I’ve finally completed a 10 game match I’ve been wanting to do for some time between Millennium’s TheKing element set to play with it’s “Normal” personality and TheKing with the “King of Alabama” personality settings created by Spacious Mind. I know “King of Alabama” won in the Aegon 1997 Revisited tournament (played in 2019, how time flies!) achieving an ELO of 2761. This is extraordinary in my opinion, and all participants did well.

The rating achieved in the Aegon 1997 Revisited tournament was at a time control of 40/40, so an average of 1 minute per move. I have been interested to see how well TheKing element’s Normal personality would perform against King of Alabama at a time control of 40/120. Would this longer time control yield more draws or would one machine dominate?

Here are the King of Alabama settings:

PLAYER
PAWN:81
KNIGHT:86
BISHOP:86
ROOK:89
QUEEN:91
SPACE:91
MOBILITY:98
KING SAFETY:81
PASS PAWN:121
PAWN WEAK:81

OPPONENT
PAWN:81
KNIGHT:86
BISHOP:86
ROOK:89
QUEEN:91
SPACE:91
MOBILITY:98
KING SAFETY:121
PASS PAWN:81
PAWN WEAK:81

SELECTIVITY:14

For TheKing “Normal”, all settings are 100% except SELECTIVITY which is Auto.
I left Contempt and Randomness at their defaults for both personalities.

I wasn’t sure of The King Element’s actual ELO rating at a time control of 40/120 so looked it up on https://www.schachcomputer.info/html...elo_liste.html
Here, tournaments are conducted at 40/120 and TheKing has a rating of 2594 after 74 games. I’m guessing this was with it’s “Normal” personality.

Game 1 of 10


The Sicilian, Taimanov/Sveshnikov opening was played.

HIARCS15 Analysis:
White: 1 mistake and 7 inaccuracies.
Black: 2 blunders and 9 inaccuracies.

The first dubious move made from the opening was by King of Alabama(KingA). In the following position King Normal (KingN) had just moved 10.Na3:

Here, KingA moved 10…d5?! This loses the pawn to White’s Knight on c3 and Black’s position quickly deteriorates. Strange to give away a Pawn especially as there did not seem to be any positional advantage.

Neither personality was perfect but in this game, KingN came out of the opening a Pawn up and more or less maintained its advantage all the way. The blunders made by KingA were in the end-game where it was going to lose anyway so I acedemic at this point.

Out of interest, at move 69 KingA announced it would be mated in 12, KingN announced mate in 13, but it actually took 17 ply moves.

So the first game goes to KingN.

I will post the next game shortly.

Game 2 of 10



The Queen’s Pawn, Franco-Indian (Keres) Defence was played.

HIARCS15 Analysis:
White: 1 mistake and 3 inaccuracies.
Black: 1 mistake and 6 inaccuracies.

The first draw in the match. I adjourned the game at move 70 at the following position:


Both Kings were consistently showing an eval of 0.00. Did not see any need to carry on shuffling pieces back and forth.

Current score 1½ - ½ to KingN.

Game 3 of 10



The Sicilian, Chameleon variation was played.

HIARCS15 Analysis:
White: 8 inaccuracies.
Black: 5 inaccuracies.

Up until move 31 (see diagram) both players were about equal.


However, KingN made the following dubious move: 32 Qf4?!
Surely it mis-calculated the target square g5 attacking the troublesome Knight on h4?

From here, KingA was keeping its head above water but did not play the endgame so well and lost it’s slight advantage with the final result being a draw.

At move 67 I adjourned the game as both players were consistently showing an eval of 0.00.

Current score: 2 – 1 to KingN.

Game 4 of 10


The English, Bremen was played.

HIARCS15 Analysis:
White: 2 blunders, 2 mistakes and 5 inaccuracies.
Black: 2 inaccuracies.

In this game there was the usual fractional evaluation swings from one side to the other until the end game where KingA was totally outplayed.

Here is the position after Black moved his Queen 69…Qa4:

An interesting position and one in which White needs to be very careful! Unfortunately, KingA makes the blunder with the move 70.Qxa4? White cannot now stop Black from Queening with passed pawns on both sides of the board. A couple of moves on and KingA’s evaluation is -0.25 looking 06|15 (6 moves brute force and 15 moves selective) ahead. KingN’s evaluation is more accurate at 1.49 with lookahead of 12|13 (12 moves brute force, 13 selective). In fact, it’s my strong belief that KingN’s more brute force approach in this end game has been more beneficial. KingA’s more selective search has maybe missed its nuances.

