Kaleidoscope Series – Lesson 6

IMPs versus MatchPoints

-------------------------------------

      The difference between MatchPoints and IMPs is simple enough.   In MatchPoints, your score is put on a list with everyone else's score.   You get one MatchPoint for every person you beat on this list, a half a MatchPoint for everyone you tie on that list.   Hence, if you do better than the 12 other people who played the hand, you would get 12 MatchPoints.   An average result would be 6 Matchpoints, (i.e. , "MPs") if 13 pairs play the hand.   If you were to beat 2 pairs and tie 3 others, you would get 3½ Matchpoints: 2 for the two you exceeded, a half each for the three that you equalled.
      On OKBridge, this is simply converted to a percentage with 100% going to anyone who beats all of the other pairs, 0% to anyone who registers the worst score.   In mathematical terms, we are merely converting a percentile into a percentage.

Difference in
Raw Scores
 IMPs

0- 10
20- 40
50- 80
90- 120
130- 160
170- 210
220- 260
270- 310
320- 360
370- 420
430- 490 10 
500- 590 11 
600- 740 12 
750- 890 13 
900- 1090 14 
1100- 1290 15 
1300- 1490 16 
1500- 1740 17 
1750- 1990 18 
2000- 2240 19 
2250- 2490 20 
2500- 2990 21 
3000- 3490 22 
3500- 3990 23 
4000+ 24 
      In International Match Points ("IMPs"), we take the difference between the scores that you and your opponents have achieved and convert that difference to an arbitrary number of "IMPs" according to this scale:
      On OKBridge, seeing percentages on the scoreboard is a sign that we are playing MatchPoints.   If decimals, the game is IMPs.
 

----- Quiz #1 -----

     1.   Drew "CANOOZ" Cannell, upon returning to compare scores with his team mates, announced that on Board #1 they would "lose 'em all".   How many IMPs did he mean ?

     2.   At which form of the game are there more likely to be "pushes", resulting in no score for either pair ?   IMPs or Matchpoints ?

     3.   At which form of the game is a "dark horse" more liable to stage an upset win ?   IMPs or Matchpoints ?

----- MPs versus IMPs and IMP Pairs -----

      IMPs were originally designed for team competition.
IMP pairs is simply an averaging of your opponents' results before such a comparison takes place; you are pitted against the "average opponent", when playing IMP Pairs.
      Playing IMPs, if you make 430, while others make 420, you have a raw score difference of a measly 10.   This is worth ZERO IMPs.
      Were you playing MatchPoints, though, 430 would beat ALL the other pairs with their 420 scores; you would receive ALL the MatchPoints on such a board.   (This, by the way, explains the popularity of NoTrump contracts among MatchPoint players.)
      Similarly, if everyone else scores +100 on a hand, you will get all the MatchPoints with any score above +100: +110 would get you the same fine result as +1100 would !  

      At MatchPoints, then, you only need to beat the score achieved by others who play the same hand.   It does not matter by how much you beat them.
      At IMPs, though, it is not enough to beat the others who hold the same hand as you.   In order to make headway, you must beat them by as much as possible !   If you score +110 as opposed to +100, you will score a lot of MatchPoints, but a net difference of (110 – 100 = ) 10 would net you zero IMPs.
+1100 versus everyone else's +100, on the other hand, is worth (1100 – 100 = ) 1000 net points, or a whopping +14 IMPs.
 

