From: Backgammon Galore : Woolsey-Bagai Match

Kit Woolsey vs. Jeremy Bagai
Game 1

Annotated by Kit Woolsey, Jeremy Bagai, and TD-Gammon

Match to 9:   Kit Woolsey - 0,  Jeremy Bagai - 0

Kit Woolsey is playing Red.
Jeremy Bagai is playing White.
Kit wins the opening roll.

Position 1: Red to play 32.

Jeremy: Probably just as good as the more common 13/11, 13/10. Perhaps even better when playing against an intermediate because he may be unfamiliar with the correct replies.

TD: Kit's opening 3-2 is the best choice in my opinion.

    24/21, 13/11      +.014
    24/22, 13/10      -.003
    13/11, 13/10      -.004

Position 2: White to play 54.

Kit: It is important for Jeremy to hit in order to prevent me from making the advanced anchor, particularly since he doesn't have any attractive fours anyway. However the double hit of 8/4*, 6/1* is just too loose. It leaves a man out of play, two direct shots, and a mess to clean up even if it works.

Jeremy: None of the usual opening 5-4's will work: 13/9, 13/8 is too passive and leaves a direct shot; 24/20, 13/8 comes under the gun on the 20 point; 24/15 leaves a double direct shot for little gain. 13/4* starts a point that I want and kicks Kit off a point that he wants. The alternative to consider is hitting twice with 8/4*, 6/1*. This might be right, but I usually avoid such tempo plays unless they have great potential gains.

TD: Correct, but not by much. While hitting an opponent off a key point is valuable, there is a lot to be said for making sure the blot hitting contest is waged on the other guy's side of the table.

    13/4*           -.107
    24/20, 13/8     -.113
    8/4*, 6/1*      -.142

Position 3: Red to play 42 from the bar.

Kit: Locking up the 11 point is clearly better than anything else. You never know when some outfield point will come in very handy later in the game.

Jeremy: Hitting is clear. I used to think that 11/9 was the best deuce in this kind of position. I'd rather make the nine point than the 11 point, and the blot would serve as a builder for five, four, and three points rather than just the five point. Now I make Kit's play. It unstacks the midpoint, bringing another checker into play, and leaves no fly shots. Note that if Kit had already made his five point 11/9 would be correct because the two checkers on the 11 point would be worthless as builders.

TD: Assets are everything. Don't even think of another play.

    B/21*, 13/11      +.212
    B/21*, 11/9       +.087
    B/21*, 24/22      +.064

Position 4: White to play 21 from the bar.

Kit: Jeremy would like to hit me off his four point again, but three men on the 24 point is just sick. You can spend the rest of the game trying to untangle the mess. I think Jeremy's play is better than B/23, 24/23. He would rather make the more advanced anchor of the three point, and since his back men aren't hemmed in yet getting off the 24 point isn't vital.

Jeremy: Three plays: B/24, 6/4*, B/23, 24/23; B/22. Hitting on the four point is good and aggressive but leaves three checkers stacked on the 24 point which is very bad. Making the 23 point is strong ("much stronger than most players realize," as Kit often says) and I'll guess it is Kit's choice. I entered on the 22 point to cover more outfield and try for a more advanced anchor. No play other than these is close.

TD: You unimaginative humans don't even consider the best play, B/23, 6/5 as a candidate! I have been teaching you the importance of playing safe in the opening in many situations, so now the thought of slotting into a double shot never occurs to you. Yet thematically it is the right play. Jeremy already has more men back, so being hit won't hurt him much. Kit has no board, so it isn't particularly dangerous. Jeremy wants to play as flexibly as possible. If he is hit he improves his chances of making an advanced anchor; if not, he can make his five point. In all fairness, Jeremy's actual play was not too far off the mark.

    B/23, 6/5      -.248
    B/24, 6/4*     -.257
    B/22           -.265

Position 5: Red to play 51.

Kit: Hitting loose with 21/20, 8/3* accomplishes little if it works and loses a lot if it doesn't. Running with 21/15 is possible, but just leaves me scrambling next turn when I should be doing constructive things. My play covers his whole outer board, makes a move toward the best advanced anchor, and brings some more firepower into play on the offensive front.

Jeremy: This play is fine although I would choose 21/15. It doesn't strip the midpoint and tries to escape a checker.

