From: Backgammon Galore : Woolsey-Bagai Match

Kit Woolsey vs. Jeremy Bagai
Game 7

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

Match to 9:   Kit Woolsey - 4,  Jeremy Bagai - 7

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

Position 1: White to play 32.

Jeremy: As I mentioned last game, the leading player should strive for an advanced anchor in order to avoid positions with nasty gammon threats. This play is better than 13/11, 13/10.

TD: Not to mention that it's just plain the best play.

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

Position 2: Red to play 32.

Kit: Once again I prefer the valuable split with reasonable safety. Here there is strong argument for 13/10, 6/4*, since the builder is six away, but I like to work on both sides of the board. The play to be avoided is 6/4*/1*. Granted it is safer than hitting on the four point and stopping, but it is much less productive when it works. Avoid dumping a checker onto the ace point in the early stages unless you really have to do so.

Jeremy: I think 13/10, 6/4*, another builder for the four point, is a little better.

TD: Jeremy's play is a tiny bit better, but it is too close to call. You two are right on target on this one.

    13/10, 6/4*      -.068
    24/21, 6/4*      -.071
    6/4*/1*          -.092

Position 3: White to play 66 from the bar.

Position 4: Red to play 41.

Kit: I have the stronger board, so I challenge Jeremy's blot in the outfield. 24/23, 8/4 doesn't put him under nearly as much pressure.

Jeremy: Attacking my blot.

TD: I get 24/23, 8/4 a bit better, presumably in order to avoid return sixes. However, Kit's play is probably ok, particularly at the match score, since he is trying to create action to increase gammon possibilities.

    24/23, 8/4      +.198
    21/20, 8/4      +.186
    8/4, 6/5        +.117

Position 5: White to play 64 from the bar.

Position 6 before roll: Red's cube action?

Kit: This is clearly a big double. I have the stronger board, he is on the bar, and I am shooting at another blot. In addition if he takes he will have no use for the cube at the match score. The take is the problem.

Jeremy: This is a great double at this score. Kit is threatening to blow me out of the water. He has the better board, is attacking another checker, and I'm still in the starting gate.

TD: Sure is. In fact with equity of .409 and reasonable volatility, it is even a thin money double.

Kit: At this match score there are arguments for making aggressive takes. The key is that Jeremy needs exactly two points to win the match, while there isn't too much difference between being ahead 7 to 5 and 7 to 6. However these arguments generally apply to positions where gammons are impossible or unlikely. In this position a good chunk of Jeremy's losses will be gammons, since he could wind up with quite a few men back. Also he will be sitting on a dead cube, and winning a gammon is meaningless for him. He has a clear disadvantage in all stages of the game, and things figure to get worse before they get better. I think he should let it go.

Jeremy: Many good players would pass this and I think they'd be wrong. Kit just doesn't have enough for me to be really worried about gammons at this point. He only has a two point board. I only have two (possibly three) men back. The blitz is still far away. I'm losing to be sure, but I can't start giving away points.

TD: I don't exactly know how to filter in the match conditions, but my guess would be that it is a borderline decision, probably a take. A fine conclusion to great cube handling throughout the match.

Position 6: Red to play 51.

Position 7: White to play 52 from the bar.

Position 8: Red to play 63.

Kit: I must fight for the five point -- if he wins the battle there he will almost equalize the game. Double hitting with 8/5*, 8/2* isn't worth it. I would lose my eight point and my position would be a mess.

Jeremy: I think this is right. Kit's five point is the most valuable point on the board for both of us, so he tries to take it. But 24/21, 14/8 can't be too bad. At much less risk it leaves a balanced position with the real advantage of one man back to my three. The reason why the play isn't clear is that Magriel's criteria for safe vs. bold plays are in conflict: Kit has fewer men back but also has the stronger board.

TD: Jeremy is right on target. Kit's play is best, but 24/21, 14/8 is right behind. Of course considering the match score Kit's play is probably even better, since it is more likely to lead to a gammon.

    14/5*           +.282
    24/21, 14/8     +.274
    14/8, 13/10     +.254

Position 9: White to play 33 from the bar.