----- The Effect on Strategy -----

      How does this difference in scoring methods affect your decisions at the table ?   Plenty !  
      At MatchPoints, making an overtrick in an undoubled, non-vulnerable 4 while others make only 10 tricks means a "TOP" board (i.e. , all the MatchPoints available).   But at IMPs, it means only (450 – 420 = ) 30 points, or 1 measly IMP !   With so few IMPs at stake, this would be considered a "flat" or "near-flat" board at IMPs.   Many such hands which are so exciting at MatchPoints ("Can you make that critical extra trick ?") are a total yawn at IMPs !  
      On the flip side, going down in 4 while all others make 10 tricks is a "BOTTOM" board, or zero MatchPoints.   At IMPs, though, you have just lost (420 + 50 = ) 470 points in total, which converts to a disastrous –10 IMPs.
      From this, we discern that, at IMPs, you would almost NEVER risk the contract for an overtrick.   The risk (–10 IMPs here) far outweighs the gain (1 IMP).   At MatchPoints, however, that overtrick may mean the difference between a top and an so-so result, so you MIGHT, if you see the chances as better than 50%, put your contract at risk to try for an overtrick (especially if you think everyone will be in the same contract as you).
      Similarly, on defence, you would always cash the setting trick at IMPs to ensure defeating the contract.   But, at MatchPoints, an extra undertrick might be worth the risk of NOT cashing the setting trick in the hopes of setting the contract TWO tricks !  

      "So," you might ask, "if going down in a game, which everyone else makes, is a disaster at both forms of the game, what's the difference ?   Isn't it just one bad board ?"

      The difference is that at MatchPoints you can recover by making an overtrick in 3 on the very next board, while everyone else makes only 9 tricks.
      At IMPs, this would only recover one of the 10 IMPs you lost going down in 4 earlier !   Because you need hands where big numbers are at stake, it is much harder to recover from disasters at IMPs, because there are far more "FLAT" OR "NEAR-FLAT" boards at IMPs !  

----- Bidding Versus Play -----

      In MatchPoints, "the play is the thing".   An overtrick or an extra undertrick means everything, if it distinguishes your score from everyone else's.   There are few "flat" boards at MatchPoints.   In a typical 24-board session at MatchPoints, you may have 2 or 3 "flat" boards.   In a typical 24-board match at IMPs, you may have 12-18 such "flat" or "push" boards.   :(
      At IMPs, bidding is paramount.   Miss a slam or a makeable game, and you're in trouble.   Squeezing out an extra overtrick or extra undoubled undertrick via careful play on the next hand will NOT recover your loss.
      The above distinction explains the expression:
"IMPs is for BIDDERS; MPs is for CARD-PLAYERS !"
It also explains why MatchPoints is a much more intense game where every board has equal importance.   A delicately bid and played 7NT contract will not gain you any more or any fewer MatchPoints than the lucky overtrick you made in 1 the hand earlier.
      At IMPs, however, you can safely fall asleep during the 1 hand — as long as you wake up for the 7NT one !   In this regard, by the way, IMPs is closer to Rubber bridge than MatchPoints is.

      Bottom line: At IMPs, the big "swing" hands with a lot of points at stake (e.g., difficult games, slams and doubled contracts) matter most.   At MatchPoints, all boards are equally important.

----- Competing for the Part Score -----

      The area where the two games diverge most is in competitive part-score bidding.   Consider this situation: You have bid 3, rather confident of making.   The vulnerable opponents compete to 3.   You can't be 100% certain, but your instincts tell you that 3 will likely go down one.   Should you Double ?
      At IMPs, the difference between 3 down one doubled versus undoubled is (200 – 100 = ) 100, or three IMPs.   If they make, though, the difference is (730 – 140 = ) 590, or eleven IMPs. Hence, you'd better be VERY confident of 3 going down, before Doubling it at IMPs !   Otherwise, you would do better to pass 3.   If it does go down one, the board will be flat, since (110 – 100 = ) 10 points means zero IMPs.
      At MatchPoints, however, you were slated for +110 in 3, so +100 in 3 will not compensate you.   You may get a bottom board, if everyone else is allowed to collect +110 in 3.   Hence, to "protect" your +110, you might Double 3 and hope to nip it one for +200 and a great score at MPs.

