Wednesday, September 24, 2014

Making sense of BBO skill levels

One of the fascinating things about BBO is the self-assignment of skill level. There are six levels to choose from plus "private". These are World Class, Expert, Advanced, Intermediate, Beginner, Novice. Probably the two most abused are expert and advanced. I've never seen anyone who judges themselves to be World Class to be incorrectly described (perhaps BBO monitors this level, I don't know). But I've seen plenty people self-assign themselves as expert whose claim to expertise in the game of bridge is questionable at best. BTW, I describe myself as expert too. Whether I am in the eyes of my fellow real-life bridge players is certainly open to question. But I feel obliged to do so in practice, by comparison with many other "experts" on BBO. I really have no idea what the criteria for these skill levels should be.

Here's the official BBO description of the levels in each case followed (in italics) by my take based on actual experience:
Novice
Someone who recently learned to play bridge — like, today, er... just now.
Beginner
Someone who has played bridge for less than one year — more like less than a month and may not have played at a club.
Intermediate
Someone who is comparable in skill to most other members of BBO — someone who is reasonably honest about their skill level but who hasn't really progressed much from the beginner stage.
Advanced
Someone who has been consistently successful in clubs or minor tournaments — someone who has been playing too long to describe themselves any other way.
Expert
Someone who has enjoyed success in major national tournaments — someone who can count to 13, especially in the trump suit; may actually know what an endplay is and perhaps has actually pulled one off; knows that underleading an ace at a suit contract is generally a bad idea.
World Class
Someone who has represented their country in World Championships — a real expert.
These BBO descriptions of course beg the question: what defines success? Do you have to win a major national tournament event to be an expert? Or is it sufficient to have placed in the overalls, or got a section top? What about bracketed KOs? Must you have won the top bracket? What about Midnight KOs? What about single-session consolation events like a Swiss? I think it's all just a bit too vague.

So, I've been compiling a list of the types of errors that I see perpetrated by the various levels. Many of these observations come from individual tournaments, or just sitting at a random table (although it's rare to find anyone who claims to be an expert sitting at such tables — I only do it when I don't have time for a real game). So, here goes:

Intermediate:
  • Fails to notice that an opponent has shown out of a suit and/or fails to appreciate that said suit is now going to block if we aren't careful.
  • Thinks that the double of a 1NT overcall is for takeout.
  • Discards potential winners instead of likely losers when declaring a hand (one suit is 97 opposite AJT6 [two tricks are certain], the other suit is AQ87 opposite 543 [82% chance of two tricks]).
  • With ♠Q9654 AK 95 ♣KQJ4, after this auction (IMPs, none vulnerable, dealer is on the left): p p 1 1♠ 2 3 4, prefers to bid 5♣ rather than waiting to see what partner is going to do (result: -100 instead of +420).
  • Bids 1♠ only with ♠AJ62 AJ86 5 ♣KJ64 when (all vulnerable) LHO opens 1 and partner doubles.
  • Raises partner's 1♠ (previous comment) to 2♠ over an intervening double with ♠Q843 75 AK864 ♣Q2. [This pair bids all the way up to 4♠ but opener's partner has KT975 of trumps and wields the axe for a top. (or 8 IMPs if it were teams)].
  • Deals and opens ♠AK8 void AQJT9862 ♣A2 with a diamond preempt.
Advanced:
  • Doesn't notice that this is a 1NT opener, playing SAYC: ♠752 J765 AK5 ♣AK6.
  • Despite having no entry while defending 3NT, and despite dummy positively bristling with winning black cards, continues a third trick in own suit (hearts) which is 100% sure to be won by declarer rather than switching to diamonds which would win his side two more tricks.
  • With this hand: ♠AJ64 J952 Q95 ♣65, and hearing the following auction (none vulnerable, starting on your left): 1♣ p 1♠ p 2♠ Dble, decides that this is a good time to penalize those pesky opponents and passes — result -470 instead of -100 (or -50 or +50).
  • With this hand ♠Q763 K5 AKQJT ♣T3 (vulnerable vs. not) after one pass on right: 1 2♣ p p 2 Dble p p p. Dummy comes down with ♠T954 T643 9 ♣AQ95. The lead is a trump which you have to win in hand. You are very fortunate in having two club stoppers which will give you time to establish a spade trick or two. But the first thing you do is to use up those club stoppers so that you have to use your good trumps to swat clubs instead of establishing your own tricks.
  • Believes that a balanced 18-count is enough to bid a major suit game opposite a single raise by partner at matchpoints.
  • At favorable vulnerability with ♠QJT8652 void AK64 ♣T9 overcalls 1 with 4♠, then after 5♣, double, 6♣ [sic], naturally rebids 6
  • Over 4 by RHO (nobody vulnerable, Goulash hand), bids 4♠ with ♠AK98 53 ♣KJT764.
  • The only way I can do this advanced player justice is to show the whole hand (from an individual):
Note that I made five calls during the auction: two redoubles and three doubles. What fun! If you click on the GIB button, you'll see that as long as I start with a high club (what else?) we are destined for +500. Unfortunately, it's now necessary to switch to a low spade to keep our 500. That seemed contraindicated to me so I switched to the J. Down to 300, although I'm still on lead and now any small black card will get us back to 500. I'm fairly keen to avoid setting up a winner in dummy because I want to protect my partner's diamonds. Alas, I continued a heart. Partner gets in and at trick 5 he plays... a small diamond! A two-trick error. Wasn't it blindingly obvious that declarer has the missing diamonds? At this point, we have gone from +500 (an 83% board) to -490 (a swing of almost 1000 points and a 0% board).