Kit: The five point is the five point. Even though this costs Jeremy his eight point it is better than B/22, 23/20*, 6/3(2). That would leave a disjointed position, while making the five point leaves him a very compact position.

Jeremy: After entering and hitting I get to make the ten, five, or three point. There are arguments for each one. The ten point brings two new checkers into play and leaves a nice structure for prime building. It just doesn't have the punch of an inside point. Note that there would be more going for this play if there were still five checkers stacked on the midpoint rather than four. The five point is the best point on the board but leaves a blot on the eight point and doesn't unstack the position. The three point is an inner board point and doesn't leave a blot but is a little deep. Any one of these plays might be right, and you encounter positions similar to this one where each play is right. When in doubt, make the five point.

TD: Sorry, guys. This is one time when the five point isn't right. The loss of the eight point and the cramped position is just too much. The smoother but disjointed three point is better.

    B/22, 23/20*, 6/3(2)     -.025
    B/22, 23/20*, 8/5(2)     -.093
    B/22, 23/20*, 13/10(2)   -.094

Position 10: Red to play 66 from the bar.

Position 11: White to play 33.

Kit: I prefer 13/10(2), 6/3(2). The ten point is very valuable in this formation with the four point open. The reason is that there is likely to be a fight for the four point, and if I win the fight Jeremy will be glad to have his ten point as a blocking point against an anchor I might form on his four point.

Jeremy: I think this is better than 13/7, 6/3(2) because slotting the bar doesn't seem worth all the return shots it would leave.

TD: Have to give Kit the nod on this one. The ten point asset is key. Odd that Jeremy didn't even consider it as a candidate.

    13/10(2), 6/3(2)      +.245
    13/10, 8/5, 6/3(2)    +.205
    13/7, 6/3(2)          +.165

Position 12: Red to play 22 from the bar.

Kit: I think this is best. B/23, 13/11(2), 6/4 may seem better distributed, but if Jeremy makes my five point as he is likely to do then the third checker on the four point will be pretty much out of play. My play brings in more firepower in case Jeremy can't make my five point. While I would like to make his four point, B/21, 13/11(2) is not the right idea. This would just invite him to attack. By double covering the four point I make it too dangerous for him to slot it.

Jeremy: This is better than coming up to my four point -- that's a point I want and for which I have three builders.

TD: I like B/23, 13/11(2), 6/4 a tiny bit better, but Kit's play is fine. Coming up to the 21 point would be very wrong, as the analysts properly point out.

    B/23, 13/11(2), 6/4      -.078
    B/23, 13/11(3)           -.079
    B/21, 13/11(2)           -.161

Position 13: White to play 33.

Jeremy: It's not that I particularly want my two point -- I don't want Kit to have his full roll to attack on his five point.

TD: The loose hit on the two point is 100% clear.

    13/7, 10/7, 5/2*      +.090
    13/10(3), 5/2*        -.010
    10/1*, 5/2*           -.051

Position 14: Red to play 66 from the bar.

Position 15: White to play 31.

Position 16: Red to play 63 from the bar.

Position 17: White to play 42.

Kit: Things sure went Jeremy's way. Now how should he proceed? He is too short on attack material to try 7/1*; if I hit back he has nothing to follow up with and he might never get his bar point back. 24/20, 22/20 is safe but unproductive. I like his actual play. He locks up the advanced anchor, which is obviously important, and brings another builder down in range of his four point. The loss of the midpoint is not serious, since he needs to bring builders down and his men on my five point cover the outfield effectively.

Jeremy: This is much better than 7/1*. I don't have enough checkers up front to go for a closeout. I want to keep my bar point in case he anchors on my ace point. I want another builder for the four point. I want some defensive security.

TD: Careless, guys. You just didn't consider all the possibilities or the correct play of 22/18, 20/18 would have jumped out at you. Jeremy doesn't want to get stuck on the 20 point, since Kit owns the blocking 11 point. Since Jeremy is ahead he will be planning to get off any anchor he makes soon, and the bar point will be much easier to clear than the five point. The builder for the four point can wait -- by itself it isn't too valuable anyway. Also, keeping the midpoint for a roll or two may be important.

