如何获取并分析Steam游戏全部评论（含Steam及其他购买渠道）

Ah, I’ve run into this exact issue before! Using start_offset to paginate Steam reviews doesn’t work reliably because Steam’s appreviews API relies on cursor-based pagination instead of fixed offsets. Offsets break when reviews are sorted dynamically (or new reviews pop up mid-scrape), leading to missing or duplicate comments. Here’s how to fix this properly:

Step 1: Start with an Initial Request to Grab the First Cursor #

Skip the start_offset parameter entirely for your first call. This will return the first batch of reviews and a unique cursor value you’ll use to fetch subsequent pages:

http://store.steampowered.com/appreviews/278620?json=1&filter=all

In the JSON response, look for the cursor field (it’ll be a long, random string like AoJ4cF9q...). This is your ticket to the next page of reviews.

Step 2: Paginate Using the Cursor #

Instead of incrementing an offset, plug the cursor from each response into your next request. Repeat this cycle until:

• The reviews array in the response is empty, OR
• The cursor value becomes * (this signals there are no more reviews left to fetch)

Example request for the second page:

http://store.steampowered.com/appreviews/278620?json=1&filter=all&cursor=AoJ4cF9q...

Step 3: Spot Steam Purchasers Easily #

Each review object in the response includes two key fields to identify how the user got the game:

• steam_purchase: Set to true if the user bought the game directly via Steam
• received_for_free: Set to true if the user got the game for free (via giveaways, promos, etc.)

Just filter for reviews where steam_purchase === true to isolate users who bought through Steam.

Critical Tips to Avoid Headaches #

• Rate Limiting: Steam will block your requests if you scrape too fast. Add a 1-2 second delay between each call to stay under the radar.
• Filter Parameter: Using filter=all ensures you collect every public review (not just positive/negative or recent ones). Omit it only if you want the default sorted set, but all is better for full collection.
• Consistency: The API returns exactly the same reviews you’d see on the Steam store page, so hidden/moderated reviews won’t be included (which matches what’s publicly visible).

内容的提问来源于stack exchange，提问作者Isabelle