{
"cells": [
{
"cell_type": "markdown",
"id": "0",
"metadata": {},
"source": [
"# CFB conference table\n",
"\n",
"**The brief:** the season-review newsletter needs the final Big Ten standings as an image: 1600 px wide for the email,\n",
"plus a square cut for social. Indiana went 16-0 and won the national title, so the table should make that obvious.\n",
"The records are built from the cfbfastR schedule through `sportsdataverse.cfb`, and the table is great_tables with\n",
"sdvplot's logo, theme and export helpers."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "1",
"metadata": {},
"outputs": [],
"source": [
"import tempfile\n",
"from pathlib import Path\n",
"\n",
"import polars as pl\n",
"import sportsdataverse.cfb as cfb\n",
"from great_tables import GT, html, loc, nanoplot_options, style\n",
"from IPython.display import Image\n",
"from PIL import Image as PILImage\n",
"\n",
"import sdvplot\n",
"from sdvplot.great_tables import gt_save_crop, gt_sdv_logos, gt_social_crop, gt_theme_sdv\n",
"\n",
"SEASON = 2025\n",
"CONFERENCE = \"Big Ten\"\n",
"OUT = Path(tempfile.mkdtemp(prefix=\"sdvplot-recipe-\")) # where the exports go; use your own folder"
]
},
{
"cell_type": "markdown",
"id": "2",
"metadata": {},
"source": [
"## 1. Get the data\n",
"\n",
"The schedule has one row per game. Stacking the home and away sides gives one row per team per game, which makes\n",
"every record a `group_by`. The ESPN team ids arrive as integers; they become strings once, at the boundary, because\n",
"sdvplot's `team_id` is always a string. The margins are kept in date order as a list, one value per game, for a\n",
"small chart later."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "3",
"metadata": {},
"outputs": [],
"source": [
"schedule = cfb.load_cfb_schedule([SEASON]).filter(pl.col(\"completed\"))\n",
"\n",
"\n",
"def side(me, opp):\n",
" return schedule.select(\n",
" \"start_date\",\n",
" \"season_type\",\n",
" \"conference_game\",\n",
" \"notes\",\n",
" team_id=pl.col(f\"{me}_id\").cast(pl.Utf8),\n",
" team=f\"{me}_team\",\n",
" conference=f\"{me}_conference\",\n",
" opponent=f\"{opp}_team\",\n",
" pf=f\"{me}_points\",\n",
" pa=f\"{opp}_points\",\n",
" )\n",
"\n",
"\n",
"games = (\n",
" pl.concat([side(\"home\", \"away\"), side(\"away\", \"home\")])\n",
" .filter(pl.col(\"conference\") == CONFERENCE)\n",
" .sort(\"start_date\")\n",
" .with_columns(won=pl.col(\"pf\") > pl.col(\"pa\"))\n",
")\n",
"in_conf = pl.col(\"conference_game\")\n",
"standings = (\n",
" games.group_by(\"team_id\", \"team\", maintain_order=True)\n",
" .agg(\n",
" conf_w=(pl.col(\"won\") & in_conf).sum(),\n",
" conf_l=(~pl.col(\"won\") & in_conf).sum(),\n",
" w=pl.col(\"won\").sum(),\n",
" l=(~pl.col(\"won\")).sum(),\n",
" pf=pl.col(\"pf\").mean(),\n",
" pa=pl.col(\"pa\").mean(),\n",
" margins=pl.col(\"pf\") - pl.col(\"pa\"),\n",
" # the last game: where a bowl or the playoff shows up\n",
" last_type=pl.col(\"season_type\").last(),\n",
" last_won=pl.col(\"won\").last(),\n",
" last_score=pl.format(\"{}-{}\", pl.max_horizontal(\"pf\", \"pa\"), pl.min_horizontal(\"pf\", \"pa\")).last(),\n",
" last_opponent=pl.col(\"opponent\").last(),\n",
" last_event=pl.col(\"notes\").last(),\n",
" )\n",
" .sort([\"conf_w\", \"w\", \"team\"], descending=[True, True, False])\n",
")\n",
"standings.head()"
]
},
{
"cell_type": "markdown",
"id": "4",
"metadata": {},
