The following scenario occurred in a tournament last weekend. The TD was using a computer pairing program (I believe Swiss Sys).
Round 2 of swiss. Four players have 1.0/1. #1 = 2279, had black in round 1. #2 = 2062, had white in round 1. #3 = 2003, had black in round 1. #4 = 1846, had white in round 1.
The pairings posted were 1-2 and 3-4.
My question is why the computer chose 1-2 and 3-4 instead of the more logical 1-4 and 3-2? The only reason I can see is that the program thought a 59 point interchange of #2 and #3 was preferred over a 157 point transposition of #3 and #4.
However, that interchange violates the basic top half vs bottom half structure of swiss pairings. As stated on page 150 of the 5th edition rule book: “While interchanges are sometimes necessary, they should not be used if sufficient transpositions are possible.” The transposition of #3 and #4 seems to comply with the 200 point rule.
What do you think? Should the pairings be 1-2 and 3-4 or should they be 1-4 and 3-2? Relevant USCF rules: 27A and 29E.
Note that this generic situation occurs somewhat frequently. It happened again the next day in the accelerated schedule of the same tournament. It could also easily happen in the 4th round of a big money event where four co-leaders are all 3-0 (two had WBW and two had BWB).
Please look at 29E5e. It answers your question explicitly. An interchange that satifies the 80 point rule is to be prefered to a transposition that does not (even though the transposition DOES meet the 200 point rule). If both the interchange and the transposition are within 80, the transposition is to be prefered (even if the transposition is 79 points and the interchange is only 1 point). There are examples in the rule book that clearly show this.
29E5e: “A transposition that satisfies 29E5a, The 80-point rule, should be preferred to any interchange.”
You are confusing terminology. The 59 point switch of players #2 and #3 is called an INTERCHANGE (switch players between top and bottom half) while the proposed 157 point switch of players #3 and #4 (switch players within bottom half) is called a TRANSPOSITION. Rule 29E5e does not apply because there is no transposition under 80 points.
No, I DO understand. Please read the text that follows this rule. There are examples that cover exactly the case you are asking about. Please just read the rule book.
The fact that the transposition (157 points) DOES NOT satisfy the 80 point rule, but the interchange (59 points) DOES is the subject of one of the examples in this rule. The rule book states in black & white that the interchange is to be prefered in this case.
29E5e Example 2 makes it crystal clear. A transposition that does not meet the 80 point rule does not have priority over an interchange that does meet the 80 point rule. So the interchange is to be prefered in that case (because it’s the smaller change).
Good! This is exactly the problem with the rule book. Example 2 on page 151 seems to me to be contradictory to other sentences written earlier. Let me elaborate.
Page 150: “While interchanges are sometimes necessary, they should not be used if sufficient transpositions are possible.” The 80 and 200 point rules provide guidelines for what transpositions are allowed. Clearly the 200 point rule applies here because we are talking about equalizing colors.
Even Example 2 states that the 200 point rule does apply. However, Example 2 claims that it is preferable to make a small interchange under the 80 point rule than to apply the 200 point rule on a transposition. Where else can I read about that? And if that is indeed what Example 2 shows, then (imho) it contradicts the above quote from page 150 and also violates the structure of rule 27A.
Can anyone show me a precisely written rule (not an Example) which states that interchanges between 0 and 80 points are preferable to transpositions between 81 and 200 points, assuming both have the same effect on color allocations?
You can’t take a rule like 27A3 and apply it without taking into account the other rules. The whole idea of an interchange is an allowable exception to 27A3. Sure 29E5d says that interchanges aren’t to be used if adequate transpositions are available. But it goes on to clarify that statement by referencing 29E5e, which makes it clear that the preference applies to transpositions within the 80 point limit. Maybe the text of the rule isn’t as clear as possible, but why say that transpositions are preferable WITHIN the 80 point rule if the intent had been that legal transpositions were ALWAYS preferable? CLEARLY transpositions weren’t meant to ALWAYS be preferred. Since the rule might not be completely clear by itself (though I didn’t think it was too bad), a couple of examples were added to make sure there wasn’t any misunderstanding. One example showed when transpositions would be preferred and one showed when they would not.
To me, this wasn’t really that hard to find. It took about 2 minutes to follow the references from rule to rule. Those references to 25E5e are in several places and were put in the rule book for a reason.
So here is when you should prefer a transposition and when an interchange (assuming that BOTH are within the appropriate 80 or 200 point limits and are equally effective at solving the color allocations):
Case 1 The interchange is within 80 points and the Tranposition is within 80 points – the transposition is to be preferred.
Case 2 The interchange is within 200 points (but not 80) and the Transposition is within 80 points – the transposition is to be preferred.
Case 3 the interchange is within 80 points and the Transposition is within 200 points (but not 80) – the interchange is to be preferred.
Case 4 the interchange is within 200 points (but not 80) and the transposition is within 200 points (but not 80) – there is no stated preference in the rules, but it seems clear to me that you should use the one that is the smaller change (since 29E5e DOESN’T give an automatic preference to the transposition).
