“Colors for the first round only may be chosen by lot instead of the table (in order to determine by lot which players receive two whites).”
However, in the pairing table in the PDF, choosing colors by lot in the first round could potentially enable player 3 to get white three times and allow player 2 to get black three times. Should the language in the chart instead read “Colors for the third round only may be chosen by lot…”?
Also, Rule 30G for Quads includes the following language for pairings:
“Round one: 1–4, 2–3; round two: 3–1, 4–2; round three: 1–2, 3–4 (colors by toss in this round)”
Rule 30G seems correct, and I would have expected the Quadrangular Tournament Chart PDF to match 30G.
Assuming that 30G is correct and the Quadrangular Tournament Chart is not, whom should I contact about updating that PDF?
I believe you are right in that it should say the third round instead of the first round (in addition to it stating this in rule 30G, it also states this
under the table for a 3 or 4 player round robin in Chapter 12 where it says “Colors in the third round are determined by toss”). Also, doing the third round colors randomly shouldn’t be listed as a variation on the quadrangular tournament chart since doing the colors randomly in round 3 is the “preferred” (and only) method given in rule 30G and the only method given for a 3 or 4 player round robin in Chapter 12.
On a side note it’s interesting that rule 30G states “Players’ numbers are assigned in order of rating” but Chapter 12 states “Pairing numbers are assigned by lot”. I’ve brought this issue up before to no avail. It would be good if this issue could get cleared up and have it stated on the quadrangular tournament chart how pairing numbers are assigned for quads.
The note on the Quad chart is for a “variant” of the standard procedure for running quads.
The way most of us have run quads for as long as I and the other TDs from the Jurassic age have done so is to follow one of three methods:
Array players in rating order and follow the quad chart and color allocation.
Array the players by rating and allow them to toss for color in the third round.
Place the players by lot on the chart and follow the color allocation.
Reasons for placing players by lot include having randomness in the order. Often the same people will show up to every quad tournament. The same four players will play each other again and again. It is easier to put them by lot rather than toss for color in any round. Over time, it appears that tossing for color in the last round of any event has become frowned upon, though that is the way we used to handle pairing/color problems in Swisses and Quads for our Jurrassic brethren who complained less about color allocation when they had some semblance of control over whether they had White or Black. No sensible beast used a color toss in the first round. The color allocation problems were few, but arguments with the TD were many even for perfectly logical and by the book pairings. Methods 2) and 3) for quads were the most usual. Method 1) was the tidiest for the USCF for reporting purposes. Rule 30G will accommodate all of them. Ignore the “variant” on the Quad report sheet, like all of the creatures of the past have done for countless millennia. If it continues to bother you, white it out when you make copies and pretend a meteorite hit the form. No one will notice as the new little mammals you are directing just want to play chess.
I have a fourth idea which I think is superior to all of these. It requires only one coin toss.
Put the four players in rating order, but use the coin toss to determine whether to number the players top-to-bottom (1-2-3-4) or bottom-to-top (4-3-2-1).
Then use the “usual” color scheme for round robins: Even-vs-even or odd-vs-odd, larger pairing number gets white. Even-vs-odd, smaller pairing number gets white. No need to toss coins again in round 3. With the single initial coin toss, it becomes 50-50 whether the highest-rated player (or any other player) will get 2 whites and 1 black, or vice versa.
This single coin toss provides all the randomness necessary.
If there are multiple quad sections, you can go a step further to make things “look” even more random: Alternate down the quads. If the top quad is 1-2-3-4, the second should be 4-3-2-1, the third 1-2-3-4, the fourth 4-3-2-1, etc.
So what? In the first round of a Swiss, the top player is always paired against the player just below the middle. In the first round of a round robin, the top player is always paired against the bottom player. There’s nothing wrong with a little regularity.
As I said, my method provides all the randomness necessary. It does not provide additional, unnecessary randomness.
Most of the TDs I know assigned the quads by lot. We noticed that the same 12 people would show up. This led to complaints that they had to play the same person and had the same color allocation every tournament. So we mixed it up. Still the same four people per quad, but different placement on the chart to change the color allocation. As I recall, one time a coin toss was tried, but the coin ended up rolling under the radiator into a dense spider web. No one wanted to reach in to get it.
If you flip a coin, it has a small but non-zero probability of landing on its edge. I’ve had it happen once or twice. (There’s a Twilight Zone episode that utilizes this.)
When I was in college, several of us tried to figure out what the ratio of diameter to thickness was to create a ‘coin’ with equal probability of head, tails and edge. It was just one of those things that engineering students do when they’re bored, like figuring how hard the wind would have to blow to blow over the John Hancock Building in Chicago, or the all-time classic written by an engineering student at Purdue, a stress analysis of a strapless evening gown.
If you call the coin flip, call sides. If it lands on the edge that is a side [edge side] of the coin. If it lands on heads or tails, that also is a side of the coin.
I think I saw that classic in a paperback book in the 1960’s. In the same book were a couple of interesting proofs:
All horses are the same color. Proof by induction on the number of horses: It’s obvious that one horse is the same color. Now suppose you know that N horses are the same color, and you have a set of N+1 horses. Remove one horse, then the others are the same color by the induction hypotheses. Now put back that horse and remove another. Those N horses are also the same color, for the same reason. Thus all N+1 horses must be the same color.
All horses have an infinite number of legs. Proof: In back, the horse has two legs. In front, the horse has forelegs. That adds up to six legs, which is certainly an odd number of legs for a horse. But the only number that is both odd and even is infinity.
But, suppose that, somehow, there is a horse with only a finite number of legs. That would be a horse of another color, which by the previous theorem does not exist.
The people who could correct it (assuming correction is deemed necessary) probably don’t follow the Forums much, and the Forums are NOT intended as a formal means of communicating with US Chess staff.
Digital Editor John Hartmann or Joan Dubois would probably be the ones to contact.
Thanks Mike for answering the original question as to who should be contacted to fix this. I will email them on it now (of course a correction is necessary).
This is a pretty poor “variant”. The original post in this thread points out that the “variant” listed on the Quadrangular Tournament Chart “could potentially enable player 3 to get white three times and allow player 2 to get black three times.”