    22/18, 20/18      +.478
    24/20, 22/20      +.442
    24/20, 13/11      +.425

Position 18: Red to play 62 from the bar.

Position 19: White to play 63.

Kit: Again looks best. One more builder for the four point, and the back man gets some breathing room. If the back man isn't moved now, things could get cramped. Once again, it doesn't appear worth the risk of losing the bar point to hit with 9/6, 7/1*, even though Jeremy would have better chances than last roll to carry out the blitz because his ammunition is closer.

Jeremy: There is more to be said for 9/6, 7/1* now that I have an anchor and more builders, but I still think keeping the five-prime is better.

TD: This time there is enough ammunition ready, so I think Jeremy should attack with 13/10, 7/1*. This appears to be better than 9/6, 7/1* which both analysts considered, because it is less committal and the bar point is easier to remake. Admittedly due to the match conditions (Jeremy doesn't need a gammon to win), perhaps he is correct not to do so. However, if he is not going to attack his approach isn't correct. The builder on the ten point just isn't as important as he thinks it is.

    13/10, 7/1*      +.550
    24/18, 9/6       +.535
    24/15            +.532
    24/18, 13/10     +.503

Position 20: Red to play 54 from the bar.

Position 21: White to play 44 from the bar.

Position 22: Red to play 54.

Kit: Naturally I stay back on the 16 point in order to pick up the third checker. This is a quick turnaround, but Jeremy's advanced anchor and strong board figure to keep him in the game.

Jeremy: You shouldn't play for safety with two checkers in the air -- this is better than 16/7* because it still attacks my outfield blot and provides a better builder up front.

TD: Things just aren't the way you guys think. It is Jeremy who has the stronger board. This means that if he gets both men in quickly, Kit will be scrambling to tidy up his blots. This could be costly, since Kit also wants to escape his back man if he can. Therefore, he shouldn't be leaving too many blots. Scooping up the other outfield blot really isn't all that important. I like 16/7*, which leaves fewer things to worry about if Jeremy enters.

    16/7*          +.119
    13/8, 11/7*    +.077
    11/7*, 11/6    -.036

Position 23: White to play 64 from the bar.

Position 24: Red to play 41.

Kit: The idea behind this play is to get an extra builder for the three point. I felt that with Jeremy having three men on the bar I could afford to be this loose. In retrospect I think I should have played the more solid 16/15*, 11/7. My play gets hit immediately if he rolls 2-2, and if he brings both men in I will be scrambling to avoid leaving a shot. Considering that I have the further problem of liberating my back men, I don't want too many complications up front.

Jeremy: I don't know if this is better than making the bar point with 16/15*, 11/7. I have three on the roof so it seems like Kit should diversify in order to make points rather than to button up. But his play does pay off immediately to double 2's, and if Kit rolls an escaping number next turn he may not want to have to worry about all those blots. I really don't know. TD?

TD: At least you guys are starting to wake up. Kit's actual play is very wrong. 16/15*, 11/7 is much better. The reasons are all the same. Kit doesn't want to have a nightmarish cleanup problem if Jeremy gets in quickly. Kit realized it in his analysis, and Jeremy wasn't sure. Next time, maybe you'll know how to handle this one.

    16/15*, 11/7      +.388
    16/15*/11         +.331
    16/15*, 13/9      +.297

Position 25: White to play 51 from the bar.

Position 26: Red to play 63.

Kit: This is a good illustration of the defect of my play last turn. I roll the perfect number to escape the back man, but must use it to tidy up in front. At least I was able to, although there is still the gaping hole on my bar point which should have been filled. I could make the three point, of course, but this would give Jeremy several combination shots at the blot on my bar point and getting hit now would be disastrous.

Jeremy: Choices. Escaping all the way leaves 22(!) shots, so it probably isn't right. Escaping part way with 24/21, 15/9 looks safer but still leaves 21 shots because all 3's and 1-1's now hit on my four point. Making the three point leaves 15 shots (double 5's now hits in the outfield). Kit plays safe. He's probably right, but I'll bet he wished he had made his bar point last roll.

TD: Sanity finally prevails. Kit's play is quite correct.

