odiousgambit's Blog

Chess: 1850 down to Kings and Pawns

I am fairly sure I have never beaten the 1850 at level 6 if we start from the beginning moves; I do remember being able to beat level 5 occasionally. I was playing more chess then and at my best. Level 5 was pretty fun, and I’d say level 6 is the last practical level for play. The computer will take an average of 2 minutes per move at L6; that includes when the best move is obvious, so 2 minutes, often more, can feel like an eternity.

I’m not there now for sure. Here I am able to win at level 1, showing the vulnerability of the computer if it can be kept from getting ahead in pieces. At level 1 it may also have left itself at a bad place when all that’s left are Kings and Pawns. At move 32 I feel that White should win, having the more advanced pawns; it’s how to beat the 1850, keep from getting behind on pieces and play a better end game. This is further tested below.

1.e4 c6
2.d4 d5
3.e5 Bf5
4.Nf3 e6
5.Bd3 Bb4+
6.Bd2 Bxd2+
7.Qxd2 Ne7
8.O-O O-O
9.Bxf5 Nxf5
10.Nc3 Nd7
11.Rad1 Qe7
12.Qd3 Nb6
13.b3 Qb4
14.a4 c5
15.Ne2 cxd4
16.Nexd4 Nxd4
17.Qxd4 Qxd4
18.Rxd4 Rfc8
19.Rc1 Rc3
20.Rd3 Rac8
21.Rxc3 Rxc3
22.Kf1 Kf8
23.Ke2 Ke7
24.Kd2 Rc7
25.Kd3 Kd7
26.Nd2 Kc6
27.c4 dxc4+
28.Nxc4 Rd7+
29.Ke4 Nxc4
30.Rxc4+ Kb6
31.Rd4 Rxd4+
32.Kxd4 Ka5
33.Kc5 h6
34.Kd6 Kb4
35.Ke7 Kxb3
36.Kxf7 Kxa4
37.Kxe6 Kb4
38.Kd5 a5
39.e6 a4
40.e7 b5
41.e8=Q Ka5
42.Qa8+ Kb6
43.Qc6+ Ka7
44.Qxb5 g6
45.Qxa4+ Kb6
46.Kd6 g5
47.Qb4+ Ka6
48.Kc7 h5
49.Qa4#

At move 32 I change the 1850 to level 6. Now, still playing White, and blundering at move 42, I am able to prevail in a somewhat different game; it takes 49 moves again. This suggests the end game did not improve for the 1850 at level 6 over level 1:

32.Kxd4 Ka5
33.Kc5 h6
34.Kd6 Kb4
35.Ke7 Kxb3
36.Kxf7 g5
37.Kxe6 Kxa4
38.Kf7 b5
39.e6 h5
40.e7 h4
41.e8=Q a5
42.Qe4+ b4
43.Ke6 Ka3
44.Qb1 g4
45.Kd5 b3
46.Kc4 b2
47.Kc3 Ka4
48.Qxb2 h3
49.Qb3#

Then I switch and I play Black, level 6, starting with move 32 again. White [now the 1850] at level 6 takes 60 moves to win. Like a mother fretting over all her babies, it just can’t see that it has the game wrapped up by ignoring those pawns. Then, after Queening, it can’t see it doesnt need to scarf up any more pawns or need any other Queens.

32.Kxd4 Kc6
33.b4 b5
34.axb5+ Kxb5
35.Kc3 Ka4
36.Kc4 Ka3
37.Kc5 Kb3
38.h3 Kc3
39.g3 Kd3
40.f4 Ke4
41.Kd6 Kf3
42.Ke7 Kxg3
43.Kxf7 Kxf4
44.Kxe6 g5
45.Kd5 Kf5
46.e6 g4
47.e7 gxh3
48.e8=Q h2
49.Qh5+ Kf6
50.Qxh7 h1=R
51.Qxh1 a6
52.Qf1+ Ke7
53.Qxa6 Kf7
54.b5 Ke7
55.b6 Kf6
56.b7+ Kf5
57.Qc4 Kf6
58.Qg4 Kf7
59.b8=Q Kf6
60.Qf8#

I test my last statement about not needing any more Queens by playing White again vs. level 6 at move 49. I am able to get checkmate on move 59 ... perhaps the conundrum caused by the Black pawn potentially putting the White King in check could have been avoided, making mate possible by move 58.

49.Qe5+ Kg4
50.Qxh2 h5
51.Ke5 Kf3
52.Qh3+ Ke2
53.Ke4 Kd2
54.Qd3+ Kc1
55.Qe2 Kb1
56.Kd3 a5
57.bxa5 h4
58.Kc3 h3
59.Qb2#

Chess: Improving

I feel I should point out again that the games can be followed with an analyzer - otherwise these posts are just a bunch of drivel - nobody is going to try to replicate the game manually for somebody at my level. With an analyzer, it's available in a flash.