TD: Kit is on the wrong theme here. He is ahead in the race, so he should be trying to extricate the back men rather than going after an advanced anchor. His actual play just invites Jeremy to do what he should have done last roll -- slot the five point.

    24/23, 21/16      +.184
    21/15             +.176
    24/23, 13/8       +.165
    21/20, 13/8       +.149

Position 6: White to play 22.

Kit: This is a little rich for my blood. Jeremy's idea is to make sure he has some advanced anchor, and he feels that getting hit on the 11 point may not be too serious. I prefer nailing the offense down with 13/11(2), 6/4(2). Getting an advanced anchor isn't too vital since I have no board, and not leaving a shot plus making the potentially very valuable 11 point looks more important to me. It should be noted that 24/20(2), often the right play with 2-2, would be incorrect here. The reason is that I haven't begun to build my board so the back men aren't in trouble yet. With the 2-2 roll such a fine developing roll, offense comes before defense.

Jeremy: Making the 20 point is nice but it does nothing offensively and leaves an awkward two. Making the four point is very strong. Making the 11 point after that would be good offensively, but would strip the midpoint and do nothing defensively. Making the four point and anchoring on the 22 point leaves a balanced position. With the anchor I'm willing to expose a blot on the 11 point which will be very useful if not hit. I like my play.

TD: Jeremy got a bit carried away here. An advanced anchor just isn't all that important. Kit's more solid suggestion is the clear winner.

    13/11(2), 6/4(2)      -.053
    24/22, 13/11, 6/4(2)  -.089
    24/22, 6/4(3)         -.096

Position 7: Red to play 61.

Kit: Clearly better than making my five point, which would leave Jeremy a lot of counterplay. By sending a fourth man back I rip away a key builder, making it even more difficult for him to build his board. At the same time, I am now playing with only one man back, which is good.

Jeremy: Much better than making the five point which would leave the blot on the 20 point under the gun with me having a full roll.

TD: Absolutely! Kit follows the proper theme -- when ahead in the race, run for home. He punishes Jeremy's last play. Not remotely close.

    20/14*/13         +.172
    20/14*, 24/23     +.108
    11/5, 6/5         +.030

Position 8: White to play 21 from the bar.

Kit: The 1-3 structure is ok here and the builder on the 11 point could be very important, so Jeremy's play is better than B/23, 24/23. Slotting plays such as B/23, 6/5 or B/23, 8/7 don't look right. While Jeremy may wind up playing a back game, he doesn't have to try to get into one yet. He would rather win frontwards.

Jeremy: B/23, 24/23 might be better if I were committed to playing a back game because the 2-3 back game is usually a tad better than the 1-3 back game. But I'm not playing a back game yet by any means, and putting the builder on the 11 point is a big plus. B/22 doesn't give me the security of two defensive points and doesn't give me the new offensive builder.

TD: Good play. Jeremy properly keeps all his options open rather than committing himself to one path.

    B/24, 13/11      -.151
    B/23, 6/5        -.181
    B/24, 8/6        -.194

Position 9: Red to play 61.

Kit: Debatable. The five point is generally better than the bar point, but in this position with Jeremy having two men pinned on my ace point the bar point is pretty strong. In addition my play leaves a key blot on the 11 point and somewhat cramps the position. I think I should have played 13/7, 8/7.

Jeremy: A big part of jumping from beginner to intermediate lies in understanding how much better this play is over making the bar point with 13/7, 8/7. The five point is intrinsically much more important than the bar point, and in this case the play is only made clearer by comparing the unstacking of the six point to the stripping of the midpoint.

TD: The five point it is, but only by a little bit. Kit's arguments for making the bar point are quite sound, but the five point is still the five point!

    11/5, 6/5      +.234
    13/7, 8/7      +.227
    24/18, 6/5     +.063

Position 10: White to play 32.

Kit: Looks best. Now Jeremy must slot, since there is no flexible way to play safe. He needs to keep his back anchors in case of an accident. If he is hit, it may well turn into a back game.

