Objectives
- Uses hardest difficulty
- Heavy luck manipulation
- Genre: Board
Difficulty
Skill Level 3 was selected as it is the hardest one available in this Backgammon port. It was selected too fast for the viewer to see so the level can be confirmed by viewing WRAM address 0x07D8.
Strategy
The best theoretical strategy would be to maximize our roll values while simultaneously forcing the AI to take the shortest turns possible. So how do we do that? By giving it nowhere to move. We should therefore force the AI on its first turn to blot (i.e., leave a single checker at a point), where we will then hit it. After, we would force the AI to make rolls so abysmal (i.e., rolls whose values on the dice are blocked by white's checkers at the corresponding point) that it will never enter back into the game. The more home points we close (i.e., fill with at least 2 checkers), the more likely the AI will be unable to move, saving some intensive RNG manipulation and therefore lots of time!
Cursor Movements
You're able to move the cursor around after the computer has finished their turn and before you roll the dice. In theory, it would be fastest to move the cursor over to the point where you're planning to move checkers from while waiting for RNG to advance. You'll find if you analyze the input roll for this TAS that I don't always do this optimally, as cursor movements can sometimes change RNG.
Game Summary and Q/A
| Turn | White | Black | Commentary |
|---|
| 1 | 5,6 | 2,6 | White consolidates six checkers on their 13-point. Black is forced to make a blot with their fatally unlucky roll—this is already their last move of the game!
|
|---|
| 2 | 6(x4) | 6(x4) | We must roll double 3's or 6's here to hit black's blot on the 18-point, hence the extremely long wait to manipulate the right outcome. We also close off the 2-point to increase the likelihood black can't enter back into the game.
|
|---|
| 3 | 6(x4) | 2,6 | Best roll possible.
|
|---|
| 4 | 5(x4) | 3,6 | We can't roll double 6's this whole TAS as any of our checkers situated on the 18-point would be blocked by the computer's checkers on the 12-point. Same goes for our checkers on the 7-point running into the opponent's on the 1-point. The next best thing is to roll double 5's to bypass this blockage. The three point is also now closed.
|
|---|
| 5 | 5(x4) | 2,6 | Same as previous turn.
|
|---|
| 6 | 3(x4) | 3,4 | We can't start bearing off checkers (i.e., moving them off the board) until all our checkers are within our home court (i.e., within the first six points). We can accomplish this with just double 1's but in theory, double 5's would make the bearing off process easier in the long run. The reason we settled for double 3's is so we could close off the 4-point, which will be important for black's 9th turn.
|
|---|
| 7 | 3(4x) | 4,6 | All checkers on our 3-point are beared off.
|
|---|
| 8 | 6(4x) | 2,4 | All checkers on our 6-point are beared off.
|
|---|
| 9 | 2(4x) | 4(4x) | All checkers on our 2-point are beared off. Black rolls extremely unlucky double 4's which are not only a 1-in-36 chance of rolling, but is the only roll which prevents black from entering back into the game. This was the whole reason we had to compromise on move 6 for double 3's.
|
|---|
| 10 | 5(4x) | — | White bears off the three remaining checkers from their 4-point, winning the game. |
|---|
Q: Why not roll double 6's on turn 1?
A: Some backgammon rules demand that any doubles rolled on the first roll of the game be discarded so as to not give the player going first an unfair advantage. This backgammon port implements this rule so it is best to roll the highest non-double value of 5,6 instead.
A Note to the Judge
Psst! If you like judging TASes of Backgammon games, I also have
a similar submission from 2.5 years ago I made for ABPA Backgammon on the Intellivision, which is in need of re-judging. It was unfortunately rejected due to the rule at the time stating TASes of real-world board-game adaptations were ineligible for publication. Nowadays with the rule repealed, the submission is easily a candidate for publication, and I'd be very grateful to anyone who finally gets it published!
nymx: Claiming for judging.
nymx: Well, this run absolutely crushes the WR, which is approximately 4 minutes. With all the RNG manipulation, I'm not surprised that it went this well. Great job.
Accepting.