I like http://www.chess.com/analysis-board-editor .... you click on the "paste PGN" and copy and paste in the moves.

This is the first game I've won against the 1850 at level one - that doesnt mean I won against an 1850-rated player but something unknown, maybe 1100-1400 ? I insisted on the beginning move, e4 e5, otherwise it was rule-enforced including touching a piece. I can't remember if I took White out of book*

Improvement has come partly from being reminded of necessary precautions in play, such as examining the possible actions of each enemy piece each contemplated move, and the extra caution needed for pieces advanced to the 4th rank or beyond [which caused the embarrassing blunder in my first post in the thread]. It all sounds simple, but when I am rusty I don't have the discipline.

The computer never resigns, so I played it out to checkmate to make member Jbyrd happy LOL.

I am Black

1.e4 e5
2.Bc4 Nc6
3.Nf3 Nf6
4.Nc3 Be7
5.d3 O-O
6.O-O d6
7.Bg5 h6
8.Be3 b6
9.Qe2 Bb7
10.Nd5 a5
11.Nxe7+ Qxe7
12.Kh1 Ba6
13.Bxa6 Rxa6
14.d4 Raa8
15.d5 Nd8
16.Qc4 Nb7
17.c3 Nc5
18.Bxc5 bxc5
19.Rfe1 Qd7
20.Re2 Rfb8
21.Rc2 Qb5
22.Qxb5 Rxb5
23.c4 Rb6
24.Re1 Rab8
25.Ree2 Rb4
26.Kg1 Ra4
27.b3 Rab4
28.Kf1 a4
29.bxa4 Rxa4
30.Ke1 Rb1+
31.Kd2 Ra1
32.Kc3 Ra3+
33.Kd2 R1xa2
34.h4 Rxc2+
35.Kxc2 Ra2+
36.Kd1 Rxe2
37.Kxe2 Nxe4
38.Nd2 Nxd2
39.Kxd2 f5
40.Kc3 Kf7
41.h5 g6
42.hxg6+ Kxg6
43.Kd3 Kg5
44.Ke3 Kg4
45.Kd3 h5
46.Kc3 h4
47.Kd3 Kf4
48.Kc3 Ke4
49.Kc2 f4
50.f3+ Ke3
51.Kc3 Kf2
52.Kd3 Kxg2
53.Kd2 h3
54.Kd3 h2
55.Kc3 h1=Q
56.Kd2 Kxf3
57.Kc2 Qe1
58.Kb2 Qb4+
59.Ka1 Ke3
60.Ka2 Kd2
61.Ka1 Kc2
62.Ka2 Qb2#

*the computer knows extensive 'opening book' but it is possible to make it play without using this knowledge - this helps

Chess Cont.

It occurs to me that I should not flatter myself to think anyone would be particularly interested in continuing to analyze my chess, or anyone else's at my level, so I will switch to blogging. It's what a blog is for I think. That's not to say I wouldn't be very happy to see any comments!

I won't repeat what I put in the thread,

http://wizardofvegas.com/forum/off-topic/off-topic/21336-chess-unexplained-weakness-of-d4-d5-e3/

except my last PGN posting, where I am starting to play a little better [thought I would win], but the Radio Shack 1850 shows again what it can do even at low levels.

1.e4 e5
2.Nf3 Nf6
3.Nc3 Nc6
4.Be2 Bc5
5.O-O O-O
6.b3 d5
7.exd5 Nxd5
8.Nxd5 Qxd5
9.c4 Qe4
10.Bb2 Nd4
11.Nxd4 exd4
12.Bf3 Qf4
13.g3 Qe5
14.Re1 Qd6
15.Qc2 d3
16.Qc3 Qg6
17.Be4 Qg4
18.Bxd3 Be6
19.Be4 c6
20.Qe5 Bb6
21.h3 f6
22.hxg4 fxe5
23.Bxe5 Bxf2+
24.Kg2 Bxe1
25.Rxe1 Bxg4
26.Bf4 Rae8
27.d3 Kf7
28.Kf2 h5
29.Re3 Bd1
30.Bf3 Rxe3
31.Bxe3 Ke6
32.Bf4 Bxf3
33.Kxf3 g5

white resigns
_

I also take Jbyrd's advice and played out the first PGN I posted by switching sides at move 22. It probably is true there is something to be learned by switching sides instead of resigning in some cases. In this case, I blunder again at move 27 as Black [stupidly failing to realize the White Queen can take a pawn and prevent disaster by putting me in check], but Black [now me] wins in the end. It shows the 1850's weakness is the end game [bear in mind it's playing at level one here though]