Jeremy: If you liked 11/8, 24/22 go back and reread Magriel's chapter 16: Safe Play vs. Bold Play. (I reread the whole book twice a year -- it's a masterpice). Magriel offers the following criteria for Safe Play vs. Bold Play: Do you have an advanced anchor? Is your board stronger (more points, fewer blots) than your opponent's? Do you have more men back than your opponent? Yes's indicate bold plays; No's indicate safe plays. I have more men back, two anchors, and an equally strong board. I should make a bold play. The logic works like this: Being hit in this position incurs a relatively small loss for me because I'm nowhere near ready to escape and/or disengage anyway. What bold play should I make? The point on the board that I want most is my five point -- so I start it. 11/9 is the only constructive deuce. 24/22 would be a mistake because I may want to play a back game if my forward blots are hit.

TD: Well done, guys. It's nice to see that you are finally following the proper theme for the position.

    11/9, 8/5      -.266
    13/11, 8/5     -.294
    8/5, 6/4       -.317

Position 11 before roll: Red's cube action?

Jeremy: Kit is correct not to double. My position is very sound, with winning chances frontwards and backwards. More importantly, it's very hard to find a sequence where I would have a drop next roll. Even when things go well Kit will almost always still be able to double me in.

TD: Double? What are we talking about? Not even close. Sure Kit has the advantage, but it isn't all that great and his threats aren't particularly strong.

Position 11: Red to play 52.

Kit: Ugly, but there isn't much else.

Jeremy: Not a great roll. First note that all of Magriel's criteria argue for Kit to make a safe play. He has fewer men back and no anchor. If he mistakenly decided to slot his four point, 1/2 of my rolls would nullify his advantage instantly. In this position the relative loss of being hit is huge. What safe play should Kit make? The blot on the 11 point is of little value because the five point is already made, so Kit plays 11/6. 13/8 would leave the blot there and strip the midpoint. With the two, Kit chooses 24/22 over 8/6. This keeps a better distribution in the front, but subjects his blot to attack. However with four men back my attack isn't likely to be very threatening, so I'll go along with 24/22. Good play of a bad roll.

TD: I guess I can excuse Kit for not finding the best play of 11/4. I had to search hard for it myself. It sure looks wrong to slot into a double shot, particularly when you have only one man back to your opponent's four. The key is that Kit's play is so, so ugly. Also, look at the neat duplication from my play -- Jeremy needs aces and threes to hit, aces and threes to cover the blot on the five point. In addition it will cost Jeremy one of his anchors to hit, and losing an anchor could come back to haunt him in the future. Actually, Kit's play is almost as good.

    11/4           +.147
    24/22, 11/6    +.142
    11/6, 8/6      +.140

Position 12: White to play 43.

Jeremy: Nothing has changed -- I still want to make bold plays. Nothing else is close.

TD: I agree -- the loose hit is quite clear.

    9/5, 6/3*        -.047
    9/5, 24/21       -.188
    9/5, 13/10       -.225

Position 13: Red to play 64 from the bar.

Position 14: White to play 54.

Kit: Jeremy is correct to get his back men moving while he has the opportunity to do so safely. Once I enter, it won't be as easy. Now he has a chance to grab my bar point and get all 15 of his men moving around the board as one unit.

Jeremy: The five is clear and then I have three reasonable fours. We can eliminate 13/9 because it divides my army into two unnecessarily. 22/18 is thematic. I'm no longer playing a back game so I start to extricate my back checkers. But at some risk. I give up my advanced anchor and leave three blots. What about 8/4? This does nothing with my back men, but they aren't in any great danger of being primed now anyway. The builder on the four point is just as useful as on the eight point. More importantly, I retain my anchor on the 22 point in case of accident. I think 8/4, 8/3 is best. I don't think I even considered it over the board. Kent Goulding wrote that most errors occur not from selecting the wrong play over the right play, but by never seeing the right play in the first place. He is correct.

TD: Jeremy should have stuck to his feel at the table. His actual play is clearly best. Taking this safe chance to liberate the back men is the right idea. 8/4, 8/3 is much too cramping.

    22/18, 8/3      +.142
    13/9, 8/3       +.096
    8/4, 8/3        +.029

Position 15: Red to play 52 from the bar.

Kit: This leaves Jeremy 24 hitting numbers, but if he misses I'm in great shape and even if he hits I will be ok if I can enter reasonably quickly. If I stop on the 23 point, I don't have a decent five. B/23, 8/3* is clearly too loose considering his four point board, and B/23, 13/8 is pretty ugly.