"source": [
"## 2. The first draft\n",
"\n",
"Hand the frame to great_tables as it is."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "5",
"metadata": {},
"outputs": [],
"source": [
"GT(standings)"
]
},
{
"cell_type": "markdown",
"id": "6",
"metadata": {},
"source": [
"Every number is there, and none of it is readable: ids, a list printed as text, a dozen decimals and\n",
"column names only the analyst knows.\n",
"\n",
"## 3. Shape it for a reader\n",
"\n",
"Records read as \"9-0\", not two columns. Each team's last game becomes one short line (\"W 27-21 vs Miami, CFP\n",
"National Championship\"), which is where the national title shows up. Columns get real labels, conference and overall records sit under spanners, and the averages get\n",
"one decimal."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "7",
"metadata": {},
"outputs": [],
"source": [
"event = (\n",
" pl.col(\"last_event\")\n",
" .str.replace(\" Presented by.*\", \"\")\n",
" .str.replace(\" at the .*\", \"\")\n",
" .str.replace(\"College Football Playoff\", \"CFP\")\n",
")\n",
"postseason = (\n",
" pl.when(pl.col(\"last_type\") == \"postseason\")\n",
" .then(\n",
" pl.format(\n",
" \"{} {} vs {}, {}\",\n",
" pl.when(\"last_won\").then(pl.lit(\"W\")).otherwise(pl.lit(\"L\")),\n",
" \"last_score\",\n",
" \"last_opponent\",\n",
" event,\n",
" )\n",
" )\n",
" .otherwise(pl.lit(\"\"))\n",
")\n",
"\n",
"table = standings.with_columns(\n",
" conf=pl.format(\"{}-{}\", \"conf_w\", \"conf_l\"),\n",
" overall=pl.format(\"{}-{}\", \"w\", \"l\"),\n",
" postseason=postseason,\n",
").select(\"team_id\", \"team\", \"conf\", \"overall\", \"pf\", \"pa\", \"margins\", \"postseason\")\n",
"\n",
"draft = (\n",
" GT(table)\n",
" .cols_hide([\"team_id\", \"margins\"])\n",
" .cols_label(team=\"Team\", conf=\"W-L\", overall=\"W-L\", pf=\"Pts/G\", pa=\"Opp/G\", postseason=\"Postseason\")\n",
" .tab_spanner(\"Conference\", [\"conf\"])\n",
" .tab_spanner(\"Overall\", [\"overall\", \"pf\", \"pa\"])\n",
" .fmt_number([\"pf\", \"pa\"], decimals=1)\n",
" .cols_align(\"center\", [\"conf\", \"overall\", \"pf\", \"pa\"])\n",
" .cols_align(\"left\", [\"team\", \"postseason\"])\n",
")\n",
"draft"
]
},
{
"cell_type": "markdown",
"id": "8",
"metadata": {},
"source": [
"## 4. Logos and the season at a glance\n",
"\n",
"`gt_sdv_logos` turns the `team_id` column into logos; the ESPN ids resolve as they are. The margins list becomes a\n",
"nanoplot, great_tables' in-cell bar chart: one bar per game, green for a win and red for a loss, so a perfect season\n",
"is a solid green row."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "9",
"metadata": {},
"outputs": [],
"source": [
"margin_bars = nanoplot_options(\n",
" data_bar_fill_color=\"#2e8540\",\n",
" data_bar_negative_fill_color=\"#c0392b\",\n",
" data_bar_stroke_color=\"transparent\",\n",
" data_bar_negative_stroke_color=\"transparent\",\n",
" show_data_points=False,\n",
" show_reference_line=False,\n",
" show_vertical_guides=False,\n",
" show_y_axis_guide=False,\n",
" interactive_data_values=True, # values on hover only, so the saved image stays clean\n",
")\n",
"with_marks = (\n",
" draft.cols_unhide([\"team_id\", \"margins\"])\n",
" .pipe(gt_sdv_logos, \"team_id\", league=\"cfb\", season=SEASON, height=26)\n",
" .fmt_nanoplot(\"margins\", plot_type=\"bar\", autoscale=True, options=margin_bars)\n",