      Bottom line: save those close Doubles for MatchPoints, not IMPs !  
      At MatchPoints, you will often strain to balance the opponents into an unmakeable contract.   Nothing does your game better than pushing the opponents to 3 and collecting +50, while everyone else is –110 against 2.   If balancing against their 2 risks the odd –800 because of a trump stack against you, so be it.   As long as you collect more +50s than –800s, you'll do fine in the long run at MatchPoints.
      At IMPs, though, those –800s had better be far fewer.   Getting +50 versus –110 is worth (50 + 110 = 160) four IMPs.   –800 versus –110 loses (800 – 110 = 690) twelve IMPs, and such disasters are harder to recover from, with fewer decisive boards at IMPs than at MatchPoints.   :(

      Bottom line: save your balancing heroics and aggressive part score competing for when you're playing MatchPoints, not IMPs !  

----- Sacrificing -----

      Let's take another scenario.   They have bid 4, vulnerable.   You figure that they have a better-than-even chance of making it.   Your side, on the other hand, can bid 4, get doubled, and go down –500, 120 points better than you will spit up for 4 if it makes.   Should you sacrifice in 4, then ?
      At MatchPoints, the answer is yes.   If they do make 4, you are have improved your score from –620 to –500.   Yes, if 4 goes down, your –500 will not look good opposite everyone else's +100.   But as long as 4 has a better than 50% chance of making, you will gain more often than you will lose in the long run.
      At IMPs, the answer not so clear.   4 will only improve your score by (620 – 500 = 120) three IMPs.   If, on the other hand, 4 goes down one, you will have foregone the +100 for –500, a 600 point difference for minus twelve IMPs.   Since your sacrifice has to be "correct" (i.e. , 4 has to make) 4 times to 1 [for those 3 IMPs to equal the 12 IMPs you might be missing], 4 should have an 80% or better chance of making, before you should sacrifice against it !   At MatchPoints, your sacrifice only had to be right more than 50% of the time to pay off in the long run !  

      Bottom Line: At MatchPoints, sacrifice against anything the opponents bid confidently.   At IMPs, sacrifice only against those contracts which are underwritten by Lloyds of London !   :)

----- When To Play SAFE at MatchPoints -----

      After an auction such as 1NT:3NT, it is reasonable to assume that virtually everyone will be in the same contract.   This being the case, at Matchpoints we must take as many tricks as we can. If this runs a slight risk or going down, so be it.
      But what if we are in a contract that many in the field might not find, we should play safely to make — just as we would at IMPs.   But how can we know where other pairs will be in a given hand ?
      If, during the auction, either you or your partner faced a "crossroads" where either of two actions looked attractive, it is fairly safe to assume that approximately half of the people took the other fork in the road.   If your decision turns out to be a makeable contract, while others will be less profitable, make your contract.   In essence, then, we say to ourselves: "Hmm, we seem to have stumbled into a good contract here.   Better make it !"
      If it appears that the others might be in a more profitable contract, we may need to stretch for overtricks.   The typical case is one where we are in 3NT, while others may be in 4 or 4.  
If we calculate that those in the other contract will be slated to make 420, we need to try for an overtrick to make 430 in 3NT.   Here we are saying to ourselves: "Gotta find some way to keep up with the Joneses !"
      At Matchpoints especially, we need to develop an additional routine to our normal "count our tricks" work habit.   We need to ask ourselves:
"What is everyone else likely to be in here ?  
 And how many tricks are they likely to take ?"
Then, we need to find a way to beat that alternate score.
Practice doing this on every hand.
----- Questions -----
  1. You are in 4 doubled and vulnerable at IMPs.   You could make an overtrick by playing a Club, but a 5-0 Club break (3.94% chance) will produce a ruff for down one.  
    Should you risk your contract at IMPs here ?

  2. You are in 4 doubled and vulnerable at MPs, a contract that few others will find.   You could make an overtrick by playing a Club, but a 5-0 Club break (3.94% chance) will produce a ruff for down one.  
    Should you risk your contract here at MatchPoints here ?