    15/9, 7/4       +.270
    24/21, 15/9     +.242
    9/3, 6/3        +.145

Position 27: White to play 55 from the bar.

Kit: Jeremy is treading on some thin ice here. His checkers are pretty well blocked in, and he could have some problems if he rolls awkwardly next turn. I think he should strive for more air with B/10, 20/15. The blot on the 15 point is virtually immune, and he will be able to handle almost anything from then on. In addition, if I escape my back man he won't have anybody in the outfield to pick me off. His plan is to bring the ammunition in quickly, but since the checker on the ten point covers his four point his play is not necessary for that purpose.

TD: Jeremy's play is awful, much worse than might be imagined. He needs breathing room badly. Kit's analysis is quite correct.

    B/10, 20/15          +.013
    B/15, 20/15(2)       -.108
    B/5                  -.149

Position 28: Red to play 52.

Kit: I don't like giving up the 11 point, but I can't afford to leave a shot now. At least my position is reasonably flexible.

TD: This is a tough one to find. After all I've been preaching about the importance of flexibility, keeping checkers in front of the enemy anchor, and playing safe when the opponent has the stronger board, I now recommend hitting loose on the ace point and stripping the whole position. How can this be right? The answer is the extreme value of the blocking 11 point. Jeremy's position is completely stripped, and blocking him from escaping with a six may cause him to collapse his whole board. This is so important here that everything else takes a back seat.

    8/1*           -.006
    11/9, 11/6     -.037
    6/4, 6/1*      -.046

Position 29: White to play 42.

Kit: Correct. Jeremy gets a badly needed checker into the outfield, while hanging back with the back man to make life difficult for me.

Jeremy: My five prime is worth too much to play 7/1*. Instead I cover the outfield and wait.

TD: No other play is even worth thinking about. Jeremy MUST get out into the outfield.

    20/14           +.086
    24/18           -.001
    20/18, 5/1*     -.028

Position 30: Red to play 54.

Jeremy: This is pretty ugly, to put it mildly, but it is the only safe play. The alternative is 8/4, 6/1*. I'm not sure which is better.

TD: I am, of course. Safety it is, now that the reason for hitting loose has vanished.

    9/4, 8/4      -.290
    8/4, 6/1*     -.317
    9/4, 6/2      -.371

Position 31: White to play 42.

Kit: Jeremy stops on the ten point in order to cover more territory if I roll 3-4 or 3-5. However by advancing the back man he makes it easier for me to play behind him, and also takes more of a risk of being pointed on. I think he should make a stronger effort to force me to leave a shot and play 14/8.

Jeremy: Very interesting. Alternatives include 14/12, 5/1*; 24/22, 5/1*; 24/20, 14/12; 24/18; and 14/8. About the only things I am sure of are that I shouldn't break my five-prime and I shouldn't break my anchor. After that I have no idea. Interestingly, this is a position for which I would trust a good computer rollout. An awful lot depends on the next two rolls, and the checker play after that might be pretty clear. TD?

TD: What kind of a play is this? Moving the back man to exactly where it is blocked and can be pointed on? Yuck! Kit's plan of forcing a shot at another blot is equally bad. The theme here is to get that back man out of there before something bad happens. Jeremy needs to control the outfield. By far the best play is 24/18. The checker is safe, and in good position to patrol the outer boards if Kit is able to spring his back man. Both of your plays are way down on the list. You both lost the thread here.

    24/18           +.312
    24/20, 14/12    +.247
    20/14           +.228
    24/22, 14/10    +.156

Position 32: Red to play 53.

Jeremy: More choices. Kit can make the three point on my head, make the ace point behind me, or come out. Making the three point leaves a lot of shots -- 22 of them if you count 4's and 6's to hit on my ace point, which I would do because of the blot in the outfield. Making the ace point smooths out his position for later attack, but does nothing to escape. Escaping gives me 13 hitters, but gains a lot when it works. I like his play.

TD: What choices? This isn't remotely close. Kit's play is a standout.

    24/16          +.111
    6/1, 4/1       -.206
    8/3*, 6/3      -.259

Position 33: White to play 21.

Kit: Jeremy's last play sure was a success. Now he properly gets off the 22 point, which is the point which is blocked on fives and sixes.

