Detailed Replay Data Analysis

[GreatGonzales]

OK GUYS. Behold the fruit of my labors. This project proved to be too much for Google Docs, so you'll have to:

Download the Excel File Here

I kind of went data crazy and created an awesome spreadsheet which dynamically calculates statistics related to Outwitters replay data. Many thanks to CodePenguin for making his awesome data available to us as a resource. If he's not the MVP of this community, I don't know who is.

I intend to iterate on this project (for example, currently I do not have any 2v2 stats). Below I will detail the features currently available in this initial release:

Games Created Over Time

In this tab, I have provided a graph which illustrates the number of games created on a particular date since the start of Season 5, which began on 3/5/2014. Why Season 5? Because that is when OSN began collecting ALL replay data, not just user-submitted replays. As you can see, there is clear - albiet slight - downward trend in player activity over time. That is to be expected with any game as old as Outwitters. It will be interesting to track this over time!

Overall Team Stats

One of the big questions that we wanted to answer using this data was how teams perform against other teams. We all know that Scallywags are a good bet vs Veggies, but what does the data say? In this tab, I've provided these statistics, broken down by league. Since team vs team win rate was the more interesting statistic, I've controlled for persons of differing leagues here; so, Fluffy vs Fluffy, ST vs ST - no ST vs Fluffy.

Team Stats by Map

Yes - the same statistics broken out by map! So, now we can see how adorables do vs feedback on Glitch for example. The answer is not particularly well in that case. Same thing here as far as controlling for cross-league play; only matches between two people of the same league are considered for these stats.

Player Stats

This tab allows you to enter the name of a player, and the document will populate replay statistics for that player. Win rate by team, win rate by map, etc. Additionally, you can enter the name of a second player to see the win rate between the two. Pretty cool!

---

Some great things about this spreadsheet:

1. These statistics are calculated using 100% of the data that is available for download on OSN. At the time of writing, that is more than 155,000 replays. (!!!)
2. When a new download becomes available each week, it only takes me about 5 minutes to add the new data to this document and re-post. I've designed the formulas so I can simply paste them down for the new records and everything will dynamically update. So, we'll always have the most up to date stats.
3. I've done the hard work of compiling all of this data - feel free to use it for your own data projects!

Please let me know if there are questions, comments or suggestions. Like I said, I intend to iterate on this a bit more, but I'm interested to hear what people would like to see.

---

Some surprising findings:

Feedback is the best team on Peekaboo?

ST Scallywags have the best record on Sharkfood Island?

ST Feedback CRUSH ST Veggienauts?

OneAlexLeft has only 3 games recorded in OSN?

[spacechef]

Sounds awesome!

[onealexleft]

(05-10-2014 04:31 PM)GreatGonzales Wrote:  OneAlexLeft has only 3 games recorded in OSN?

4 now :>.

[onealexleft]

Really awesome to see these stats come from the data dumps!

[awpertunity]

This is awesome. And confirms how bad I am at War Garden.

[lawtai]

This is great! How are you planning on managing all the data though? The file is already 40+MB's, and will only continue to grow.

[Mag!cGuy]

This is... badass!

And, I like that I don't even have to type my name in the players stats.

The page "games created over time" is bugged to me.

[mr_13en]

Wow! Mad props for doing this GG!

And this definitely confirms Skull Duggery is my worst map

[Poc 13]

Does this have to be downloaded on a computer? It's not working for my iPhone, it says unable to read document unknown exception

[Chemoeum]

I do not have Excel.