Page 3 of 3

Re: First Move Tendencies on PlayDip

PostPosted: 17 Jan 2014, 19:59
by QuiritatioSusurratori
I dont understand this post:

And I'll probably get in trouble at work for taking the time to do this, but split out by openings by country (again, all of the below have occurred in >1% of the games). This table is different; it's listed as ALL / WIN_OR_DRAW / SOLO So Vie Hold / Tri -> Alb / Bud -> Ser is used 2.75% of the time by any player, 2.76% of the time by someone involved in the draw, and 3.07% of the time by those who end up soloing.


How can the numbers keep getting bigger? If 2.75% of the time an austrian player opens like that, how can 2.76% of the time an austrian player that ends in a draw open like that?

Re: First Move Tendencies on PlayDip

PostPosted: 17 Jan 2014, 20:02
by gareth66
He is quoting percentages of player in his various categories who make a particular opening.

Re: First Move Tendencies on PlayDip

PostPosted: 17 Jan 2014, 20:57
by QuiritatioSusurratori
gareth66 wrote:He is quoting percentages of player in his various categories who make a particular opening.

I guess I'm pretty dumb, but I still do not understand. Could you elaborate slightly more. English also isn't my first language, so I have some trouble understanding what you just said.

Re: First Move Tendencies on PlayDip

PostPosted: 17 Jan 2014, 21:12
by Petunia
QuiritatioSusurratori wrote:I dont understand this post:

And I'll probably get in trouble at work for taking the time to do this, but split out by openings by country (again, all of the below have occurred in >1% of the games). This table is different; it's listed as ALL / WIN_OR_DRAW / SOLO So Vie Hold / Tri -> Alb / Bud -> Ser is used 2.75% of the time by any player, 2.76% of the time by someone involved in the draw, and 3.07% of the time by those who end up soloing.


How can the numbers keep getting bigger? If 2.75% of the time an austrian player opens like that, how can 2.76% of the time an austrian player that ends in a draw open like that?


I'll give it a shot. You're reversing the sampling relationship.

You're thinking: Austria opens [whatever] 2.75%, Austria opens [whatever] and draws 2.76%, etc.

In fact:
OF ALL AUSTRIAN PLAYERS: Vie HOLD, Tri-Alb, Bud-Ser makes up 2.75%
OF AUSTRIAN DRAWS/WINS: Vie HOLD, Tri-Alb, Bud-Ser makes up 2.76%
OF AUSTRIAN WINS: Vie HOLD, Tri-Alb, Bud-Ser makes up 3.07%

Re: First Move Tendencies on PlayDip

PostPosted: 18 Jan 2014, 09:33
by QuiritatioSusurratori
Ah That makes sense. Of course. Im sorry for not understanding immediately. So the statistically best openings are Those in which the percentage of austrian results is bigger than That of austrian plays.

Re: First Move Tendencies on PlayDip

PostPosted: 18 Jan 2014, 15:56
by asudevil
QuiritatioSusurratori wrote:Ah That makes sense. Of course. Im sorry for not understanding immediately. So the statistically best openings are Those in which the percentage of austrian results is bigger than That of austrian plays.


EXACTLY

Re: First Move Tendencies on PlayDip

PostPosted: 20 Dec 2016, 17:16
by AngelEyesGBU
Not surprisingly, it doesn't appear that players making impossible opening moves (i.e. ITALY F NAP=> TUN) :roll: ever participate in wins or draws. :lol:

Re: First Move Tendencies on PlayDip

PostPosted: 21 Dec 2016, 02:27
by Jegpeg
AngelEyesGBU wrote:Not surprisingly, it doesn't appear that players making impossible opening moves (i.e. ITALY F NAP=> TUN) :roll: ever participate in wins or draws. :lol:


Mind the bump.

Actually the OP said that 0% means less than 1% so there might be 1 or two players opening Nap - Tun that manages to participate in a win or draw. We also know that 1.47% of Russian players who get a win or draw open with StP(sc) to Swe.

Re: First Move Tendencies on PlayDip

PostPosted: 26 Mar 2017, 10:57
by Zubb
The data is interesting, but I feel like the comparison done regarding move frequencies is not entirely correct - comparing absolute figures in percent by subtraction doesn't give us relevant information and is hard for interpretation.

With rare (or even just more uncommon) opening, which have smaller absolute numbers, the large proportional difference will look smaller than a minor proportional difference in popular (high %) moves.

Example:
Trieste MOVE Albania: 65.43% / 77.15% / 11.72%
Trieste HOLD: 13.3% / 9.11% / -4.19%
Trieste MOVE Adriatic Sea: 10.62% / 5.23% / -5.39%
Trieste MOVE Venice: 8.02% / 7.44% / -0.58%

First look says that moving to Albania is statistically the most important thing - 11.72% seem big compared to other figures. However if you compare 77.15% (win+draw frequency) and 64.53% (total frequency) you will find out that successful Austria players move to Albania ~20% more often than usual. However, if you do the same for moves to Adriatic sea you will find out that successful Austrians move there TWICE less often.

So, my point is - if we are trying to answer the question which moves are "better" we should be looking at comparative and not absolute difference in usage in successful/unsuccessful games. This is even more true for openings - there are some uncommon openings that never end well for the country trying them and some of the more popular ones give you significantly better odds to win/draw.