Jeremy: Three legal fives. 13/8 strips the midpoint, stacks the eight point, and is generally non-constructive. Yet it is the safest play against my four point board and might be correct. 8/3* is the most constructive if it works, but has the most to lose when it doesn't. 23/18 doesn't leave a second blot and keeps a good general distribution, but does leave 24 numbers to hit on a point that I would dearly love to make. I have no clue which is better, but would probably play B/23, 8/3*.

TD: The problem with Kit's play is that when Jeremy does hit he also starts the bar point, which is the point Jeremy wants to make. Bringing the fifth checker to the eight point is a bit ugly, but Kit's position can handle it. The key is to see that Kit's main priority is to run the back man, not shuffle checkers in the inner board, so a fifth checker on the eight point doesn't hurt him all that much. His back checker is much safer hiding on the 23 point out of harm's way.

    B/23, 13/8       -.019
    B/18             -.056
    B/23, 8/3*       -.068

Position 16 before roll: White's cube action?

Jeremy: Now I am correct not to double. Making the bar point with a 6-1 followed by a fan doesn't look like a pass -- I would still have four men back and Kit would be a favorite to enter next roll. It looks like boxes-fan is the only market losing sequence, which clearly isn't enough.

TD: Double? Who are you kidding. You're barely winning. Don't put such thoughts into innocent people's minds. Get ahead in the race first.

Position 16: White to play 31.

Kit: Jeremy splits his back men, virtually ensuring that he will be able to make some advanced anchor if he needs it. Reasonable, but I prefer going all out for the offensive bar point with 13/10, 8/7*. Making that bar point is just huge for him now.

Jeremy: Leaving the bar point slotted is a worthwhile risk. 13/10 to maximize builders for the bar point is much too loose -- it might be right if Kit didn't have his five point and I had an advanced anchor. That leaves 24/21 and 18/15. 18/15 has the benefit of creating 5-3 and 4-4 as bar point making numbers, but isolates the back checkers. 24/21 is likely to produce an advanced anchor, but leaves a lot of blots when Kit enters with doubles. It's close, but I still like my play.

TD: Jeremy's play is slightly the best, but all the reasonable plays are pretty close. Kit's concept of going all out after the bar point is not far off.

    24/21, 8/7*      +.134
    18/15, 8/7*      +.131
    13/10, 8/7*      +.126

Position 17: Red to play 61 from the bar.

Position 18: White to play 43 from the bar.

Kit: Looks best. The anchor in my inner board is much better protection against getting blitzed than making my bar point. Jeremy now covers the entire outfield, and if his outfield blot is hit he has a return six.

Jeremy: This is better than making the bar point. I don't want to leave three blots in Kit's board vulnerable to attack.

TD: I'm afraid you guys are a bit confused on this one. Just make the defensive bar point like a normal computer. This puts full pressure on the outfield, and you will have good chances to make another anchor in addition. The blots in the inner board aren't in much danger since Kit isn't about to be attacking with Jeremy having a four point board. Your play gives Kit a much smoother ride home.

    B/18           -.146
    B/21, 18/15    -.217
    B/22, 18/14    -.257

Position 19: Red to play 53.

Kit: This play is far from clear. It is safe and well-balanced, but now Jeremy has the opportunity to make my three point and have a very powerful defensive structure. The alternative is 8/3*, 5/3, making an important offensive point and unstacking my heavy six and eight points as well as putting Jeremy on the bar. The downside of course is the blot left on his bar point. He would have a few return shots at it, and even if it isn't hit I would still have to get it to safety next turn. It's a tough choice, but if I had to do it over again I think I would make the three point.

Jeremy: Good play. Making the three point would leave eight immediate return shots and the blot would not be guaranteed safety on the next roll. Kit's play leaves no blots, smooths out the position, and will most likely lead to a perfectly efficient cube. If you wanted to make the three point, go back and reread Magriel's chapter 16 again.

TD: Making the three point is just barely better. Really, it is too close to call. The very sound arguments for both plays just balance each other out.

    8/3*, 6/3       +.343
    18/13, 8/5      +.340
    18/10*          +.232

Position 20: White to play 42.

Kit: Jeremy thinks his best defensive chances lie in hanging back on my ace point. I do not agree. Any double anchor is quite effective for generating shots. By staying back, he risks eventually getting attacked at just the wrong time. I strongly believe he should have played 24/22, 15/11. Even though the timing is not great for a 3-4 back / holding game, this play will make life very difficult for me. Sure, I will be able to dump checkers behind his rear anchor, but I will still have plenty of problems. In addition, after making my three point he can't be blown away.

