Slot, Split, or Safe?

The position below is very common, and clearly, Red rolled a 4-1 and has three choices. 

1) split his back checkers
2) slot his 5 point
3) play safe

The answer to this question can change very quickly depending on the score of the match, the position of the cube, and of course, exactly where all the checkers are.  At times, any one of these plays could be right.  In this situation, however, only one of the three choices is clearly right.  Decide for yourself, and then scroll down for explanations from two of the top experts in the world I rely on for insights into plays like this.

As you can see from the extremeGammon evaluation, it is clearly right to split the back checkers in this instance.
I asked Perry Gartner if there is a rule of thumb or some general clues that tell us when this is the best play or when to make the other plays.

Here was Perry's response:

You are asking a question that needs a long article or book chapter to answer fully. I haven’t written that chapter yet, but neither has anyone else although there are lots of kernels out there that could be put together to answer this question. I will give some general guidance in discussing this problem.


Let’s look at the considerations and virtues of the game plans.


1.     Racing.  Red pip count is close to even.  Racing is viable if it can be consummated. Freeing the back checkers from being primed is essential to advancing the racing game plan. When checkers are in your opponent’s home board RGP(racing game plan)  can be initiated by 4 potential directions of play.  Priming, attacking, holding, and running.


When you prime your opponent you commit part of your resources to containing him and this effort enhances your prospects for escaping. That’s because at least some of his checkers are not available to stop your advance out of his home board and if he wants to thwart your prime he usually needs to play to confront your prime on your side of the board.  You are constrained by your structure to making immediate priming moves. The 6 and 3 need a lot of help, slotting is very dangerous and this roll doesn’t create conditions that allow you to catch up quickly enough.


The same can be said for attacking.  In this position you are way behind in the priming battle and your potential for attacking in your inner board is not present.


Advancing your anchor puts you in a position to run with the right double or escape one at a time under the appropriate conditions.


Escaping into the outfield, and hooking up to points made in the outfield is another way out.


2.     Priming and attacking game plans. At the early stages, making inner points (depending on the order of course) advances both these game plans. You are behind in this aspect.


3.     In this position the best anchor possibilities are the bar, or 22 point. I think of the 22 as a borderline advanced anchor. We are tending more to not being satisfied with it as it generates powerful double/takes where your equity is close to as bad as it gets without passing. But even the 23 beats the 24.


 Splitting has enormous defensive contact potential. Any single checker builders brought down from the opponent’s midpoint are subject to many more hitting numbers you get compared to when not split. 


The play of  moving to your own 8 also has defensive value in that your 8 point can remain made when the 3rd checker leaves it.  The 8 is a conduit to bringing checkers into the zone, a way-station so to speak is one of its central roles.  When fortified with a 3rd or 4th checker it inhibits movement of the opponent’s checkers in your home board. That’s why moving to the 8 is the second best play.


When your opponent’s home board is much better than yours, splitting and slotting at the same time is massively wrong as your forces are just too weak to contend with his power. Your home board is not enough of an inhibitor, as it will not exact a sufficient price for his going on the attack and hitting loose wherever he can. You need to hoard your rolls for use in his home board and not worry about what’s happening to your forces on the other side.  In fact look at it as splitting is a confrontation with you on the short end of the stick, but you are doing it to bite the bullet now rather than waiting for later where is promises to be more dangerous. 

There is the concept of equalizing positions before you split. It means focus on making a strong home board and later on split where a hit back will extract a price or his home board will disintegrate before yours does. This theory is most practical when you are significantly down in the race, isn’t usually as good when you are even in the race, and when your opponents structure is far superior to your own you are a big dog to equalize positions. That would be demonstrated in this position by either just coming down to the 8 or going to the 9 and slotting the 5.

Would you have preferred 5-1 to the roll of 4-1?  The answer is "yes" despite the enormous value of the 9 point, if it can be made, and the larger number of points is can participate in making.  But the danger of another blot, even an indirect at this juncture is preferable to avoid if you can.  The ace is mandatory to split so the 9 becomes best with 4-1.

When you split, the checkers on your opponent’s 8 and 6 are the only potential attackers, other than some doubles.  8 checkers in the zone are 2 less than 10 which is used as signal of extreme danger.  There are 27  4’s and 6’s but not enough ammo to have a robust follow through. The mid is too far from the action to have its checkers play a significant role in the early exchanges in most variations of this position.  The blot on the 8 is a condition that your opponent wants to remedy.   He wants to make it or hit you.  If he makes it, his prime and attacking chagces are strengthened so don't let him have it on a sliver platter.  If he hits you on the 23, you rate to come in, do hit him back sometimes, and end up on the 22 sometimes as well, both pluses for your game.

After reading Perry's explanation, Dana Nazarian was kind enough to add his thoughts as well:

Interesting.  To me, this was a very easy problem.  Here's my take:

Perry's explanation shows the depth of knowledge he has.  The decision to slot or split is often governed by whether or not you want to play a prime/prime game or not.  Here, nearly every factor argues against a prime/prime game, namely:

  • White already has a much better prime than Black, so why go for a game plan where you're outclassed.
  • Black has already made his 3 point.  That makes it extremely difficult because he would be committed to making both his 4 and 5 point.  He would have to roll pretty precisely to do that.
  • Even IF (big IF)  Black successfully makes a decent counter prime, the game is still a toss-up since the race is even (in prime/prime, you want to be losing the race so your timing is better).

Here's 2 more reasons why I don't like the slot in this case:

  • .  Lack of covers.  Even if white misses the shot,  you're not a lock to cover it.  If you throw a 3 and have to break your 8 pt, its good, but not that great.* Slotting is often much more attractive when you essentially get a HUGE advantage when missed (e.g. like when you can effectively cube someone out without even rolling in an extreme example).
  • And finally, the most obvious reason- 17 shakes hit!  In almost all variations, this is a terrible turn for black.

EVERYTHING is saying don't slot, so it's out of the question.

Then the question is, split or safe.  Here's my take on that...splitting is very good (actually essential)  since:
1.  White is forming a very good prime.  If you don't do something fast, you will soon be dead meat.
2.  White has a better board and could also attack, but lacks wood in the zone.  Now is the best time to split.  Waiting even one more shake invites severe blitzing danger.  You can split now without a lot of fear of attack.
3.  You put pressure on white's blot on 8.
4.  The split gives you a builder at much less risk.  I think the risk/reward tradeoff here is about right.. if you split and are worried about getting hit on your 9 pt, even if you do, you will have several shots back from the bar and a chance to at least make a more decent anchor than the 24 pt.

In general, when making the decision of split vs. safe, I tend to take calculated risks early when my game is behind.  That's exactly the case here. 

So, here's the my SIMPLIFIED take-away of this problem:
1.  Slot or not?  - Do you want prime vs. prime?
2.  Split vs safe? - Do you need to take a calculated risk?