diff --git a/hotdog/command/urlinfo_/imdb.py b/hotdog/command/urlinfo_/imdb.py
index c3e2aa9..4db2c68 100644
--- a/hotdog/command/urlinfo_/imdb.py
+++ b/hotdog/command/urlinfo_/imdb.py
@@ -33,7 +33,7 @@ def thumbnail(url, width=182, height=268):
     The default settings are what IMDb currently uses for desktop display.
     """
     resize = f"UY{height}"  # there's also 'UX' to resize on width
-    offset = "2,0"  # by setting non-0 for the first value the image is fitted
+    offset = "0,0"
     crop = f"CR{offset},{width},{height}"
     al = "AL"  # not sure what this is, doesn't seem to do much but they use it.
 
@@ -102,10 +102,10 @@ class Extracted:
     genres: List[str]
     description: str
     published: date
-    duration_s: int
-    rating_value: float
-    rating_count: int
     creators: List[str]
+    duration_s: Optional[int] = None
+    rating_value: Optional[float] = None
+    rating_count: Optional[int] = None
 
 
 async def extractor(info):
@@ -120,8 +120,10 @@ async def extractor(info):
         return
 
     ld = json.loads(parser.value)
-    assert ld["@context"] == "http://schema.org" and ld["@type"] == "Movie"
-    assert ld["aggregateRating"]["@type"] == "AggregateRating"
+    assert ld["@context"] == "http://schema.org" and ld["@type"] in (
+        "Movie",
+        "CreativeWork",
+    )
 
     creators = []
     for k in "director", "creator":
@@ -136,12 +138,17 @@ async def extractor(info):
         genres=ld["genre"],
         description=ld["description"],
         published=date.fromisoformat(ld["datePublished"]),
-        duration_s=parse_period(ld["duration"]),
-        rating_value=float(ld["aggregateRating"]["ratingValue"]),
-        rating_count=ld["aggregateRating"]["ratingCount"],
         creators=creators,
     )
 
+    if "duration" in ld:
+        info.extracted.duration_s = parse_period(ld["duration"])
+
+    if "aggregateRating" in ld:
+        assert ld["aggregateRating"]["@type"] == "AggregateRating"
+        info.extracted.rating_value = float(ld["aggregateRating"]["ratingValue"])
+        info.extracted.rating_count = ld["aggregateRating"]["ratingCount"]
+
 
 async def handle(message, url, info):
     ex = clone(info.extracted)
@@ -165,10 +172,12 @@ async def handle(message, url, info):
 
     details = [
         f"🖋 {' ∕ '.join(ex.creators[:2])}",
-        f"⏱ {pretty_duration(ex.duration_s)}",
-        f"⭐️ {ex.rating_value:_.01f} ⁄ 10 (👤 {ex.rating_count})",
-        f"🏷 {' ∕ '.join(ex.genres)}",
     ]
+    if ex.duration_s:
+        details.append(f"⏱ {pretty_duration(ex.duration_s)}")
+    if ex.rating_count:
+        details.append(f"⭐️ {ex.rating_value:_.01f} ⁄ 10 (👤 {ex.rating_count})")
+    details.append(f"🏷 {' ∕ '.join(ex.genres[:3])}")
     lines = [
         f"<b>{ex.title}</b> (<b>{ex.published:%Y}</b>)",
         f"{', '.join(details)}",