----- Rubber and Board-A-Match -----
      In Rubber bridge, the pair with the better cards will almost invariably win.   Even a few 620's and 1430's will more than compensate the –50's and –100's that the other side will score for defeating contracts.   At Rubber bridge, then, "cardracking" (i.e. , the ability to hold good cards) is of paramount importance.   A good rubber bridge player will never risk hir contract for an overtrick.   In this regard, then, Rubber is exactly like IMPs.
      But because Rubber is played by pairs, not teams, one would more accurately state that Rubber is like IMP Pairs.   Aside from the conversion from raw scores to IMPs, though, what is the difference between the two ?  
      This contrast can be expressed in one word: "equity".  
At IMP Pairs, all the pairs sitting in our direction will be holding the same cards.   Our task at IMP Pairs, as in life, will be to do the very best that we can with the cards that we are dealt.
      In a sense, then, Rubber could be considered "IMP Pairs without a field" of players with whom we would compare our results.   This explains its decline in popularity relative to IMPs.
      In every Fall ACBL National event, there is an Board-a-Match event called the "Reisinger".   Board-a-Match is a team event, where whichever team does better than the other wins the board and gets one point.   Ties or "pushes" don't count (i.e. , each team gets a half a point).   So if one team makes 4 for 420 at one table, while their counterparts on the other squad at the other table make 430 via an overtrick in 3NT, the second team wins one point.  
      In this way, while Rubber may be considered "IMPs without a field", Board-a-Match could be viewed as "Matchpoints without a field".   As such, notwithstanding the unquestioned prestige of the Reisinger, Board-a-Match is the least common form of the game.
----- The Effect on Systems and Conventions -----
      An effective pair may prepare differently for an important event, depending upon the form of the game.   Before a Matchpoint event, we might wish to brush up on our playing skills by leafing through Louis Watson's classic tome, "Play of the Hand" and on defence.   We may review such works as Hugh Kelsey's "Killing Defence" or Rainbow Lessons #10 ("Defensive Signalling") and #11 ("Defence").   We would also want to concentrate on our competitive bidding.   We might look at references on the Law Of Total Tricks (e.g., "Following the Law", "Points Schmoints" and "To Bid or Not to Bid"), overcalling (as with Mike Lawrence's "The Overcall") and on Doubles (e.g., Rainbow Lesson #8: "Competing with Doubles" and, perhaps, KaleidoScope Lesson #3: "Defensive Doubles").
      In preparing for an important IMPs contest, we would tend to concentrate on our slam bidding.   Some pairs will play 2/1-GF or Forcing Club system at IMPs, SAYC at Pairs.   At IMPs, some pairs will even play 2/1-GF in first and second seat (when slams are more likely) and SAYC in 3rd and 4th seat (when Partner's pass dims our slam hopes).   Some 2/1-GF pairs will play a forcing 1NT response to 1 or 1 at IMPs, but prefer to retain a non-forcing 1NT response at Matchpoints.   This will require that they use a gadget such as a non-forcing 2NT reply (10-12, doubleton in Opener's major) or the Toddler 2 response (which can be investigated at: http://www.escape.ca/~cpw/toddler.htm) while playing Matchpoints.   Others may prefer a weak 1NT opening at Matchpoints, strong 1NT's at IMPs.
      Many expert pairs play DONT at IMPs, CAPP at MatchPoints.   To understand this, we need to review these two conventions:
DONT: Over their 1NT opening bid:

WestNorthEastSouth
1NTDblePass2C
  Double = 1-suited takeout.  
  Advancer asks via 2.
WestNorthEastSouth
1NT2C
2D
2H
2, 2, or 2 = 2-suited
with the bid suit
and any higher ranked suit.

Cappelletti: Over their 1NT opening bid:

      Double = Penalty
      2 = Any 1-suiter.    Advancer bids 2 to ask.
      2D = Both majors.
      2 = Hearts and an unspecified minor.
      2 = Spades and an unspecified minor.

