually can get the nightly version to work a little better
install.packages("arrow", repos = "https://arrow-r-nightly.s3.amazonaws.com")If both of these give you problems, please see the official arrow help for installation here: https://arrow.apache.org/docs/r/articles/install.html)
Each subfolder (i.e. scottfishbowl/draft) is designed to be opened as
a multi-file Arrow dataset.
Here’s an example of accessing the league summary table:
suppressPackageStartupMessages({
library(arrow)
library(dplyr)
})
#> Warning: package 'arrow' was built under R version 4.0.5
safeleagues_leagues <- arrow::open_dataset("safeleagues/league/") %>%
dplyr::collect()
safeleagues_leagues
#> # A tibble: 1,838 x 14
#> league_id league_name franchise_count qb_type
#> <chr> <chr> <dbl> <chr>
#> 1 10089 SafeLeagues Dynasty 143 - Basic (SF) 12 2QB/SF
#> 2 10108 SafeLeagues Napoleon 02 - CP Challenge #1 12 1QB
#> 3 10124 SafeLeagues Dynasty 120 - Premium (SF/TE) 12 2QB/SF
#> 4 10608 SafeLeagues Dynasty 078 - Rival (SF) 12 2QB/SF
#> 5 10627 SafeLeagues Dynasty 127 - Basic (SF) 12 2QB/SF
#> 6 10672 SafeLeagues Dynasty 090 - Basic (SF/TE) 12 2QB/SF
#> 7 10974 SafeLeagues Dynasty 142 - Basic 12 1QB
#> 8 11122 SafeLeagues Redraft 09 - Starter (SF) 12 2QB/SF
#> 9 11205 SafeLeagues Redraft 07 - Starter (SF) 12 2QB/SF
#> 10 11360 SafeLeagues Dynasty 186 - Basic (SF/TE) 12 2QB/SF
#> # ... with 1,828 more rows, and 10 more variables: idp <lgl>,
#> # scoring_flags <chr>, best_ball <lgl>, salary_cap <lgl>,
#> # player_copies <dbl>, years_active <chr>, qb_count <chr>, roster_size <dbl>,
#> # league_depth <dbl>, year <int>Arrow is very useful because you can pass filters, sorts, and selects to
the dataset before loading it into memory. For example, if you want to
look at waiver wire adds (and not drops) in Scott Fish Bowl’s 2020
leagues, you can do so like this:
sfb_adds <- arrow::open_dataset("scottfishbowl/transactions/") %>%
filter(type_desc == "added", year == 2020, sfb_type == "official") %>%
arrange(desc(bbid_spent)) %>%
select(league_id, player_id, player_name, pos, team, bbid_spent, franchise_id, franchise_name, timestamp) %>%
collect()
sfb_adds
#> # A tibble: 20,400 x 9
#> league_id player_id player_name pos team bbid_spent franchise_id
#> <int> <chr> <chr> <chr> <chr> <dbl> <chr>
#> 1 13411 14811 Ahmed, Salvon RB MIA 100 0041
#> 2 13411 10313 Dalton, Andy QB DAL 100 0063
#> 3 13411 14265 Fulgham, Travis WR PHI 100 0051
#> 4 13411 10313 Dalton, Andy QB DAL 100 0055
#> 5 16428 14096 Ozigbo, Devine RB JAC 100 0062
#> 6 17910 14095 Johnson, Ty RB NYJ 100 0059
#> 7 17910 10313 Dalton, Andy QB DAL 100 0057
#> 8 17910 10313 Dalton, Andy QB DAL 100 0001
#> 9 17910 14065 Rypien, Brett QB DEN 100 0093
#> 10 17910 14813 Robinson, James RB JAC 100 0032
#> # ... with 20,390 more rows, and 2 more variables: franchise_name <chr>,
#> # timestamp <dttm>Many hands make light work, especially when maintaining open data! Here are some ways you can contribute to this project:
-
You can open an issue if you’d like to request specific data or report a bug/error.
-
You can sponsor this project by donating to help with server costs!
Please note that this project is released with a Contributor Code of Conduct - by participating, you agree to abide by these terms.
This data is released under CC0.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent