by Snowie 4 Pro
5 April 2011
Snowie 4 Pro analyzes the first of two 11-point Final matches from the 2011 ABT Ohio State Backgammon Championships played between John O'Hagan and Greg Merriman.
There are 13 games in this match. Game 1 starts below.
Thanks to Mary Hickey for providing this match file.
In case you lose your place or do not have time to view the entire match in one session, bookmark this link now - GV Bot Section - you can always pick up from where you left off. That's the link to our full archive of analyzed and annotated matches, listed by date.
If you have never seen an analyzed backgammon match, look below each board diagram where you will find Snowie's evaluation of the position. The move marked with an asterisk, and highlighted in yellow in the list, was the one made by the player on roll. For more details on how to read and understand Snowie analysis read the following few paragraphs taken from the Help section of the Snowie program, then scroll further down to begin viewing the match. Enjoy!
|* 4||2||13/9||0.091 (-0.024)|
|0.6% 15.7% 53.1% 46.9% 12.8% 0.5%|
The first line shows the following information:
|* 4||rank of the given move (* before the rank indicates that it is the played move)|
|2||ply of the evaluation|
|0.091||cubeful equity of the move|
|(-0.024)||difference of equity with best move|
The second line (hidden if the Detail Mode has not been checked) displays
the complete cubeless evaluation of the move:
|Player's wins||Player's losses|
If you check the Evaluation Parameters, you get additional information
about how the evaluation was computed.
|Cube action equity|
|0.6% 19.4% 66.9% 33.1% 6.7% 0.2%|
|Proper cube action: No double||5%|
The first two lines display the cubeless evaluation of the position. The evaluation can be of type 1-Ply, 2-Ply, 3-Ply, Mini-Rollout, Rollout or Database. Below the cubeless money equity of the position, the detailed probabilities are also displayed. They are in the same order as the detailed line for the checker play.
The next section displays the cubeful equities of the possible cube actions. From the point of view of the player in action there are two possible actions: Double or No Double. If the player in action decides to double, then there are two options for his opponent: either take or pass the offered cube. Therefore, there are three possible actions which can be played: No double, Double, take or Double, pass. The first line corresponds to the correct action for both sides. The second line displays the line which corresponds to the wrong double/no double action but with the correct response of the opponent. Finally, the last line corresponds to the equity you would have if your opponent would do the wrong take/pass decision to an offered double.
In the example above the cubeful equity of not doubling is 0.711 pts/game. Now, if you offer a double and your opponent correctly takes the cube you will only earn 0.696 pts/game on average and finally if you double and your opponent passes you will earn 1 pt. Of course your best equity is the 1 pt you would win if you double and your opponent passes but since you have no control on what your opponent will do, the best action here is not to double so that you secure an equity of 0.711 which is better than the equity of 0.696 you have if you double.
As we have seen, the best equity you can earn here is the 1 pt if your opponent passes. Suppose you have the suspicion that your opponent might pass but you are not sure. Is Snowie able to tell whether you should whip it anyway to profit from his occasional incorrect passes? Yes, Snowie is able to tell you that. Snowie gives you a borderline frequency at which your opponent is supposed to pass to make the theoretically incorrect double practically correct. In our example Snowie says that this frequency is 5%. Therefore, the double becomes correct if you think there is a 5% chance that your opponent passes the cube!
|The rest of this article (410.20 K) is premium content. Please subscribe below.|