Current score 3 – 1 to KingN.

Game 5 of 10


The English, Nimzowitsch, Flohr variation was played.

HIARCS15 Analysis:
White: 2 mistakes and 2 inaccuracies.
Black: 2 mistakes and 5 inaccuracies.

This game was fairly even for both sides until about half way through. See the following position:

Here, White had just played 33.Ra1 and Black has two passed Pawns marching down the board. His Pawn on a3 is the most troublesome. From here Black played 33…Ra5 which is a mistake according to H15. The better move is 33…a2 The plan then is for Black to move it’s Knight and Rook down forcing White to defend and ultimately losing it’s Knight in the process. I don’t think this is necessarily beyond TheKing’s ability to see, but maybe the more selective searching of KingA missed this particular variation?

Looking a few moves on in the following position, White had just played 39.Kb2:

KingN is now in a better position and a pawn up. It also probable Black’s Pawn on a2 will be lost at some point. KingA’s evaluation was -0.7 whilst KingN’s was 1.52 which is the closer value to H15’s 2.18. It’s from this point that KingN’s endgame is better.

Here is another example where KingA’s more selective searching may be the cause of missing nuances of an endgame:

KingN as White had just played 47.c5 Marching towards the queening square. Here, KA played 47…g6!? Now, White can advance his Pawn again! Black should have played 47…Kd7. In this position KingN was searching ahead with 9 brute force and 10 selective moves (09|10) whilst KingA was searching with 02|11, so I think this is where the more brute force option selected by KingN has been more beneficial. But this is just an impression.

Current score 4 – 1 to KingN
It’s looking a little one-sided but things could change.

Game 6 of 10


The English, Nimzowitsch, Flohr variation was played again. This makes it interesting to compare with Game 5.

HIARCS15 Analysis:
White: 2 blunders and 5 inaccuracies.
Black: 2 inaccuracies.

This game was not good for KingA. He lost a Pawn early in the game and never recovered.
In the following position Black had just played 40…Kg6:

KingN(Black) is in a reasonably good position and White will have to fight hard. However, KingA blundered with the move 41.Qc3?? What!? An exchange of Queens leaves White with Pawns marching down from both sides of the board with his King caught in the middle.
Play continued 41…Qxc3 42.Kxc3 The expected Queen exchange. 42…Kf5 Black brings his King into position for the fight:

Here, KingN search depth is 16|17 whilst KingA’s is 09|18 but the extra selectivity does not help. Here he blunders again and moves 43.h3?? He obviously does not care about Black’s Pawn on a4. Surely 43.Kb4 would have been better? Now, White cannot prevent Black from Queening.

Current Score 5 – 1 to KingN

Game 7 of 10


Unusual opening played A00: Sokolsky (Polish/Orangutan) and the first win for King of Alabama.

HIARCS15 Analysis:
White: 2 mistakes and 11 inaccuracies.
Black: 1 mistake and 8 inaccuracies.

KingN shows it can also mess up in the endgame. Take this position from the game. Black had just moved 62…Rc5:

White now played 63.Rae1? Possibly thinking of then moving it to e3. However, this give Black time to double his Rooks on the b file and White wont be able to defend against the Rooks and advanced Pawn, as can be witnessed. The better line would have been 63. Rgf1+ Ke4 64.Rf2 Kd4 65. Re5 now White has a double attack on the troublesome Pawn on a2. 65…Rca5 66.Rd2+ Ke3 67.Rb2 Ke4 Giving an equal position:


KingN did not play this endgame so well, but KingA certainly showed what it can do.

Current Score 5 – 2 to KingN

Game 8 of 10


The Sicilian, Nimzowitsch-Rossolimo Attack was played.

HIARCS15 Analysis:
White: 1 blunder, 1 mistake and 7 inaccuracies.
Black: 2 inaccuracies.