Jeremy: Hmmm. I think I like this play better than 15/9 because the blot on his bar point is only attacked by two builders and may also provoke an exchange of hits earlier. But the real question is why not 24/22, 15/11 creating the 3-4 back game. I remember thinking that I didn't have enough timing for the back game -- meaning that I would likely be forced to leave one of my anchors prematurely. However, that doesn't mean the play is wrong. If I do have to leave his three point early I'll usually have the same holding game I'm resigning myself to with my actual play. On the other hand, if Kit rolls big doubles and my timing is restored I'll be much happier with the two anchors. I now think 24/22, 15/11 is best.

TD: Jeremy's committal play is not best, as he saw in the post-mortem. Simply make the anchor and sit on the position, and see which way the wind blows.

    24/22, 15/11      -.320
    24/22, 13/9       -.347
    22/18, 15/13      -.378

Position 21 before roll: Red's cube action?

Jeremy: Once again, not quite a double. Kit needs another point somewhere to ensure a smooth bearin. As long as the bar point is open I'm likely to have a take, so there is little danger of losing his market. Also quite important is that my board is excellent.

TD: Correct. Kit isn't far from a double, but the position just isn't volatile enough to send the cube over with an equity of .378. He needs a bit more.

Position 21: Red to play 53.

TD: Making the three point is correct, but not as automatic as one might think. 13/5 is definitely in the ball park, due to the extra flexibility from the spare on the eight point.

    8/3, 6/3       +.416
    13/5           +.376
    13/8, 6/3      +.180

Position 22: White to play 43.

TD: So you humans thought this play was automatic, did you? Well, not only is it not automatic; it is wrong. The correct play is 21/18, 13/9. Don't underestimate the power of the defensive bar point in this sort of position. The remaining back men aren't in too much danger of being attacked, and there will be the chance to make a second anchor which will be very big.

    21/18, 13/9       -.402
    18/11             -.440
    18/14, 13/10      -.461

Position 23 before roll: Red's cube action?

Kit: This is a decent double. I have a big racing lead, so if I can come home safely I will almost certainly win. I now have a good offensive structure, so I have the additional chance of attacking his back blot. If I roll doubles and clear the midpoint or if I point on him and he flunks I will lose my market, so doubling looks ok.

Jeremy: A classic double. Kit has a 39 pip lead against a four point holding game. My blot on his ace point is pretty insignificant and largely overrated by many players. It won't cause him many problems.

TD: Right on! Equity of .440 and a fair amount of volatility. A fine double.

Kit: A very clear take. Granted Jeremy is well behind in the race, but he will have good chances to get a shot or two. Of great importance is that his offensive board is perfect, so if he hits a shot he will have a very strong redouble immediately. If his board weren't so good, he might not have a take.

Jeremy: Absolutely. Kit's bar point is open and my board is perfect. If either of these cube actions passed you by, you need Robertie's Advanced Backgammon. It is, by far, the best compendium of reference positions like this which need to be mastered by any serious player. Magriel and Robertie: those two alone will take you a long, long, way.

TD: Definitely! Equity of -.440 is clear take. Not bad cube handling for a couple of humans.

Position 23: Red to play 43.

TD: I'm afraid you guys are playing too fast without considering all the possibilities. Kit's main goal here is to clear his midpoint, and making the two point has nothing to do with that. He leaves himself a thin position with only one spare left in his inner board. He should play 13/6. This starts to clear the midpoint and keeps more of an attack force in place. Making the two point isn't so important here, because Jeremy probably won't be hanging back on Kit's ace point too long anyway.

    13/6          +.462
    6/2, 5/2      +.439
    6/3, 5/1*     +.234

Position 24: White to play 65.

Kit: Once I have made my two point Jeremy's back man isn't covering anything useful, so he is correct to spring it out. Before it was an asset sitting back there; now it is just a liability.

Jeremy: Correct. Once Kit has made his two point, the annoyance factor of the blot on his ace point goes down considerably. Instead, it becomes a target which Kit can attack. Notice how quickly the blot bacame irrelevant if you thought it was a factor when Kit doubled.

TD: Right, but quite close. Hanging back still has some value. However, running does negate Kit's last play of making the two point.

