ALL			Top50			DIFF	
Picks	     #	Pick%	Win%	     #	Pick%	Win%	Pick%	Win%
1,2,3	 94650	50.9	49.7	  3201	39.7	48.7	-11.2	- 1.0
1,2,4	 25253	13.6	50.8	  1207	15.0	50.3	+ 1.4	- 0.5
1,2,5	 12200	 6.6	54.9	   805	10.0	57.1	+ 3.4	+ 2.2
1,2,6	  2030	 1.1	57.0	   127	 1.6	52.8	+ 0.5	- 4.2
1,2,7	  1373	 0.7	28.3	     6	 0.1	50.0	- 0.7	+21.7
1,3,4	 21478	11.5	50.9	  1120	13.9	50.4	+ 2.3	- 0.5
1,3,5	  5914	 3.2	49.5	   349	 4.3	48.4	+ 1.1	- 1.0
1,3,6	   785	 0.4	54.5	    67	 0.8	50.7	+ 0.4	- 3.8
1,3,7	   837	 0.4	29.4	     1	 0.0	 0.0	- 0.4	-29.4
1,4,5	  6029	 3.2	50.8	   358	 4.4	49.2	+ 1.2	- 1.6
1,4,7	   159	 0.1	35.2	     0	 0.0	-	- 0.1	-
1,7,7	   124	 0.1	26.6	     0	 0.0	-	- 0.1	-
2,3,4	  9855	 5.3	48.2	   461	 5.7	46.0	+ 0.4	- 2.2
2,3,5	  3757	 2.0	47.8	   263	 3.3	48.3	+ 1.2	+ 0.5
2,3,6	  1167	 0.6	50.5	   104	 1.3	54.8	+ 0.7	+ 4.3
2,3,7	   423	 0.2	29.3	     1	 0.0	 0.0	- 0.2	-29.3

Table shows for each possible combination of picks how often it occurred and how often it won. The first set is for all games, the second set is for the 4,000+ games between top 50 players. The last two columns are differences in Pick% and Win% between the two groups. If there is a "7" in the pick that means the player was randomly assigned a territory because they picked less than six.

Some of the numbers make for too small of sample sizes to say much.

Things of note:

* Top50 players get 1,2,3 10% less than the whole (makes sense)
* Getting 1,2,3 is a slight disadvantage (I usually hate it)
* Getting 1,2,6 is the best for the whole, and I expect it would be in the top50 with more games (you get your top 2 and know all 3 of theirs)

the win% shows how often the players getting these picks win the game, right?

~~~~~yeah it should be so. that's why you say getting the "1,2,6" is the best.

Edited 6/24/2015 14:06:32

This show how important it is to make 6 picks, just look at this.

2/3/6 wins 50.5% of the time
2/3/7 wins 29.3% of the time

Similar low win rates goes for other combinations with random picks too:
1/2/7 28.3% win rate
1/3/7 29.4% win rate
1/4/7 35.2& win rate
1/7/7 26.6% win rate

Always make 6 picks


Also Smileyleg, can you do this for territories blockaded too, I love your statistics :D

Thanks for doing this! I love the huge bump in 1/2/7- looks like that level of intel has very high value, especially when combined with the ability to go for the FTB/two-turn bonus you wanted to start with.

The difference also seems to imply that one distinguishing mark of a good player is their ability to use intel gleaned from picks (and counterpick their opponents so the 1/2/7 combo is actually valuable instead of just a collection of random badly-picked territories which won't let them take full advantage of the situation).

What's also interesting is that this occurs in 1/2/7 and not in 1/2/6, which seems to imply a few things as well.

Edited 6/24/2015 14:28:21

@Hennns, I think there's a bit of a confounding effect here though. Anyone who is any good will always make 6 picks, so you know off the bat that if someone got their "7th" pick, they're likely to be a below-average player. You'd have to compare the win rates of players who get their "7th" pick to the win rates of players who make less than 6 picks but don't get a random pick. Your overall message is correct though. Always make 6 picks!

@knyte, the win rate for 1/2/7 among the top 50 players is rather meaningless, since it only happened 6 total times.

Edited 6/24/2015 14:37:24

@Beren: Ah, didn't notice the sample size.

How exactly does one get 7th pick on the 1v1 ladder? I had assumed this was impossible.

7th pick = you got something you didn't pick, so you made 5 or fewer picks.

125 and 126 are usually lethal if you picked appropriately because your 5th/6th in this scenario is generally a counter. You have intel on 2/3 picks and a counter. Your 6th could even be a counter to a triple pick. So yeah if I get 123, 125, or 126, I know I'm probably going to win.

I disagree about 123 being a disadvantage. The reason it happens a lot less between the top 50 players is because both players can read the map appropriately. Getting 123 usually means your opponent read the map wrong and you have really nice expansion.

Thanks all!

Hennns: Sure, I can do a blockade one.

Forbidden: As knyte said, and I explained in the text :)

Interesting thing I just noticed: if you don't get your first pick you have a definite disadvantage unless you get 2,3,6 -- although 2,3,6 does quite well among top50 players (I'm usually happy when I get it). That's more of an advantage for getting first pick than I would've guessed.

Well getting 123 can also mean you read the map wrong.

If you get 126 and lose you've definitely made a mistake (it's happened to me and I knew it).

Picks 	#	Win%
1,7,7	124	26.6

lol :D Losing after your opponent got 2 random picks must look lovely.

Interesting thing I just noticed: if you don't get your first pick you have a definite disadvantage

Fizzer made a thread about that, if memory serves me right. Well see if I can find it.

EDIT: I can't find it :( I think he said something about the player who gets 1st pick wins 10% more games. This is all memory based through.

I can be confusing it with the rating formula. There 1st pick is valued as 10 points.

Edited 6/25/2015 00:36:27

Oh wow, this shit again? Try no one cares.

^ Ignore this troll.

Why are you so sad Cornucopia?

You sure seem to care!

He's not in Kansas anymore

what map? :)

I always do at least 6 picks, to avoid randomness.

Cornucopia: Oh wow, this shit again? Try no one cares.

What happened to rule #1: You shall be respectful to other players at all times.?

Edit: Once (in https://www.warlight.net/MultiPlayer?GameID=8683194), I found out all 3 of a player's picks, on turn 0.

Edited 6/30/2015 20:40:28

I always do 7 picks, just to avoid the random one. I always new there was a slight possibility of getting a random one, but I've never had to use it.

I don't think that gives you any benefit. A "pick 7" scenario would be something like this:

Player A makes only 5 picks because he thinks he's read the map perfectly and that he should get at least 3 of his first 5 picks. Player B makes the same picks as Player A and ends up getting all 3 of theirs- so Player A is stuck with one random pick (a "pick 7").

Outside that, it's completely impossible to get a random pick if you make at least 6 picks each game. The only way to get a random pick is if you have an overlap of 3 or more picks with your opponent- the system doesn't just arbitrarily decide to screw you over, at least not to my experience. Yours might be different.