Moves can be foudn here.
This was an excellent round. We obtained several turnovers and removed a full suit of Diamonds. We also cleared the mess in Column 1. However, we couldn’t quite make the most of our good fortune with the dreaded one-hole-no-card scenario occurring after taking all the easy good cards.
At the risk of sounding like a broken record, one-hole-no-card usually indicates we failed to take maximum advantage from a position of strength. But a poor start meant we never had a position of strength to begin with. Attempts to avoid 1HNC would have entailed some other strategic risk – therefore I cannot declare that Bart had a finer muck-up. Spoonerisms aside, a detailed analysis is unfortunately beyond the scope of this post. In any case our position could easily have been much worse. Game on!
Bart played a very interesting gambit at the end. Unable to convert the empty column into an extra turnover, he chooses to split an in-suit build to at least maximise the chances of recovering an empty column. He feels the position is sufficiently good to justify playing safe – and if we do get an empty column or two then hopefully the in-suit builds will take care of themselves. I can’t claim this is a bad decision, but I’m sure we’ll find out soon 😊
Meanwhile, we hope IM Bug a speedy recovery after his accident.
One thought on “Round 3 Summary”
I find that I have to leave a reply to a thread in order to get alerted to anyone else replying to that thread, at least with the WordPress setup I’m using. I get alerted to new threads, and to new replies where I’ve made a reply already, but not on old threads if I haven’t weighed in. Occasionally I’ve missed something important that way. Hopefully my new effort to avoid that situation won’t lead to make very many useless replies…
Good summary of the situation, I think. But somehow “1HNC” doesn’t feel so bad here as in other circumstances. With 5 columns empty for but one atomic unit, getting 2 or even 3 spaces in short order after the deal seems pretty likely to me. In general, when a column’s got a king in it, you necessarily give up a space to uncover a card, and I don’t see as much is lost if we delay that for a round. Columns 3 and 6 are kind of complicated, but 7 and 9 require gobbling up just one scarce resource to uncover a card, and 8 requires just two.
GM has been educating us about 1HNC for some time. The problem I wish to highlight needs a snappy name… “Can’t Get Started” or CGS? Surely others have seen this many times? You’re in a good situation, and you get a new deal of cards that looks promising. You see many avenues of progress, if only you had a little breathing room. But you can’t make a single space, and you can’t get started. One special case of this is the final deal which simply allows no moves. If we ignore the chance that a landing card will happen to be in-suit with the card it falls on, my simulation shows just above half a percent of the time there will be no moves available — 0.58%, roughly. Your perfectly played game goes up in smoke. However, there are many more examples of cases that are almost as bad — you can move one card, or two cards, but then you are stuck at that point. I would guess it’s more like 5-10%?
Now, getting a space after a deal always requires moving at least one thing — what landed on your top-candidate column. And then it requires having a place to put it. With 5 such top candidates, as in this deal, your chances are pretty good but not at all certain. My contention is that the A-7 in column 2 increases significantly the chances. Because if we can expose that sequence, there are four other columns with a latent home for it if we can only move what’s on top of them. It should have a similarly beneficial effect on the ability to create 2 spaces instead of 1.
I could test this hypothesis by hand, by getting my program to give us a random 10 cards of those left, and after 100 trials or so comparing the number of spaces we can produce. My plan of putting the A-7 in the space can’t be considered in isolation. What is the best competing alternative that people felt would have been better? I have one candidate from IM Bug. GM? Of course I can’t just measure spaces produced, but note on the other side of the ledger benefits to the alternative move.
Even if my hypothesis is correct, it would also be of great interest to know how big the effect is.