I tested the script with the 6-man (81 GB) vs the 3-4-5 man (200 - 100 MB) of data inputted into the StockFish engine itself by using the (roblox-cheat-engine) and found a few problems. But a few good things
Good:
Improved the positioning sometimes of the moves with the 6man and 3-4-5 man syzygy only by a slight margin, but not much.
Doesn't make unnecessary moves that often.
Does know some openings theory that improved on and used strategies against lower to mid elo people.
Bad:
Does not know "when" to castle from certain danger and keeps instead moving the king into the center of the board making an inaccuracy.
Keeps on doing "okay moves" on winning postionings and failing to find the best ones on "depth of 30" (tested with 6-man and 3-4-5 man)
When playing as white or black you have to manually play your opening you like to choose. Stockfish keeps on trying to avoid openings "you" want to try and play that are much better.
Failing to find potential forks where you can get pinned.
Overall:
Sometimes the script will allow you to play the best moves most of the time against unexpected players that are near lv20 or higher depending on what they play. However, you have to manually castle at the right time when it's potential.
Using the 6-man or 3-4-5 man does help, however. It did not learn better openings. (Queens Gambit accepted or declined, Chigorian, Bird Defense, Catalan, etc). It rarely plays the London or even ruy lopez, but in a few games it did and wins for you sometimes. It mostly plays four knights, or the scotch which is mostly sometimes drawish depending on what you do.
The engine itself plays like a 1800 to 2300 only. It does not feel like it's at it's maximum at all unless you turn on discard thinking time. Allowing it to think more aka ponder before making a move (which will take awhile and catch you suspicions quite often.)
I tested the script with the 6-man (81 GB) vs the 3-4-5 man (200 - 100 MB) of data inputted into the StockFish engine itself by using the (roblox-cheat-engine) and found a few problems. But a few good things
Good:
Improved the positioning sometimes of the moves with the 6man and 3-4-5 man syzygy only by a slight margin, but not much.
Doesn't make unnecessary moves that often.
Does know some openings theory that improved on and used strategies against lower to mid elo people.
Bad:
Does not know "when" to castle from certain danger and keeps instead moving the king into the center of the board making an inaccuracy.
Keeps on doing "okay moves" on winning postionings and failing to find the best ones on "depth of 30" (tested with 6-man and 3-4-5 man)
When playing as white or black you have to manually play your opening you like to choose. Stockfish keeps on trying to avoid openings "you" want to try and play that are much better.
Failing to find potential forks where you can get pinned.
Overall:
Sometimes the script will allow you to play the best moves most of the time against unexpected players that are near lv20 or higher depending on what they play. However, you have to manually castle at the right time when it's potential.
Using the 6-man or 3-4-5 man does help, however. It did not learn better openings. (Queens Gambit accepted or declined, Chigorian, Bird Defense, Catalan, etc). It rarely plays the London or even ruy lopez, but in a few games it did and wins for you sometimes. It mostly plays four knights, or the scotch which is mostly sometimes drawish depending on what you do.
The engine itself plays like a 1800 to 2300 only. It does not feel like it's at it's maximum at all unless you turn on discard thinking time. Allowing it to think more aka ponder before making a move (which will take awhile and catch you suspicions quite often.)