Fascinating data! This is incredibly useful. Thanks for the hard work, y'all!

I'd like to meet the person who felt the need to support London to the North Sea S1901.

I can fairly easily port the code for other variants, but obviously they won't be quite so detailed, and I don't think they'd be that useful. Let me know if you really want to know though. Also, I'd appreciate any feedback on different ways to categorise the data, other factors to take into consideration, etc.

Live versus non-live. That's what I really want to see.

2ndWhiteLine, you missed the funniest opening so far.

Brest Support move to Gascony from Marseilles - Paris Support move to Gascony from Marseilles - Marseilles Move to Spain

For the record, it has 0 wins.

Interesting idea, though I'd like to see them in a pie-graph form to show what is the most successful (or some type of ordering method). Have you thought about putting it into an excel spreadsheet (or google docs spreadsheet) to allow both graph creation and reordering?

Incredible, nice work, Vaft! Thanks for putting this together!

Draugnar, coming up!

fortknox, aren't the percentages visually pleasing enough? ;) On a serious note, yes, I could paste it all to Google Docs. It seems exhausting to me though. Does someone know a quick and automatic way for the php output to produce xls or ods?

yebellz, thank you, and you're welcome!

I counter with Constantinople Move to Smyrna - Smyrna Move to Syria - Ankara Move to Black Sea. Doesn't quite compete with my recent German opening of Kiel to Berlin, though.

This is great stuff, thank you Vaft

Really interesting! And some things not as expected at all! I'll take a closer look later...

Very interesting. Well done. Looks like Russia gets the most wins, and Italy is at the bottom (by a big margin).

spyman, look at the gunboats ;)

@Vaft - We'll talk, but I have some ideas about putting the states into a database and then writing some .Net pages to render graphs and sortable/filterable lists to make it even more useable.

Thank you Vaft, great job done, hopefully it will help to avoid stupid openings in the future.
Also It might push openings towards a path, where they become self fulfilling and more players will use such openings, which are more successful, resulting future's stats will be effected by this stat.

All this hopefully results better games on the long run and the development of players. Like less 'austrian attack' opening for italy, and so on.

But sometimes it's just fun to lie your ass off to Austria and attack as Italy. Other times it's even more fun to lie your ass off to Italy and attack as Austria. Of course, it almost always results in a jugg in both cases. :-)

Draugnar, now that you mention it, it wouldn't be hard at all to rewrite the code to save everything into SQL tables instead. I just wouldn't know what to do with them, but if you have elegant solutions, I'll hear them and claim it was my idea all along and then steal your job and possibly your identity.

For those who are curious why the webpages look like shit, it's because this is not only for you guys, it's been a learning experience for me. I've used this stat-making as an excuse and a tangible task to learn php and SQL. If you want to look at my code, that's fine, but it's not elegant nor fast and I think it would make a serious programmer vomit. It works though. If you spot any errors, let me know!

Let me get to the weekend and then we'll chat. Go ahead and rework it to send it to a table, though. I can take it from there and get it up to a database.

I think the database missed my epic Con -> Smy -> Syr, Ank -> Con opening in the Winter 2011 Leagues. Resulted in a draw so byaaaahhh

Comma delimited list. Put the same data out as you are with the tables, but replace the </TD> with a comma so your data is just comma delimited... then you can put it in a '.csv' file and import it to excel without any issue.

Eden, you mean this game, gameID=57892 , where you did not in fact open to Syria? It's accounted for ;)

In case anyone is wondering, it has 3 draws and 2 defeats.

oh word, oops.

sweet! I AM THE 60%

@roka- the austrian attack is not necessarily a foolish opening. In an analysis i did on opening moves, i subjected the openings to a series of t-tests which is a deeper look than merely looking at raw percentages which can be misleading and while the austrian attack was very comparable to the lepanto, there is some measure of argument that it could be a stronger opening. Their respective standard deviations were too broad to make any definite conclusions but they certainly are stronger than other options available to italy. ultimately, though, stats alone wouldnt be enough to sway

@roka- the austrian attack is not necessarily a foolish opening. In an analysis i did on opening moves, i subjected the openings to a series of t-tests which is a deeper look than merely looking at raw percentages which can be misleading and while the austrian attack was very comparable to the lepanto, there is some measure of argument that it could be a stronger opening. Their respective standard deviations were too broad to make any definite conclusions but they certainly are stronger than other options available to italy. ultimately, though, stats alone wouldnt be enough to sway such a decision and it would have to come down to board dynamics.

I apologize if this got double-posted...phone error.

These are just statistics. They have to be analysed in the proper framework. Assume the following scenario:

The conventional wisdom states that Obriani is the best Italian opening.
Experienced players who read articles on the subject learn this, and play the Obriani a lot.
The Obriani wins more than any other opening, not because of the opening, but because of the caliber of players who play it.

Is this likely? I don't know. Is it possible? Certainly.

Statistics are just numbers. They are fun to look at and can give valuable information, but one has to be careful with them.

Live games are up! Have fun looking at them, although the sheer number of CDs may make you throw up.

Where's the all holds Italian opening? Didn't it have like a 6% win percentage?

Bwahahahaha!!!!!!! I have snuck a virus on all y'all's machines from my website! Bwahahahaha!!!!!!

J/K!

<*rusn off snickering at the joke*>

dubmdell, it did, in my alpha version that couldn't tell that openings were the same when the orders came in a different order. Did I mention that it was untested and probably completely unreliable? I should have :D The real win percent for NMRs is somewhere around 1% for all countries. I excluded it from the final tally.