    24/13           -.460
    13/7, 11/6      -.469
    13/8, 11/5      -.470

Position 25: Red to play 61.

Jeremy: No shots.

TD: Safety first. Not close.

    13/6          +.484
    13/7, 6/5     +.300
    13/7, 8/7     +.294

Position 26: White to play 52.

Kit: This is better than 13/6 Jeremy wants maximum builders to make the bar point.

TD: I like 13/11, 13/8, which also spreads the builders out. However, I will admit that these kinds of technical plays are not the strongest part of my game. They generally aren't too important in the grand scheme of things. My real edge is in overall positional judgment decisions, where I have it all over humans because of my vast experience and ability to accurately and objectively weigh all the relevant factors.

    13/11, 13/8      -.459
    13/8, 11/9       -.481
    13/6             -.489

Position 27: Red to play 53.

Kit: This figures to be better than 13/5. The spare on the eight point allows me to handle an awkward six. After 13/5, both 6-1 and 6-4 would leave a shot next roll.

Jeremy: 13/5 looks more natural, but leaves a shot if his next roll is 6-4 or 6-1, whereas Kit's actual play doesn't. That looks like the only difference so I think Kit's play is right, even though I know I would have played 13/5 without a thought. Perhaps I won't next time.

TD: My evaluator says 13/5 is the best. However, Kit's play might be right. The problem is that you humans want your answers quickly, so you don't give me time to examine all the possibilities thoroughly and you can't build fast enough computers. Consequently I am limited to looking ahead only one move (i.e. Jeremy's responses in this case) and can't see what will happen after Kit's next roll so the 6-1 and 6-4 danger he speaks of escapes me. Therefore, I guess I'll just have to trust you guys in positions such as this one, although I am reluctant to trust human backgammon judgment very far. I guess it's just as well -- if I made the best move all the time you guys probably wouldn't play with me any more. Really, this sort of nitpicking technical analysis bores me anyway -- let's get back to the more exciting positional judgment area.

    13/5           +.506
    13/8, 6/3      +.466
    8/5, 8/3       +.409

Position 28: White to play 53.

Kit: Jeremy is trying to maximize his builders for the bar and two points. I think 13/8, 9/6 is a bit better. This locks up the eight point, which is valuable, and bringing a spare to the six point can't be all bad.

Jeremy: I can't slot the bar or two points, so I bring in another builder. Stop playing for money if you liked 21/13.

TD: I agree with Jeremy -- spread them out. However, once again my answers here my be somewhat suspect.

    13/10, 9/4      -.428
    13/5            -.442
    9/4, 8/5        -.448

Position 29: Red to play 64.

Jeremy: No shots.

TD: Safe is safe. Not even close.

    8/2, 6/2       +.338
    13/3           +.150
    13/9, 13/7     +.020

Position 30: White to play 61.

Kit: Even though Jeremy might get a shot next roll, he is quite correct to slot the two point. He needs that point in all variations, and slotting it is the best way to make it. If he plays 9/3, 8/7 things could get awkward next roll.

Jeremy: One point which Kit has emphasized over and over in his wonderful Matchqiz is the need to slot points in positions like this instead of being afraid of the parlay: opponent rolls a number that leaves a blot, you roll a number to hit that blot, opponent rolls a number that comes in hitting your slotted point. This is a pretty remote parlay. In addition, if I don't start the two point now it may be hard to make later. Slotting and covering is by far the easiest way to make points. Note that making the bar point is a big blunder, allowing Kit to clear his midpoint safely if he rolls an ace.

TD: Good thinking, guys. Like I always said -- you can't make points without starting them.

    9/2          -.268
    8/2, 5/4     -.292
    9/3, 8/7     -.296

Position 31: Red to play 65.

Kit: This is a classic pay me now or pay me later situation. I can play safe with 8/3, 8/2 and pray for rain next turn, or I can do what I did. It is usually correct to pay later, but there are exceptions and I think this is one of them. There are several relevant factors: 1) If I get away with my play I will be in great shape -- a clear favorite to not leave any more shots. 2) The pay later play is very dangerous. My position would be completely stripped, and it would be quite likely that I would have to leave far more than an 11 number shot in the next roll or two. 3) Right now Jeremy has a blot in his inner board, but he figures to cover it next roll and have a perfect board. Thus, getting hit will probably be more serious next roll. All these factors are important. If one or more of them didn't exist paying later would probably be better, but here it looks right to shoot it out now.