In this game, KingA had a slight edge for the first half but then was outplayed for the remainder of the game. In the following position KingN (Black) played 33…h4

The idea being if 35. gxh then White’s in trouble! However, H15 thought 33…Qg6 was better.
White then played 34.Rg1 Obviously concerned with his Pawn on g3, but this takes the Rook away from the center. Better would have been 34.g4.
Play continued with 34…Qh6 (see diagram)

Here, White played 35.Rh1?! This does not serve much purpose, there is still pressure on the g3 Pawn. But, KingN is not without issue. He misses a chance to put White into real trouble:

From this position KingN played 35…Bf7?! Better would have been 35…Bc4 White dare not capture this Bishop! If 36.bxB Rd2+ 37. Ke1 Rgd8 and it’s looking grim. If instead he captures with his Queen then 36.QxB is even worse. 36…Rd2+ and White cannot retreat with 37.Kf1 as it will be mate in 2, so the only option is 37.Qe2 and the outlook is bleak. However, this is academic as things just worsened for White.

Here is an example of poor endgame play this time from KingA:

It’s not looking good for White, he’s going to lose but from this position he moved 52.Kf4? Better would have been 52.Ne4+ Kg6 53.Nf2 blocking off Black’s Rook. However, actual play continued with 52…Rg2 (see diagram):

Now it was White’s move and KingA blundered with 53.Nxa2?? Why sacrifice his Knight? 53.Ne4+ would have given a bit of a fight! At move 66 KingN announced mate in 12, and at move 67 KingA saw that it would be mated in 14. It took 18 ply moves in the end.

Current Score 6 – 2 to KingN

Game 9 of 10


The Reti, Nimzowitsch-Larsen Attack was played and KingA shows what it can do!

HIARCS15 Analysis:
White: 1 mistake and 5 inaccuracies.
Black: 1 inaccuracy.

Up until move 10 in the opening it was fairly evenly matched. See diagram:

Here Black has a slightly better position. White’s Bishop on c3 is under attack from Black’s Knight and Bishop so he needs to handle this correctly. There are a few options here:
If 11.exd5 Bxc3 threatening the Rook 11.Rd1 Nd4 and White’s in difficulty. A better line goes:
11.Bxd5 Bxc3 12.Rd1 Ne5 13.h3 Nxf3+ 14.gxf3 and White has a much better game.
However, KingN opts for 11.Bxg7?! Play continues with 11…Nf4 attacking the Queen. 12.Qf1 the safest as 12.Qd1 maintains Blacks skewer with his Bishop on g4. 12…Kxg7 13.O-O-O Qa5 and Black is in a healthy position. In fact, from here KingA stormed through to a win in the shortest game so far at 58 moves.

I think KingA’s extra selectivity helped secure this game, and from the opening it found the best moves. A good win!

Current Score 6 – 3 to KingN
So, even if KingA wins the next game it wont win this match, but seems to be catching up.

Game 10 of 10


The Queen’s Gambit Declined Slav variation was played and KingA wins back another game in style.

HIARCS15 Analysis:
White: 1 mistake and 3 inaccuracies.
Black: 4 mistakes and 5 inaccuracies.

Both players were about equal up until White’s move 14.fxe4 in this position:


From here, KingN played 14…Rxh2? Gains a Pawn but sacrifices position. Play continued with:
15.Rxh2 Bxh2?! A missed chance to put White into check with 15…Bg3+ causing White to lose castling rights. 16.Bg5 and White’s in a good position with both of Black’s Knights pinned and his Bishop not doing much. White can and does bring his Rook to the centre to allow advancing his Queen’s Pawn. From here it was all down hill for KingN.

Final Score 6 – 4 to KingN.

King Normal: 2 wins as White, 3 Wins as Black and 2 draws
King Alabama: 1 win as White, 2 wins as Black and 2 draws

That gives KingA an ELO rating of about 2524 in this match with a time control of 40/120

Ten games is not enough for a 100% accurate evaluation, but I think it does show that the longer time control has slightly narrowed the ELO difference from the 40/40 time control. From this small set of results I can only give my impression of how the King personalities played. My impression is that King of Alabama seemed not so good in the end game due to it’s higher selectivity possibly missing some of the nuances of the positions. TheKing Normal in the end game always increases the amount of brute force searching. That being said, when The King of Alabama found the correct line it would then play very well. The key point here is the option to turn the Auto selectivity off. I noticed, that TheKing module with Auto Selectivity turned off would still use some brute force searching but always much less than the Auto mode in the same positions. So maybe, and I stress maybe, higher selectivity does not work so well at longer time controls, and thinking when turning off the Auto Selectivity that we are losing some of the magic behind this setting.

It could be interesting to see how King of Alabama fairs with less selectivity or even set back to Auto for games at 40/120. Perhaps this personality would called King Brutus of Alabama?

I think this was a fun little tournament and by no means conclusive of the personalities performance, just possibly a hint.