      Why play DONT at IMPs but NOT at Matchpoints ?
Consider the 2 Overcall from this Advancer's point of view:
   S Q 10 x x    H x x    D Q x x    C J x x
WestNorthEastSouth
1NT2DPassPass?
      [Playing DONT,] removing the 2 overcall to a Major will work out well, if that Major happens to be Spades.   However, it may be disastrous, if Overcaller's Major is Hearts.   Hence, most players will PASS here.   If Overcaller's Major turns out to be Spades, this will not be a disaster at IMPs: +90 versus +110, if both contracts make 8 tricks.   At Matchpoints, however, [this will indeed be a disaster for us, and ] we MUST be in the right strain.
WestNorthEastSouth
1NT2HPass2NT
Pass3D
WestNorthEastSouth
1NT2S
      Playing CAPP, Overcaller would bid hir Major directly (e.g. , 1NT-2 or 1NT-2).   If Hearts, the above Advancer will remove to the minor via 2NT.   If Spades, the pair will have found its major suit fit.   DONT, then, is SAFER — a critical consideration at IMPs.   CAPP will find any major suit, IF there is ONE, but will risk going to the 3-level, if the pair's fit is in a Minor.
      All of these adjustments address the basic needs of the partnership.   At IMPs, we need to avoid losing big amounts, while generating any that are available to us.   We don't "sweat the small stuff" at IMPs.   Only at MPs, do we need to fine tune our bidding to get the very best score.
----- In Conclusion -----
      Matchpoints, then, is a simple matter of edging our the opposition.   It is an "EITHER/OR" proposition.   You either beat their score or you don't.   The effect is that games, doubles and sacrifices need only a 50% chance of success.
      IMPs is a game of AMOUNTS where the risk-reward ratio is usually more complex than a straight 50-50.   Rather, it is a game of percentages.   For example, a vulnerable game is worth bidding at IMPs, if it succeeds 38.71% of the time.   A non-vulnerable game contract requires only a 45.24% chance of success.  
      This is because the cost of going down one in, say, 4 is the loss of the 140 for 3, plus the 50 point penalty.   Compare this 140+50=190 we "lose" to the +420 we would make if the game were to make and we come to 45.24%.   These more complex percentages, in contrast to the 50-50 world of Matchpoints, gives rise to the expression:
"Matchpoints is a coin toss; IMPs is a crap shoot !"
      And that, in a nutshell, is the difference between IMPs and MatchPoints.   :)
---------- Final Quiz -------------
  1. After 1NT:3NT, Partner leads the 9 and we hold:
    Me S K x x    H A Q 4 3 2    D Q 9 xx    C x .
    Dummy    S x x x    H K J 5    D J 10 x x    C A K x x .
    1. Play at IMPs when Declarer plays the 5 from Dummy ?
    2. Play at MPs when Declarer plays the 5 from Dummy ?

  2.   K Q x x x x Q x x x x x x,    None vul.
    WestNorthEastSouth
    1S2H2S3H
    PassPass??
    1. The auction proceeds back to you as Responder.   What should you do here at IMPs ?
    2. The auction proceeds back to you as Responder.   What should you do here at MatchPoints ?
      WestNorthEastSouth
      1S2D2S3D
      PassPass??
    3. The auction proceeds back to you as Responder.   With the opponents bidding diamonds this time, what should you do here at IMPs ?
  3. Which of the following are more common at MatchPoints than at IMPs ?
          (a) Competitive Doubles
          (b) Help Suit Game Tries
          (c) Balancing Doubles on marginal values
          (d) Safety Plays
  4. We have bid 3, confident of +110 there.  
    1. The opponents, who are NOT vulnerable, compete to 3.
      We calculate that 3 will go down one slightly more often than 3 will make.   Down two is very unlikely.   Should we Double ?
    2. The opponents, who ARE vulnerable, compete to 3.
      We calculate that 3 will go down one slightly more often than 3 will make.   Should we Double ?


Valid HTML 4.0 Transitional saved from url=http://www.firesides.ca/kaleido6.htm