|
Looking for
The Spawning Tool is a handy way to aggregate data from tons of replays and generate statistics to learn about the game. You can:
- Upload personal and tournament replays
- View build orders for each player
- Browse replays by build order, event, map, matchup, player, game length, and more
- Look at statistics about win rates, unit counts, abilities cast
So whether you're a casual viewer who wants to know more about their favorite players, a player who wants to copy professional build orders, a professional player who wants to analyze their own and others' play style, a caster who wants to mix statistics into their casts, or any other StarCraft community member, Spawning Tool may be helpful for you!
Broadly, the goal of Spawning Tool is to make quantitative thinking pervasive in the StarCraft community. Like sabermetrics in baseball, statistics and concrete analysis gives all of us another way to understand the game. Spawning Tool aggregates all of these replays and presents summary data about the games to make us think differently. For more thoughts, read here and here.
If you're interested, there are lots of ways to contribute.
- Play around with the site and share any interesting statistics you come up with
- Tag replays with build orders (let me know, and I can give you permission
- Send me feedback on bugs or features or ideas you would like to see
- If you're handy with design, coding, or statistics, I could use collaborators
SCREENSHOTS (Sorry they're big for the screen) Build Orders + Show Spoiler + Browse Replays + Show Spoiler +
LINKS Spawning Tool home page Twitter Blog updates Change Log
|
This is attractive and you are attractive.
|
|
Wow, the research tool is much improved since last time I stopped by. You've got some awesome fine grained data there.
I think that the number averages are pretty misleading. Lets say that Hyun goes muta-ling-bane with no roaches and roach-bane all in with no mutas as two primary strategies. There is no way that hyun wins games with an average of 40 roaches. When he goes roach all in he probably makes many more than that. Averages across both game types really skew the unit counts in the chart.
I don't have fantastic ideas for resolving this issue but perhaps you could only average over games that unit was produced and add a second column with that number, possibly broken down by win/loss. That way you'd be able to say that he only went muta in 4/32 games but when he did he averaged 30+ over the course of the game, losing each one of them; 3/3 OP!
|
What language is the site written in? I'm a PHP/Python developer, I could help if you are interested.
|
On August 31 2013 01:39 ShadesofGraylin wrote: Wow, the research tool is much improved since last time I stopped by. You've got some awesome fine grained data there.
I think that the number averages are pretty misleading. Lets say that Hyun goes muta-ling-bane with no roaches and roach-bane all in with no mutas as two primary strategies. There is no way that hyun wins games with an average of 40 roaches. When he goes roach all in he probably makes many more than that. Averages across both game types really skew the unit counts in the chart.
I don't have fantastic ideas for resolving this issue but perhaps you could only average over games that unit was produced and add a second column with that number, possibly broken down by win/loss. That way you'd be able to say that he only went muta in 4/32 games but when he did he averaged 30+ over the course of the game, losing each one of them; 3/3 OP!
Thanks for the feedback; yeah, I'm excited to see more of the data out there.
Totally agree that the averages aren't really meaningful at the moment; it is tremendously biased by game length and strategies. For the first part, I was thinking that narrowing production over a fixed time period can present better idea (e.g. how many Roaches produced from 10-20 minutes). For the second, my hope is to label all of the replays well enough so that instead of just saying "Hyun makes Roaches in ZvT", we could tack on the build order tags to say "Hyun makes a lot of Roaches he goes Roach/Hydra in ZvT". That, of course, will require a lot of hand-labeling, but I think that work is intrinsically valuable.
|
On August 31 2013 03:11 ObsessionSC2 wrote: What language is the site written in? I'm a PHP/Python developer, I could help if you are interested.
It's all python/django. PM me if you had thoughts or are interested.
|
Great work!
Why is the supply count checked only every 10 seconds? With the new replay format you can do it continuously - only chronoboosts and cancels would need some tweaking and possibly heuristic solutions.
|
On August 30 2013 16:47 -Kaiser- wrote: This is attractive and you are attractive. Thats a pretty cool first post in my opinion.
|
On August 31 2013 19:17 slmw wrote: Great work!
Why is the supply count checked only every 10 seconds? With the new replay format you can do it continuously - only chronoboosts and cancels would need some tweaking and possibly heuristic solutions.
Player stats are given every 10 seconds. It's the easy and good solution (small enough and equidistant).
Otherwise you would have to track all the units building and finished building and dying yourself. I think this is possible because tracker events say when units are finished building or killed.
|
On August 31 2013 19:52 SinCitta wrote:Show nested quote +On August 31 2013 19:17 slmw wrote: Great work!
Why is the supply count checked only every 10 seconds? With the new replay format you can do it continuously - only chronoboosts and cancels would need some tweaking and possibly heuristic solutions. Player stats are given every 10 seconds. It's the easy and good solution (small enough and equidistant). Otherwise you would have to track all the units building and finished building and dying yourself. I think this is possible because tracker events say when units are finished building or killed.
Yeah, it is possible using the new events. In the case of studying openings/build orders, I think it's important to get the supply counts accurately. I wrote a sc2reader plugin for myself which does this for Terran openings.
|
On August 31 2013 19:57 slmw wrote: Yeah, it is possible using the new events. In the case of studying openings/build orders, I think it's important to get the supply counts accurately. I wrote a sc2reader plugin for myself which does this for Terran openings. I agree that more accuate supply counts are possible if you are very careful about it. Protoss and Zerg are likely going to be trickier than Terran though. Woul you be interested in sharing your code?
|
Yup, you guys nailed it: I'm using the playerstats, which only update every 10 seconds. I initially did it because it was easy, and it hasn't been a big priority for me to revisit it yet. Since you guys noticed, though, I'll put it in my backlog of features to address. Thanks for the feedback!
https://github.com/StoicLoofah/spawningtool/issues/19
|
SPAWNING TOOL BIRTHDAY UPDATE
A year ago, I launched Spawning Tool, and it’s grown tremendously in that time. It started out as an experiment in using Blizzard’s replay format to grab build orders. Since then, it has become a site for organizing and labeling replays not only to steal build orders but also to analyze replays in bulk.
So where are we now? By the numbers, Spawning Tool has:
1. 9,051 replays uploaded 2. 132,364 replay tags 3. 23,092 lines of code
I like to hope that Spawning Tool has contributed meaningfully to our understanding and analysis of StarCraft. Highlights are:
1. Comparing win rates by supply difference (part 1, part 2, reddit) 2. Putting PvT Blink Stalkers in perspective (blog) 3. Finding close/cross position win rates (blog, reddit)
I ended up going on hiatus for quite awhile around the beginning of this year, but I have cranked out a few changes recently to highlight as well:
1. Tags are now directly searchable so you can understand the hierarchy and dig down into specific builds and players 2. Added spawn positions for players to mark cross and close positions 3. Started using machine learning to suggest build order tags for replays 4. Added an easy accept/reject option for rapidly labeling build orders 5. Drag-and-drop file upload 6. and lots of other bug fixes, optimizations, and changes
Of course, I have an ask for all of you as well:
1. Label a few build orders just by accepting or rejecting suggested builds. The archive of replays is as good as its searchability, and build orders still require human expert knowledge 2. Fill out a survey about your experience with Spawning Tool. I would love to know where to take the site from here
I think that's about it! Let me know if you have any feedback or suggestions on the site. And if someone could let me know how to edit my original post, I will update that as well.
|
|
|
|