1.d4 d5
2.e3 Nf6
3.Nc3 Nc6
4.Nf3 e6
5.Bb5 Bb4
6.Bxc6+ bxc6
7.Bd2 O-O
8.O-O Bb7
9.a3 Ba5
10.Na4 Bxd2
11.Qxd2 Ne4
12.Qd3 Qd6
13.Rfe1 Rfb8
14.b4 Qe7
15.Qb3 a5
16.Nc5 Nxc5
17.dxc5 axb4
18.axb4 e5
19.Rxa8 Rxa8
20.Qb2 e4
21.Ne5? f6
22.Ng4 h5
23.Nxf6+ Qxf6
24.c3 Kf7
25.Ra1 Rxa1+
26.Qxa1 Qd8
27.Qa5 Qa8
28.Qxc7+ Kg6
29.Qd6+ Kh7
30.Kf1 Qa1+
31.Ke2 Ba6+
32.Kd2 Qb2+
33.Kd1 Bd3
34.f4 Qc2+
35.Ke1 Qe2#

Reserecting EV of Crap/Slots

Reserect's blog got me thinking.

Taking the Wizard's example for slot SD [8.74 see link] and assuming the the slot return is 93% [which I am guessing is fairly middlin'], and looking at 2 SDs of variance, I wanted to look at the multiple of EV this could be in Slots and compare it to Craps.

If a slot player's EV was -$500 and each bet was $5, playing such a slot, we know then that he bet about $7145 dollars in 1429 pulls [his EV would be -$500.15 actually]

Sq Rt of 1429 = 37.8 approx

37.8 * 8.74 = 330.37 units
2 SDs would be 660.74 units or $3303.7 , let's say $3305 if never betting other than $5

if running bad by 2 SDs, that would mean player lost $3805 [adding above to EV]

That's 7.6 times EV approximately; that same EV is about 13% of 2 SDs running bad, to look at it another way.

_

If a Craps player's EV was -$500 at 3x4x5x playing passline with full odds only, then at 0.374% HE he put about $133,689.8 at risk. SD for one bet is about 4.9.

Average bet is 3.78 units* and let's say it was a $10 flat bettor. That'll mean the player made about 3537 bets. Sq rt of that is about 59.47.

59.47 * 4.9 = 291.403 units, about $2914 dollars

2 SDs $5828; running bad means player lost $6328, that's about 12.7 times EV of -$500. That $500 is about 7.9% of 2 SDs running bad, to look at it another way.

_

So I conclude that a Craps player sticking with low HE and high variance can look at his EV and have his real results run wildly off.

A Slots player, always stuck with high variance, still gets a better handle looking at EV, which is not to say running bad won't clobber him.

It's somewhat subjective to conclude this.

I am used to accepting 20-40 times EV with Craps in short sessions. For 3500 bets, however, that kind of variance will not hold up. In the case of slots, 20-40 times EV just isn't going to happen. So it is another indication that a player needs to go by standard deviation to get a handle on his 'swings'

*saved info from Alan Shank, our Goatcabin, however I can no longer find the post that says this.

Wizard's link: http://wizardofodds.com/games/craps/basics/#toc-TheOdds

disclaimer: I am perfectly capable of screwing up the above math.

edits

Back to the Greenbrier part 3

Of course there was no way I wasnt going to play Craps when I have a chance.

Still 3x4x5x there, but $5 min is much more commonly available than before, at least during the week.

After having given it a lot of thought, I decided that I would try to make the transition from someone who has his free odds off during the come-out to someone who left them on. I was always the former player type for the usual reason, but also because I hadn't comfortably found a way to communicate the latter style to the dealer. It especially seemed kind of nuts to me to have to "train" each dealer in the rotation to do what I wanted. Finally it occurred to me that the thing to do is just be sure to say "always on, please" the very moment you give the chips to the dealer. What a simple expedient, yet it escaped me for the longest!

Well, "odds always on" sure adds excitement to the game. It is a real smack in the face when the dice do the nasty on you on the come-out and a real thrill when your number is hit instead. It's a good thing, though, that the Wizard has so confidently declared it is the thing to do, as I think it would otherwise be in the column of selective memory giving the wrong answer. In fact I think for the 2 sessions I played, I came out worse for the change in strategy. It makes such an impression though to lose those odds on the come-out that it's possible I broke about even and don't even know it.

One thing for sure, I couldn't seem to prosper with my come betting. The regular line bet worked out pretty good for me, but the come betting wasnt working. I broke even in the first long session, maybe 3 hours, but on the last session we had a lot of ladies shooting and I'm afraid I bought in heavy on the idea this would bring good luck. Instead we were treated to "girl can't catch a break" shooting. I took my beating cheerfully but the trend was finally too much for me and I colored up down $200 in a short 90 minute or so session.

I came away pretty convinced my new strategy was going to lower HE [as promised by the Wizard] while simultaneously increasing variance. Nice combination, but I got to experience the fact that variance is a double edged sword yet one more time!

Still a lot of fun. That Greenbrier Craps table has given me a lot of sessions that break even, but I've lost plenty of times and won a modest amount maybe once.

Armed with my new strategy though, I've got to give it another go one of these days!