Expert:
  • Holding ♠J97 J93 KQ4 ♣AJ54 (none vul) and having dealt and opened, after the auction (opps silent): 1♣—1—1NT—2♠, passes (not a tragedy because about a third of the declarers in 3NT went down).
  • Holding ♠AKQ53 JT76 742 ♣J (all vulnerable) and hearing partner open 1♣, responding 1♠, hearing double by LHO (a passed hand) followed by pass and 1NT, decides to bid 2♠. [When the same auction happened at another table, a real expert doubled for blood and ended up with 800].
  • Holding ♠J9 Q52 AQ6 ♣AK983, having opened 1NT and hearing partner transfer into hearts and then bid 2♠, bids 2NT [The opponents mis-defend but we still get a poor score because everyone is making 10 tricks in hearts, some in game, some not].
I hasten to add that I have occasionally made some pretty big blunders on BBO myself when not really concentrating. Perhaps these good people that I reference here were talking on the phone, watching the telly, or otherwise multi-tasking. But here's the difference. When I make an error, I apologize to partner after the hand. The perpetrators of the foregoing mistakes seemed to be completely oblivious.

Wednesday, September 17, 2014

Passed hands may make only one free bid

It is a truth universally acknowledged that once you have made a (natural) pass you can no longer mastermind the auction. Indeed, unless partner makes an even more limiting call, such as 1NT, you have made partner captain of the side. So, partner is the one who gets to do any "operating"—safe in the knowledge that you won't hang him. Examples of operating include shaded ("third-hand") openings and "pressure" bids (preempts opposite a passed hand with less offense than usual).

So, just what are your limits as a passed hand in terms of bidding? I would say one free bid or, at the very most, two. Of course, you should answer partner's opening bid normally if you have responding values—that's not a "free" bid. You might even make a 2/1 bid since he will know that you cannot have opening count. And if partner forces you to bid again, you will honor his request. You may raise his suit, even his preempt although you will be careful about doing so since he may have been making a pressure bid.

But if the auction gets competitive, you can't just keep on bidding till the cows come home! I recall a hand from a year or two ago where my partner made three bids after having passed originally and hearing only the one bid from me. The result wasn't pretty but fortunately the details of that hand are lost.

Here's a classic example of what I mean from a recent random game on BBO (I changed partner's name to protect the guilty but used Victor Mollo's favorite epithet):


Admittedly, North made only two bids after he had passed, one of which was not a free bid. But the jump to 5 was itself the logical equivalent of two bids. We can make 3, they can make 3♠. It's hard to be sure what East would have done over 3. Probably with a singleton diamond, he would take the push. What would the effect of a 4 bid by partner have been? Hard to tell. I would guess that the opponents would either go quietly (we lose 1.7 IMPs) or bid on to 4♠ (we gain about 5 assuming we don't double which would be more like 10 or 11). As it was, they didn't double 5 and we "only" lost 4.3.

In the following hand, the BBO Robots were kind enough to furnish me with another example (see the following).



The first diamond call by partner isn't really a free bid, given that he has 8 points. To pass would be Quixotic. The support double in round two confirmed a full opening bid (11-21 hcp) with three-card diamond support, and suggests a balanced hand, although presumably doesn't deny a 5-card club suit. The likely result of the double, it seems to me, would be +200 or -790. Predictably perhaps, we suffered the latter fate and a (shared) bottom.

Personally, I don't think the North robot has any business making a double with no surprises in store for the declarer, and having already made two bids after passing at his first turn.

So, to conclude: once you are a passed hand, resolve all marginal calls by taking the more conservative action (often that would be pass). The only justification for bidding aggressively is having very good shape. Doubling for penalties without having the goods, is just asking for trouble.

[Note: 2014/09/18: I changed the title back to the original]

Friday, September 12, 2014

Eight is enough

My young partner Alexander and I couldn't get our teammates from the most recent Eight is Enough at the Watertown, MA sectional (which we had won), but we got another very good pair: Steve and Vincent. We began the evening in a three-way whence we emerged with the disheartening total of 9 victory points. One of these teams ended up in second place, but that's small consolation. So, now we had only two more rounds—we would need blitzes in both matches in order to finish in the overalls.

One rather odd aspect of this event was that every single board in the first two rounds was played (at our table) in a no-trump contract: eleven at 3NT, one at 1NT. The first board of the next set was also a notrump contract. Together with three more in the final set, we ended up playing 2/3 of all contracts without the benefit of a trump suit.

Things turned around for us in the third round: we won by 39 IMPs. There were three pushes and three swings in our direction. Two of these were grand slams bid by our teammates but not by our opponents. We can't take any credit there. But here is a hand that merits some comment: my hand was ♠JTxxx Kxx x ♣QTxx. We were at favorable vulnerability and I passed as dealer. The next player also passed. Alexander opened 1♠ and the next player doubled. One of my favorite tactics is to feign strength. I could perhaps have jumped to 4♠ but who knows whether they might successfully reach 5 or 5? Instead I splintered with 4. The next player passed and Alexander signed off in 4♠. We were not doubled and drifted off two tricks for -100. Meanwhile, our teammates were able to bid a comfortable 3NT on their combined 25 hcp. Here's the whole hand...


This is thinking out of the box, for sure. Third-hand bidding is an art in itself. If prospects for your own game look dim, and the vulnerability is favorable, it's reasonable to try to get in the way of their game. So far, spades hadn't been claimed and, if I was to end up on lead, a spade out might be quite helpful, thus the 1♠ bidding card. I was as surprised as anyone to find out that declarer had only three. And I will have to be careful if this becomes a habit ("Alert: at least three spades, may have a really bad hand"). But coming as it did with three boards still to play,  it was a great shot across the bows of the opponents.

Oh yes, we added 38 VPs in our last two rounds and ended up tied for fourth place.