cornucopia down, trminator up! I like these numbers, though I don't play the 1vs1 ladder. And I liked watching those two games linked. Good job!

Thanks. Now if only I could impress Cornucopia...

@ Cornucopia: If nobody cared about these numbers and data, this thread wouldn't be notable. Who wouldn't want to do analysis of the 1v1 ladder?

I've been playing a bit with the numbers.
Unless I did something terribly wrong, I got:

On average how many bonuses are captured in given turn:

T1	T2	T3	T4	T5	T6+	total
0.1565	1.5085	0.9484	0.7608	0.6087	3.4061	7.3889

Which frankly I find that a bit odd. Surprised so few bonuses captured turn 3+, so many few total, especially considering those numbers are joint for 2 players. Maybe some games simply ended early, but still, only 3,37 in first 4 turns?

Widzisz: your averages look correct.

I think it may seem a little low at first, and it's definitely possible I made another mistake, but the more I think of it the more it makes sense.

This is among better players and better players will often have countering picks and picks which border each other immediately, etc.

As it stands it shows each player has an average of 1.7 bonuses after turn 4, which seems about right to me. Often you can't safely complete your 3 bonuses in the first 4 turns--doing so typically requires almost all of your armies to be used for expansion which often isn't an option.

Widzisz, you included Hawaii, Japan, Alaska and Korea in your calculations. Excluding these we get the following numbers:

T1	T2	T3	T4	T5	T6+	total
0.1565	1.5082	0.9446	0.7474	0.5814	3.0569	6.9950

They seem right. Turn 2 is common for 2 bonuses, then turn 3 is 1 bonus, turn 4 would have to be an outside expansion or bad luck or just bad expansion. By turn 5 you've usually met your opponent on a border, so outside expansion is low. 3.37 sounds correct.

Widzisz, you included Hawaii, Japan, Alaska and Korea in your calculations.

Yea.

Lolowut, it's for both players total. I always thought typical strategy is to capture 1st bonus 1st/2nd turn, and the rest before turn 4, 5 if unlucky. So for first 5 turns it should be close to 6 bonuses, while on average it's below 4 (3.93, this mean on average not even 2 bonuses per player).

Edited 6/24/2015 22:21:35

Ah. I see what you mean now. Consider that the majority of players on the ladder don't play it correctly in the first place.

Small request, any chance to get games duration? I mean "how many games ended turn X".

So for first 5 turns it should be close to 6 bonuses, while on average it's below 4 (3.93, this mean on average not even 2 bonuses per player).

With early contact it's often hard to go for 3 bonuses in 4-5 turns

Another request: track the income at each turn for each player in a game and find the mean income and standard deviation at each turn. This would probably take a while but I would love to see that data.

Edited 6/27/2015 09:01:46

Which games did you sample? are they yours?

read carefully, professor ;)

The data comes from about 4,000 games in the last 18 months between players who were in the top 100 at the time I ran it

My next one will have income by turn, income difference by turn, win percentage by income difference per turn, and will include game duration.

Thanks a ton, smileyleg!

Very nice data collection indeed. I'm assuming that these data are for ladder games only, and not including auto games?

I'm most surprised by the fact that the T6+ percentage for Alaska and Japan are 100%, surely out of all those hundreds of thousands of games someone has taken them before turn 6? Evidently not. I'm tempted to set WL history :P

@Pulsey:

The data comes from about 4,000 games in the last 18 months between players who were in the top 100 at the time I ran it.

Missed that, thanks.