What do you mean "promised" results? They are just statistics.
Can you trust the stats in openings explorer?
Sort:
For an example a line can score very well for a long time and get good statistics. If someone refutes it and the line is not played anymore it will still keep its good stat
Feb 8, 2012
0
#4
tigergutt wrote:
For an example a line can score very well for a long time and get good statistics. If someone refutes it and the line is not played anymore it will still keep its good stat
That doesn't make the stat wrong.
mark100net wrote:
What do you mean "promised" results? They are just statistics.
I assumed that the the W/D/B stats are taken from the actual game database, therefore how can a move show as 50% draws before you play it and zero draws after its on the board?
I agree i just dont think he should worry to much on stats. For example 1.e4 d5 2.d4 scores 61% for white
[COMMENT DELETED]
Vease wrote:
mark100net wrote:
What do you mean "promised" results? They are just statistics.
I assumed that the the W/D/B stats are taken from the actual game database, therefore how can a move show as 50% draws before you play it and zero draws after its on the board?
Give the sequence of moves, so that we can see what you are talking about.
It might be that you are misreading the database.
Feb 8, 2012
0
#9
Vease wrote:
mark100net wrote:
What do you mean "promised" results? They are just statistics.
I assumed that the the W/D/B stats are taken from the actual game database, therefore how can a move show as 50% draws before you play it and zero draws after its on the board?
I think maybe you are misinterpreting the statistics. 50% draws only shows how many draws have resulted from the current...not every position that could result suqsequently.
Once you go to another position, the statistics change to those resulting from the new position (or go away completely if there are no games in the database reaching your new position).
I also agree with tigergutt...if you are paying this much attention to the stats it may be too much. I don't think they are relevant at all when there are only a handful of games containing the current position. At that point you need to play the games out and see where the game was actually lost.
Numbers is the best way to lie.
You can't trust statistics and you can't learn an opening or pick your lines from databases.
Don't do that , it's simply wrong.
Only if there is a good amount of trials. If an opening was played once and has a 100% win rate, it might be terrible. But if it was played 7000 times with a 60% win rate, it's excellent
You are misreading the database. The grey bar shows draws not wins for White.
Michael-G wrote:
Numbers is the best way to lie.
You can't trust statistics and you can't learn an opening or pick your lines from databases.
Don't do that , it's simply wrong.
Absolutely, I agree entirely. I am trying to learn the 4.Nd7 line of the Caro Kann for Black and reached this position after a quick playthrough. These are not all the most popular options for either side, but its the line I was interested in. I just happened to notice that there were no drawn games in the database after 11.Ndb5 for black but on 11.Bc2 for White the stats show that 50 percent of the games are drawn. I presumed this was for Ndb5 because the stats for each move in the tree are different.
[COMMENT DELETED]
ozzie_c_cobblepot wrote:
Can you post the link to games explorer?
Solved. See two up.
Ugh, what a fool..Sorry folks, I'm sure I saw something different..anyway it still looks promising , but White has to cooperate...
I have just played through a line that I was interested in and at move 11 for Black if he plays Ndb5 the stats show W 12.5% D 50% B 37.5%..who wouldn't look at this for Black? Actually enter the move though and the stats show white wins 62.5% and Black 37.5%.
I have noticed in other lines where the resulting numbers don't seem to match up with the 'promised' results...Am I missing something here or is it a real problem?