Output of an analysis

Chequer play analysis

Cube decision analysis

Game, match, or session statistics

You can get a summary of the analysis from the game, match, or session analysis. The game analysis is a summary for the current game whereas the match or session statistics is a summary of all the games in the match or session. The match analysis is available in the GUI from Analysis->Match Statistics or at the botton of exported files.

Chequer play statistics

This section provides a summary of the chequer play statistics. The following information is available

• Total moves : The total number of moves in the match.
• Unforced moves : The number of unforced moves, i.e., all chequer play decisions which had more than one legal move.
• Moves marked xxx: The number of moves marked very good, very bad etc. The analysis will mark moves very bad, bad etc. based on the threshold you've defined in the analysis settings. gnubg can currently not automatically mark moves, e.g., good, but you can mark moves good yourself.
• Error rate (total): The first number is the total amount of normalised equity that the player gave up during this game or match. The number in parenthesis is the un-normalised counterpart; for money play the all errors are multiplied with the value of the cube, and for match play the total match winning chance given up by the player is reported.
• Error rate (per move): The error rate per move is the total error rate divided by the number of unforced moves. Note that is different from Snowie 4 that defines the error rate per move as the total error rate divided by the total number of moves for both players, i.e., the sum of total moves for both players. In general, your error rate per move will be lower in Snowie than in gnubg.
• Chequer play rating: gnubg will assign a rating for your chequer play ranging from Awful! to Supernatural. See the description for the overall rating below.

Luck analysis

This section provides information about how Ms. Fortuna distributed her luck. The following information is available:

• Rolls marked xxx: The number of moves marked very lucky, lucky etc. Moves marked very lucky are huge jokers that improve your equity with more the +0.6 relative to the average equity. gnubg normally uses cubeful 0-ply evaluations to calculate the luck, but you can change that under the analysis settings. See below for a complete overview of what is considered very lucky, lucky,etc.
• Luck rate (total): The total luck for this game or match reported both normalised and unnormalised.
• Luck rate (per move): The luck rate per move reported both normalised and unnormalised.
• Luck rating: Based on the luck rate per move gnubg will assign you a humerous luck rating. See below for the possible ratings.

Thresholds for marking of rolls:

Deviation of equity from average	Roll is marked
>0.6	very lucky
0.3 - 0.6	lucky
-0.3 - 0.3	unmarked
-0.6 - -0.3	unlucky
< -0.6	very unlucky

Luck ratings:

Normalised luck rate per move	Luck rating
> 0.10	Cheater :-)
0.06 - 0.10	Go to Las Vegas immediately
0.02 - 0.06	Good dice, man!
-0.02 - 0.02	none
-0.06 - -0.02	Better luck next time
-0.06 - -0.10	Go to bed
< -0.10	Haaa-haaa

Cube statistics

This section provides a summary of the cube decision statistics: the number of cube decisions, missed doubles, etc.

• Total cube decisions: The total number of cube decisions, i.e., the sum of no-doubles, doubles, takes, and passes.
• Close or actual cube decisions: Similar to "Total cube decisions", except that no-doubles are only included if they're considered "close". gnubg considers a cube decision close if the relavant equities are within 0.25 from eachother or if the position is too good.
• Doubles, Takes, Passes: The total number of doubles, takes, and passes in the game or match.
• Missed doubles around DP, Missed doubles around TG: The number of missed doubles around the double point and too good point respectively. If the equity is below 0.95 the position is considered to be around the double point else it considered to be around the too good point. If you miss doubles around the double point, it usually means that you double to late, whereas missed doubles around the too good point means that you're too greedy and play on for gammon when the position is not too good.
• Wrong doubles around DP, Wrong doubles around TG: The number of wrong doubles around the double point and too good point, respectively. If you have missed doubles around the double point that generally means that you double too early where you really should hold the cube, whereas missed doubles around the too good point means that you double positions where you really should play on for a gammon or backgammon.
• Wrong takes, Wrong passes: The total number of wrong takes and passes, respectively.
• Error rate (total): The accumulated cube errors for this game or match reported both normalised and unnormalised.
• Error rate (per cube decision): The error rate per cube decision is the total error rate divided by the number of close or actual cube decisions. Note that is different from Snowie 4 that defines the error rate per cube decision as the total error rate divided by the total number of moves for both players. In general, your error rate per cube decision will be lower in Snowie than in gnubg.
• Cube decision rating: gnubg will assign a rating for your cube decisions ranging from Awful! to Supernatural. See the description for the overall rating below.

Overall rating

The last section is the overall summary.

• Error rate (total): The accumulated chequer play and cube decision errors reported by normalised and unnormalised.
• Error rate (per decision): The error rate per decision is the total error rate divided by the number of non-trivial decisions (i.e., the sum of unforced moves and close or actual cube decisions). Again, please note that is different from Snowie 4 that defines the error rate per decision as the total error rate divided by the total number of moves for both players. In general, your error rate per move will be lower in Snowie than in gnubg. An investigation of approximately 300 matches showed the on average the gnubg error rate with be 1.4 times higher than your Snowie 4 error rate.
• Equivalent Snowie error rate: For easy comparison gnubg will also print the total error rate calculated by the same formulae as Snowie 4.
• Overall rating: Based on your normalised error rate per move gnubg will assign you a rating ranging from Awful! to Supernatural. See the table below for the thresholds.
• Actual result: The actual result of the game or match. For money game this is simply the number of points won or lost during the game or match. For match play the number is calculated as 50% added to the result of the game or match in MWC. For example, losing a match or winning a match corresponds to an actual result of 0% and 100%, respectively. Winning 1 point in the first game of 7 point match is worth 6% using Kit Woolsey's match equity table, hence the actual result is 56%.
• Luck adjusted result: The luck adjusted result is calculated as the actual result plus the total unnormalised luck rate. This is also called "variance reduction of skill" as described in Douglas Zare's excellent article Hedging Toward Skill. This should give an unbiased measure of the strengths of the players.
• MWC against current opponent: For match play gnubg will calculate your MWC against the current opponent. The number is calculated as 50% - your total unnormaluised error rate + your opponent's total unnormalised error rate. If your opponent is really lucky but plays rather bad, this number can become larger than 100%, since he due to the extra luck has the possibily to give up even more MWC. However, gnubg will report this number as 100%, and 0% for the opposite situation. Note that this number is biased towards the analysing bot, e.g., a 0-ply analysis a game between gnubg 2-ply and 0-ply will suggest that 0-ply is a favourite, which it is clearly not. For an unbiased measure use the luck adjusted result above.
• Guestimated abs. rating: writeme

Threshold for ratings:

Normalised total error rate per move	Rating
0.000 - 0.002	Supernatural
0.002 - 0.005	World Class
0.005 - 0.008	Expert
0.008 - 0.012	Advanced
0.012 - 0.018	Intermediate
0.018 - 0.026	Casual Player
0.026 - 0.035	Beginner
> 0.035	Awful!