Jeremy: Another issue that Kit analyzes often in Matchqiz is the classic "Pay me now or pay me later" question. The factors are: Is my opponent's board improving or crashing? If I pay later will I be likely to leave more and worse shots? If I pay now and am missed will I be home free? In this case the answers are: improving, much worse, and very nearly. Kit should pay now, and he does. If this sounds an awful lot like what Kit says it's because I study his Matchqiz almost daily. In the last 33 moves I have had occasion to draw upon the three backgammon sources that have influenced me the most. In their order of complexity, Magriel, Robertie, Woolsey: these three alone will have you winning tournaments. Of course if you really want to be good, read everything, take lessons from pro's, and quit your job.

TD: Now that's more like it. Enough of the technical stuff -- back to the real decisions which win or lose games. Pay me now or pay me later decisions are right up my alley. In this case, paying now was right by a country mile, for all the reasons the boys said. Well done, guys. You're finally starting to get your priorities straight.

    13/8, 13/7      +.344
    8/3, 8/2        +.188
    13/2            +.010

Position 32: White to play 42.

Kit: Covering the blot is correct. There is virtually no gammon danger, and if Jeremy hits a shot he will definitely want to have his two point.

TD: Correct. No reason not to cover the blot.

    8/2            -.622
    13/11, 8/4     -.645
    13/7           -.658

Position 33: Red to play 21.

Kit: I'm not so sure about this play. It is clearly safer than the alternative of 8/6, 7/6 for next roll, but it does leave a stripped position and two points to clear. However the fact that it is 100% safe next turn and doesn't figure to be so bad after that makes it look like the winner.

Jeremy: Hmmm. At first glance 8/6, 7/6 appears to be better because it clears a point rather than creating one. However that play leaves a shot on 6-1 and 6-4 (once again) whereas this play doesn't.

TD: Yeah, I know. My algorithm gives 7/5, 2/1 as best, and you guys laugh at this play. Well you can look ahead in this position better than I can, so I'll bow down and accept your judgment. I accomplished the important part -- clearing the midpoint safely -- now you guys can figure out how to lock up the win.

    7/5, 2/1      +.696
    8/6, 7/6      +.680
    8/7, 3/1      +.671

Position 34: White to play 44.

Position 35: Red to play 53.

Position 36: White to play 21.

Position 37: Red to play 43.

Position 38: White to play 32.

Position 39: Red to play 44.

Jeremy: Obviously better than 7/3, 5/1(3) in that it leaves 11 shots instead of 14.

Position 40: White to play 33.

Kit: This is not right. The key is that I have 15 men left, so taking only one checker off is quite sufficient. If I had 14 men left then this play might be correct because it would prevent me from taking two checkers off on some rolls. His play can be quite costly if I roll something like 6-1, hit and take the man off, and he flunks for a while. He should get both men moving.

Jeremy: 21/18, 21/12 looks like a slightly more efficient use of the last three in that it gains an extra crossover. However, I think my play is better because it forces Kit to safety his blot with an ace, two, or three instead of taking a second checker off. Although, now that I look closely I see that Kit has an odd number of checkers to take off (15) and looks likely he'll not miss in the future so taking only one off now isn't any loss. Of course, he could roll four aces and miss, in which case the extra checker could matter. The plays are very close.

TD: You guys argue this one out. It's too detailed for me.

    21/9           -.936
    21/12, 5/2     -.949
    21/12, 21/18   -.952

Position 41: Red to play 64.

Position 42: White to play 32.

Kit: 21/19, 9/6 is slightly better technique for getting off the gammon, since that play brings one checker exactly to the six point. In practice, it is very unlikely to matter.

Jeremy: 21/19, 9/6 is better for getting off the gammon, which might be an issue if Kit rolls three consecutive doubles and I roll small. My play is probably wrong by a thousandth of a point.

Position 43: Red to play 31.

Position 44: White to play 53.

Jeremy: It transposed anyway.

Position 45: Red to play 33.

Position 46: White to play 63.

Position 47: Red to play 61.

Position 48: White to play 43.

Position 49: Red to play 61.

Match to 9:   Kit Woolsey - 2,  Jeremy Bagai - 0

Go on to next game: Game 2

Return to: Backgammon Galore : Woolsey-Bagai Match