On Playing One Point Matches

Bernhard Kaiser (FIBS : one_pointer)

Reprinted from rec.games.backgammon

Bernhard Kaiser (one_pointer on FIBS) has updated his article on strategy for playing one-point matches. It is due to appear in an English backgammon magazine next month, although I do not know which one. He is temporarily without r.g.b. access, so I am posting it on his behalf (after doing some minor editing).

Cheers, - Darse.

Playing One Point Matches

Bernhard Kaiser is a 26-year-old student of mathematics from Germany and has been playing tournament backgammon for eight years and on the Internet server FIBS for the last two years. Yerry Felix, consultant to Games & Puzzles, asked him to write an article about one-point matches / double matchpoint. The text was edited and typeset by Paul Lamford [PL].

To set the scene, let me say I am rather experienced at this special score, because I played a great deal on FIBS under my alias one-pointer. I was interested in how I would fare playing only double-matchpoint situations, and I managed to climb to second place in the ratings.

First of all lets summarize the basic properties of a one-point match (OPM). You might say, one-point-matches, I never play them, so why should I learn about them?. But, of course, one point matches are basically the same as double-matchpoint (DMP) situations, where both players only need to win the last game. The special characteristic of DMP, as opposed to money games or other match-situations, is that both players only want to win the game, and do not care whether they win a gammon or lose a gammon. This irrelevance of gammons leads to situations in DMP, where you have to make different moves from those you are used to in money games.

I will now try to give some examples of such situations and to outline some general rules. Later in the article I will try to give some tips how to play OPM against very weak players. There may not be such weak players in your tournaments that you have to adapt your moves [don't you believe it PL], but on FIBS it is very important also to get maximum results against players with a rating, of, say, 1400 or less. [Equivalent to a BIBA rating of about 900 PL])

• 5-4. In a money game you can choose between 13/9 13/8, and 13/8 24/20. Mostly I pick the second choice. But at DMP the situation is different. Here it would be a slight error to play 24/20, 13/8. Why? Because this play tends to give you an advanced anchor. Although to have such an anchor is nice, at DMP its not your highest priority. You don't need the safety of an advanced anchor as an assurance against getting gammoned, because a gammon doesn't cost you any extra. Lets have a look at 13/9, 13/8: This one tends to lead to slightly more gammonish situations, but with you being a slight favourite to win one. So you are favourite to win a gammon with both plays. But, paradoxically as it seems, you don't want to be the favourite to win a gammon at DMP, because you cant use this extra equity! So this play, although probably better than 13/8 24/20, doesn't feel so great at DMP either. But is there anything else ? Yes there is! A mostly underrated third play is 24/15, which, in a money game, is nearly as, if not equally good as the two previously mentioned. But lets have a look at it at DMP: If you don't get hit with a three, you have the better game in most cases. If you do hit, your opponent has much the better game, mainly because of his increased gammon chances. In fact, after you play 24/15, your opponents chances of winning a gammon in a money game are better than yours! But you are still the favourite, because you are simply winning more games. These are just the plays you want to make at DMP, giving your opponent extra gammon-chances, but increasing your own winning chances a little. Computer roll-outs confirm this theory; while all three plays are rather equal in overall equity, the third play, 24/15 wins considerably more total games than both the others (about 51.7% against 50.5 % and 50.0 %). I am rather sure it must be the right play at DMP.

• 6-2, 6-3, 6-4. A similar reasoning could be applied to these opening moves. Since 24/18, 13/x is a bid for an advanced anchor, its not that attractive at DMP. So I prefer running (24/14) with 6-4. Because playing 24/18, 13/x with 6-3 and 6-2 is considerably better than running in a money game, and also gives some useless gammon chances to your opponent, the right play is not so clear here. I tend to alternate between running and bringing a man down with the 6-3 and play 24/18, 13/11 with the 6-2 (also agreeing with my computer roll-outs).

• 2-1, 4-1, 5-1. Playing 24/23 13/x tends to result in an anchor, and 13/x, 6/5 tends to produce a gammonish position. Its not clear at all, which is the better approach. Do we have a third play here? Yes! Just playing 13/10. But although its much closer than in a money game, this try is unfortunately just a little bit too weak. So my preference is to play 6/5 with every ace, because I like to keep my position smooth; but I could easily be talked into the split. [note: In the original thread, Toni Wuersch commented "don't slot unless you are already way behind; if way behind, slot more aggressively". This seems reasonable, given the large setback in race position (pips) whenever the slot is hit. -drb]

• 4-3, 3-2. With these rolls I think at DMP its better to bring two men down instead of going for the anchor, for the same reasons given above.

