diff --git a/mtg_dictionary.py b/mtg_dictionary.py
new file mode 100644
index 0000000..13f6513
--- /dev/null
+++ b/mtg_dictionary.py
@@ -0,0 +1,50 @@
+import json
+
+# Generates a list of every word (with frequencies) used in MTG cards.
+
+# The 'oracle_cards.json' file is Scryfall's "Oracle Cards" file.
+# "A JSON file containing one Scryfall card object for each Oracle ID on Scryfall."
+# https://scryfall.com/docs/api/bulk-data
+with open('oracle_cards.json', encoding='utf8') as f_oracle_cards:
+    data = json.load(f_oracle_cards)
+
+
+def word_list(card):
+    if 'booster' not in card or card['booster'] is False:
+        return []
+    if 'type_line' not in card:
+        return []
+    if card['set_type'] in ['funny', 'vanguard', 'memorabilia']:
+        return []
+    if 'oracle_text' not in card:
+        return []
+    name = card['name']
+    text = card['oracle_text']
+    # remove the name; eliminates misleading occurrences of some words
+    text = text.replace(name, '')
+    text = text.replace("—", " ")
+    remove_these = "®½π—•−∞☐"
+    for char in remove_these:
+        text = text.replace(char, " ")
+    words = text.split()
+    words = [word.strip(' ,.()[]!@#$%^&*=;:"\'?<>\n\t').lower() for word in words if word.islower()]
+    return words
+
+
+d = {}
+for card in data:
+    if 'oracle_text' in card:
+        card_words = word_list(card)
+        for word in card_words:
+            if word not in d:
+                d[word] = 1
+            else:
+                d[word] += 1
+
+f_oracle_cards.close()
+
+output = open('mtgdict.csv', "w+", encoding='utf8')
+for word in d:
+    output.write(f"{word},{d[word]}\n")
+output.close()
+