" .cols_label(team_id=\"\", margins=\"Game by game\")\n",
" .tab_spanner(\"Margin\", [\"margins\"])\n",
")\n",
"with_marks"
]
},
{
"cell_type": "markdown",
"id": "10",
"metadata": {},
"source": [
"## 5. Theme it and say what it means\n",
"\n",
"A theme does the typography and rules in one call (`gt_theme_sdv`, the SportsDataverse house style, here). The title says the news, the subtitle\n",
"how to read the table, and the source note credits the data. Indiana's row gets a soft fill in its own red, and a\n",
"footnote owns up to the ordering: teams tied on conference record are listed by overall record, which is not the\n",
"conference's tiebreaker."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "11",
"metadata": {},
"outputs": [],
"source": [
"indiana = standings.filter(pl.col(\"team\") == \"Indiana\")\n",
"fill = sdvplot.team_colors(indiana[\"team_id\"][0], \"cfb\") + \"1f\" # the primary color at 12% opacity\n",
"\n",
"final = (\n",
" with_marks.tab_header(\n",
" title=f\"Indiana ran the table: 9-0 in the {CONFERENCE}, 16-0 overall and national champion\",\n",
" subtitle=html(\n",
" f\"Final {SEASON} {CONFERENCE} standings. Bars are each game's margin, in date order: \"\n",
" \"wins and \"\n",
" \"losses.\"\n",
" ),\n",
" )\n",
" .tab_source_note(\"Data: cfbfastR via sportsdataverse-py | Table: sdvplot + great_tables\")\n",
" .tab_footnote(\n",
" \"Teams tied on conference record are listed by overall record, then by name.\",\n",
" locations=loc.column_labels(columns=\"conf\"),\n",
" )\n",
" .tab_style(style.fill(fill), loc.body(rows=pl.col(\"team\") == \"Indiana\"))\n",
" .tab_style(style.text(weight=\"bold\"), loc.body(columns=\"team\", rows=pl.col(\"team\") == \"Indiana\"))\n",
" .pipe(gt_theme_sdv)\n",
")\n",
"final"
]
},
{
"cell_type": "markdown",
"id": "12",
"metadata": {},
"source": [
"## 6. Export for the newsletter and for social\n",
"\n",
"`gt_save_crop` renders the table in headless Chrome, trims it with an even border and, with `width=`, scales it to\n",
"the email's 1600 px. `gt_social_crop` centers the same table on a square canvas for Instagram, never cropping it:\n",
"a tall table just gets side padding."
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "13",
"metadata": {
"sdvplot_gallery": {
"alt": "great_tables standings of the 2025 Big Ten with team logos, records and a bar per game, Indiana highlighted as 16-0 national champion",
"title": "Big Ten final standings table"
},
"tags": [
"gallery"
]
},
"outputs": [],
"source": [
"newsletter = gt_save_crop(final, OUT / \"big_ten_1600.png\", width=1600)\n",
"square = gt_social_crop(final, OUT / \"big_ten_1080x1080.png\", aspect_ratio=\"1:1\", width=1080)\n",
"for f in (newsletter, square):\n",
" print(Path(f).name, PILImage.open(f).size)\n",
"Image(newsletter, width=800)"
]
},
{
"cell_type": "markdown",
"id": "14",
"metadata": {},
"source": [
"The square cut keeps the whole table and pads the sides:"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "15",
"metadata": {},
"outputs": [],
"source": [
"Image(square, width=540)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"name": "python"
},
"sdvplot": {
"description": "Build the final Big Ten standings table for a newsletter with logos, records and game-by-game bars, exported at 1600 px and as a square for social.",
"label": "CFB conference table",
"position": 2
}
},
"nbformat": 4,
"nbformat_minor": 5
}