Prolog Chess AI, Alpha Beta Pruning - prolog

Question

I am relatively new to Prolog. Currently I am trying to do an AI for a Minichess Game, named Diana chess or Ladies chess (https://en.wikipedia.org/wiki/Minichess). It's basically chess but with a 6x6 Board where you don't have a queen and only one knight.
For my AI, I am mainly looking at the book Prolog programming for artificial intelligence by Ivan Bratko (3rd Edition). There he shows an implementation of the alpha-beta algorithm.I am also looking at the Prolog-Code someone posted on GitHub, where he implemented the same code for his Checkers game. (https://github.com/migafgarcia/prolog-checkers/blob/master/checkers.pl)
The code should think a certain number of turns ahead and choose the best move out of them. Before I started programming in Prolog I did a lot oh Java programming, that's why "Zug" in the first line of my Code is the return of NextMove. I hope this all makes some sense.
My Problem is that when I run my code and, for example, I let the AI run against a human (me), when I start as white and do my first move (b2b3), the AI just doesn't work, the backtracking kinda stops right before NextMove and that way I can't get the best move back. Furthermore when the AI thinks for the minimizing player and needs to choose his move, he always takes the move with the highest value, but, when my understanding from the algorithm is not wrong, shouldn't it take the lowest Value?
I am grateful for every comment on this or for any feedback at all.
Here is the code: https://pastebin.com/sc4gLdQe
alphabeta(Player,MaxiPlayer,Alpha,Beta,Board,NextMove,Eval,Depth) :-
Depth<2,
NewDepth is Depth+1,
findall(
(Player,Figur,VonL,VonR,NachL,NachR,Gegner,Board),
zugbewegung(Player,Figur,VonL,VonR,NachL,NachR,Gegner,Board),
Moves
),
write('All Moves:'),
sysoutMoves(Moves),nl,
bounded_best(Player,MaxiPlayer,Alpha,Beta,Board,Moves,NextMove,Eval,NewDepth),
write('NextMove: '),
write(NextMove),nl,
!.
alphabeta(Player,_,_,_,Board,_,Eval,_) :-
value(Board,Eval,Player),
!.
bounded_best(Player,MaxiPlayer,Alpha,Beta,Board,
[(Spieler,Figur,VonL,VonR,NachL,NachR,Gegner,_)|Moves],
BestMove,BestEval,Depth) :-
sysoutMove((Spieler,Figur,VonL,VonR,NachL,NachR,Gegner,_)),
sysoutMoves(Moves),
ziehen(Board,(Spieler,Figur,VonL,VonR,NachL,NachR,Gegner,_),NewBoard),
gegner(Player,NextPlayer),
alphabeta(NextPlayer,MaxiPlayer,Alpha,Beta,NewBoard,_,Eval,Depth),
good_enough(Player,MaxiPlayer,Moves,Alpha,Beta,Board,
(Spieler,Figur,VonL,VonR,NachL,NachR,Gegner,_),
Eval,BestMove,BestEval,Depth).
good_enough(_,_,[],_,_,_,_,_,_,_) :- !.
good_enough(Player,MaxiPlayer,_,_,Beta,_,Move,Eval,Move,Eval,_) :-
minimizing(Player,MaxiPlayer),
Eval > Beta,
!.
good_enough(Player,MaxiPlayer,_,Alpha,_,_,Move,Eval,Move,Eval,_) :-
maximizing(Player,MaxiPlayer),
Eval < Alpha,
!.
good_enough(Player,MaxiPlayer,Moves,Alpha,Beta,Board,Move,Eval,BestMove,BestEval,Depth) :-
new_bounds(Player,MaxiPlayer,Alpha,Beta,Eval,NewAlpha,NewBeta),
bounded_best(Player,MaxiPlayer,NewAlpha,NewBeta,Board,Moves,Move1,Eval1,Depth),
better_of(Player,MaxiPlayer,Move,Eval,Move1,Eval1,BestMove,BestEval).
new_bounds(Player,MaxiPlayer,Alpha,Beta,Eval,Eval,Beta) :-
minimizing(Player,MaxiPlayer),
Eval > Alpha,
!.
new_bounds(Player,MaxiPlayer,Alpha,Beta,Eval,Alpha,Eval) :-
maximizing(Player,MaxiPlayer),
Eval < Beta,
!.
new_bounds(_,_,Alpha,Beta,_,Alpha,Beta).
better_of(Player,MaxiPlayer,Move1,Eval1,_,Eval2,Move1,Eval1) :-
maximizing(Player,MaxiPlayer),
Eval1 >= Eval2,
!.
better_of(Player,MaxiPlayer,_,Eval1,Move2,Eval2,Move2,Eval2) :-
maximizing(Player,MaxiPlayer),
Eval2 >= Eval1,
!.
better_of(Player,MaxiPlayer,Move1,Eval1,_,Eval2,Move1,Eval1) :-
minimizing(Player,MaxiPlayer),
Eval1 =< Eval2,
!.
better_of(Player,MaxiPlayer,_,Eval1,Move2,Eval2,Move2,Eval2) :-
minimizing(Player,MaxiPlayer),
Eval2 =< Eval1,
!.
maximizing(Player,MaxiPlayer) :-
Player == MaxiPlayer.
minimizing(Player,MaxiPlayer) :-
Player \== MaxiPlayer.