Rule 29E5e states a particular scenario when transpositions are preferred over interchanges. Are we supposed to read more into that rule than the words say?
For example, I still do not see a specific rule that states that interchanges within 80 points are preferable to transpositions between 81 and 200 points. I only see a blanket statement that says that if a transposition is legal (e.g. within 80 or 200 points, depending on the exact scenario), then it is preferable to any interchange. If no transposition is legal, then the TD should investigate possible interchanges.
Your outline of four cases is far clearer than the rule book, imho. However, I’m not willing to accept the full accuracy just yet. As far as I can tell, the rule book only explicitly deals with Cases 1 and 2. Case 3 is the subject of this thread and Example 2 on page 151. And we could even discuss Case 4 in a different thread (but I won’t).
I also won’t address the issue of what is “right”. In my example, poor player #2 would have faced an 1846 opponent as his natural pairing but, due to the color situation, ended up playing a 2279 instead! He can certainly make the argument that such a tremendous switch is harsh, especially since an alternative pairing against a 2003 was possible.
I don’t quite see the problem. The last paragraph on p. 150 says, “If pairing a round in which 29E5b, The 200-point rule, is used … an interchange involving a smaller rating switch than a transposition should be preferred unless the transposition satisfies the 80-point rule.” In the case you give, you can equalize colors either with a 157-point transposition, or a 59-point interchange. The rule says you are supposed to choose the latter. If, on the other hand, the choice was between 59-point interchange and a 79-point transposition, you should choose the transposition.
fpawn, I think you are interpreting the word ‘sufficient’ in a very narrow sense. Sufficient doesn’t mean allowable; it means reasonable, sufficient to make it acceptable. From all the examples and rules quoted above, sufficient would mean a transposition under 80 points.
I agree that it should not be left to the reader to interpret the word ‘sufficient’ though.
The “rules” in the rule book aren’t meant to be read in isolation. That’s WHY all the cross-references and examples are given. Without all the examples and explanations the rule book could be a lot smaller but wouldn’t explain things nearly as well. If a case is presented in an example then it’s part of the “Rules of Chess”.
Example 2 in the rule 29E5e clearly covers the same case fpawn asked about. There’s no need to “interpret” the word “sufficient”. The interpretation is spelled out right there in the rule book.
If you think this needs additional clarification or explanation in the rule book, let the rules committee or Tim Just know. I think the example makes things pretty clear, but there’s always room for improvement.
The whole subject of transpositions and interchanges is very complicated. It’s one of the reasons I’m glad there’s pairing software.
OK, thanks! I guess that answers my question. Not that I fully agree with it, but that’s just my opinion. All I really wanted to know is that the pairing program did what it was supposed to do.
By the same token, you could argue that if the raw (“natural”) pairings are:
2000 BWB vs 1560 BWB
1600 WBW vs 1550 WBW
– and if you make the switch:
2000 BWB vs 1550 WBW
1600 WBW vs 1560 BWB
– then this transposition is harsh from the standpoint of the 1550 player. Yet, from the overall tournament perspective, it’s only a 10-point switch, and I’m sure that any TD in his right mind would make this switch.
Why is it that some players accept harsh transpositions, but not harsh interchanges? They seem to feel that players in the top half, even by just a few rating points, have a God-given right to an easy pairing in the next round.
A better philosophy (for players) would be, “Gee, I’m very close to the middle. Thus, my pairing could go either way. I could be playing the top player OR the bottom player.”
That question is easy to answer. Why? How do people explain the basics of the “swiss system” to a newbie? First you say that players of equal score will play each other. And then you say that you split each score group into half and match the top half with the bottom half. At least that’s how I was taught when I started playing tournaments many moons ago.
In other words, there is indeed a mentality that top half vs bottom half is a God-given right. You see players at tournaments all the time try to figure out whether they will get paired up or down. Translated: they are trying to see if they are in the top half (paired down) or the bottom half (paired up). They get excited if they are #5 of a 10 player point group.
I think more people would understand an interchange if there was no way to do a transposition to solve the pairings problems. But in my example the 157 point transposition is quite legal–just that for some odd reason (imho) the 59 point interchange is preferred. And that’s where the confusion starts.
Very true. And there is a “wink-wink” clause in the rules that even reinforces that mentality: “While interchanges are theoretically acceptable … interchanges do violate basic principle 27A3, Upper half vs lower half, and so tend to catch players by surprise, causing them to express their concerns. …” Translation: Even when an interchange is preferable to a transposition, you may want to use the transposition instead, so the players won’t squawk.
I hate wink-wink rules. If plan A is better than plan B, go ahead and use plan A, even if the players need to be educated a bit.
Another area where many players need a little education is in the matter of pairing players due the same color. Some people mouth the phrase “higher-rated gets due color”, not realizing that this is only rule 5 on a list of 5, and kicks in only if rules 1-4 do not decide the issue. I once had an Expert try to tell me that a higher-rated player with bye-WB should receive the white pieces over a lower-rated with BWB, ignoring the “equalization has priority over alternation” rule.