General rules. What style of game do you want to reach?

a) Blitz: Of course a blitz is not the type of game you want to play, because if you win, you mostly win a worthless gammon, but if things do not go well, you become an underdog quickly. There are many plays in blitz situations which you have to handle different. Lets have a look at an example: Your opponent opens with 5-4 and moves 13/8, 24/20. Then you roll a 5-5. In a money game making both inner points is routine and the best play by a wide margin, because you are winning a lot of gammons. At DMP I would rather play 13/3(2)! I think this gives you just slightly more winning chances. Generally its important to keep your position smooth and flexible at DMP. You should not be stuck in a single game plan too early, especially not a blitz. So you shouldn't hit loose on the ace or make your ace-point, if you don't have to, because hitting on the ace favours a blitz, which isn't attractive at DMP, and it reduces the chances of other game types, such as a backgame, which is quite a good plan at DMP (see below). Of course you should not overdo this rule. Sometimes hitting on the ace-point or going for the blitz is just plainly the best approach.

As stated, a backgame is a plan you would of course prefer to play at DMP than in a money game. With the right timing you are often the favourite to win, as gammons don't count. The problem is to get the right timing. So you should not become committed to a back game from move one. This could backfire quickly. You only should tend to like a backgame more than usual. And if playing a backgame, you should go for a full-blooded one, meaning that you have to do all you can to preserve your timing. Believe me, I have won a fair number of games just by making hara kiri plays such as leaving blots on every point in my home board, because my timing was about to deteriorate. Remember, you don't have to fear a gammon or, not unimportantly, even a backgammon! With 15 men on two of your opponents deep home points, the backgammon chances are high, but your winning chances are better than you might think!

In some ways similar to backgames are ace-point or deuce-point games. With proper timing, you have winning chances in both games of about 20%, making these types of game considerably stronger than in money game because there is no gammon danger. So, if things are going badly, and you have a low anchor, you are still not out of the game, no matter how many men you have on the anchor or even on the bar. So you should tend not to give up the opponents ace-point too early, because its an assurance of some winning chances until the end, no matter what happens. This is a further reason for my leaving the back men unmoved on some opening rolls (such as 4-1).

Lets make a short summary of all these points:

• a) In the opening, going for an advanced anchor is less important than in money games.

• b) Play positionally and smoothly; be reluctant to hit on or make the ace-point. A blitz is not the type of game you should aim for.

• c) A backgame is quite an attractive game plan, but don't go for it from the outset. If playing a backgame, do all you can to preserve your timing.

• d) Ace-point and two-point games are still not desirable, but considerably better than in money games. So you should think carefully about breaking a low anchor prematurely.

• e) Don't overdo these rules! In most positions (say about 95%) the best play in a money game is also the best play at DMP!

Playing against a much weaker opponent.

If you are playing against a much weaker opponent (this could happen if you play on FIBS, or if you want to prove to a beginner that backgammon has nothing to do with luck, or if you are playing one single game for $500 against a pimp, who has not the faintest idea about the game, as I did recently), you should at least consider the following thoughts of mine:

• a) In the opinion of many people, you should try from the beginning to complicate the game, to force the beginner to make errors. This surely isn't wrong at all, but I have a slightly different suggestion: Firstly, I think, you should try to get a solid racing advantage! If you are 30 pips or more ahead, your win is almost guaranteed, because most beginners don't know the concept of building up their board while waiting for a shot. I really get the feeling that if I am way ahead in the race, I cant lose against a very weak opponent. Only if you cant manage this, say if the race is still rather even or worse, should you complicate the game, meaning that you start playing very positionally, slotting points, and falling back on a backgame or other type of holding game.

• b) I think a rather important point against a beginner, more so than against an average player, is to hold the ace point as long as possible! Then you'll be in the game till the end, no matter what happens. Also a beginner will make terrible errors at the end of the game, so with proper timing you could easily become even the favourite by just playing an ordinary ace-point game! If you make an advanced anchor, your opponent could easily out-roll you with some lucky numbers and there may then be little scope for him to make major errors. For example, with the opening rolls of 6-2, 6-3 and even 6-4, I usually play 13/7, 13/x against a known weak player! Why is this? Well, lets assume you subsequently make the 18-point after opening with 24/18. Then a very weak opponent may roll some big doubles, get behind you, and you'll have virtually no chance in the race. By holding your ace-point as long as possible, you give your opponent the possibility to make errors until the end, and most of the time he or she will make them! Also, making your own bar point is rather strong against a weak opponent, because it avoids some jokers such as 6-6.

• c) This brings us to a minor third point -- avoid Jokers. You should try not to give your weaker opponent any game-winning jokers, even if its mathematically correct to do so. So you should, if possible, avoid positions, where he has, for example, a few chances to roll boxes and win the race easily. If you don't give him many jokers, he will be doing hard to win, because of his weaker play. [This is why you should also try hard to stop a weak player making an advanced anchor when he or she has a race lead, the opponent is then often one roll away from winning PL]

• d) You shouldn't overdo any of these points, either. Often opponents are not as weak as they first appear and sometimes much stronger than their rating! By the way, I lost against the pimp!

Any comments are welcome!!

Loner, how about an article from you ?? :)

Bernd, alias one_pointer