TD: Not quite. Jeremy is no longer blocked on sixes after his play, but what about fours, twos, and aces. Kit's prime may not seem like much, but bad things could happen. Jeremy should play 20/18, 10/9*. What if Kit rolls 1-1 or 4-4? Well, maybe he won't. We computers look at the big picture and don't concern ourselves about the freak occurences.

    20/18, 10/9*      +.307
    22/20, 10/9*      +.264
    10/9*/7           +.222

Position 34: Red to play 41 from the bar.

Position 35: White to play 33.

Kit: I don't think this is right. Jeremy needs fives to escape, while fours are his worst number. Therefore I think he should play 9/3, 7/4(2)*, so his fours go to the ace point.

Jeremy: I can't see any difference between this and 9/3, 7/4(2)*.

TD: Kit is correct. His reasons are quite valid.

    9/3, 7/4(2)*          +.637
    9/6, 7/4(2)*, 5/2     +.596
    7/4(2)*/1(2)          +.384

Position 36: Red to play 42 from the bar.

Position 37: White to play 42.

Position 38: Red to play 65 from the bar.

Position 39: White to play 31.

Kit: This time Jeremy properly shifts to the five point so his good numbers are not duplicated. It would not be correct to slot the ace point. If I roll an ace, he would be one checker short with which to attack. His goal is to keep building up the pressure, so that when I do roll that ace he is in position to pounce.

Jeremy: I see that I need fives to escape so I diversify with the ace, giving me fours to help close my board. The same concept as duplicating your opponent's numbers but in reverse.

TD: Back to the technical stuff. I have 14/10 a slight winner, but it is tough to argue with the experts' analyses so I suppose they are right. Btw, maybe slotting the ace point isn't all as bad as Kit says it is.

    14/10           +.783
    14/11, 6/5      +.781
    14/11, 2/1      +.780

Position 40: Red to play 61 from the bar.

Position 41: White to play 11.

Position 42: Red to play 62 from the bar.

Position 43: White to play 22.

Position 44: Red to play 52 from the bar.

Position 45: White to play 31.

Position 46: Red to play 31 from the bar.

Position 47: White to play 66.

Position 48: Red to play 66.

Kit: Not a happy situation. My position is awkward and I am unlikely to get a shot, but I am so far behind in the race that running seems worse. Maybe he'll be nice enough to roll a 6-4, the only number which leaves me a shot next turn.

Jeremy: Kit sees that he will be 22 pips and three crossovers down if he runs, so he stays back to make things more complicated. 6-4 is the only number which leaves a shot.

TD: Looks pretty close to me, but Kit's play comes out best.

    9/3(2), 8/2(2)        -.848
    24/12, 9/3, 8/2       -.850
    24/6, 8/2             -.854

Position 49: White to play 64.

Jeremy: I think this is right, but I'm by no means sure. 6/0, 6/2 looks easier to clean up next turn if not hit and Kit may be forced to leave with a six anyway. My play however, keeps Kit behind a five-prime so that when he does hit he isn't home free. This needs a rollout.

TD: Personally, I doubt if it is even close. Jeremy's play looks by far the best to me.

Position 50: Red to play 43 from the bar.

Position 51: White to play 41.

Position 52: Red to play 22 from the bar.

Position 53: White to play 11.

Kit: Correct. Nothing leaves a shot next turn, and Jeremy probably won't have to leave anything at all. 6/4(2) or 6/5(2), 2/0 would risk leaving a shot if he rolled large doubles next.

Jeremy: Generally I want to be clearing points instead of keeping contact, but 6/4(2) would be very wrong in that double 6's, 5's, and 4's become blot numbers, and Kit gets two large doubles to get back in the race instead of just one.

Position 54: Red to play 31 from the bar.

Position 55: White to play 53.

Position 56: Red to play 21 from the bar.

Position 57: White to play 42.

Position 58: Red to play 31 from the bar.

Position 59: White to play 52.

Position 60: Red to play 51 from the bar.

Match to 9 (final):   Kit Woolsey - 4,  Jeremy Bagai - 9

Return to: Backgammon Galore : Woolsey-Bagai Match