This past Monday night we had the first round of a four round tournament at our Chess Club. I am the TD of record and my friend Wayne is the President and has the club’s laptop with SwissSys on it.
In this tournament we have one person that is completely unrated with this being his first rated event. He is the father of 3 boys that have been active players in our club and our local scholastic scene.
His game had the only upset with him beating a 1503 rated player.
For the second round he will be number 8 of the 1-0 group of 8.
We believe that one of the other winners will not be coming for the second round as he will be out of town. This makes the unrated fellow number 7 of that group of 7.
His due color is Black, the same as the due color of the number 6 player in that group.
With the pairing program we ran SwissSys to see what the pairings would be. It had the unrated player paired against number 3 in the group and put number 6 down to play the highest rated player in the 0.5 group.
Both Wayne and I wondered why it did this. As an experiment, we gave this unrated player a low rating, 599, to see how it would pair him if he had a rating instead of being unrated. Well, it had this player, number 7, go down to play the highest rated 0.5 player and had number 3 paired against number 6.
So, SwissSys pairs the player differently when he is unrated than when he is rated.
Is this pairing treatment of the unrated a USCF rule or just something the program is doing?
If we paired this tournament with cards instead of the program, should we give the unrated this “special” treatment?
As stated in the first sentence of rule 29D1, “In the case of an odd number of players, the lowest-rated player, but not an unrated player, is ordinarily treated as the odd player and is paired with the highest-rated player he or she can play in the next lower group.”
IMHO it is highly undesirable to drop an unrated player down to the next lower score group. If somebody must be dropped, it should be one of the lowest rated players.
I’m too lazy to look it up right now, but I think there’s even an explicit rule to that effect. If so, SwisSys was simply following the (highly intelligent) rule.
There’s no particular reason to assume the unrated is always the weakest player in his score group. In fact, unless the unrated has all wins or all losses, that assumption is highly illogical. He made it into that score group somehow, didn’t he? In the absence of outside information (i.e. a rating), it makes the most sense to assume he belongs about in the middle of his group.
In fact, you should feel free to consider the unrated as belonging anywhere in his score group, even in the top half, if it would improve the overall pairings in some way (e.g. colors).
For similar reasons, I dislike the idea of always listing unrateds at the bottom of the wall chart. I prefer to use an age-based guesstimate to determine wall chart position. Something like 1399 for adults (ages 14 and up), 1299 for age 13, 1199 for age 12, etc.
In later rounds, you can adjust the guesstimates based on what you have seen in the earlier rounds. Pairing cards or program, be flexible and use common sense.
I’ve rarely given unrated players ratings for the sole purpose of wall chart placement. If I’ve given ratings, it’s either based on perceived true strength, or in the case of using WinTD, I give ratings to all unrateds, but they are always below the lowest rated. It’s been a while since I’ve done this with WinTD and I don’t recall exactly why I did it, but it had something to do with bad pairings, I believe.
Anyhow, re: giving a player a rating for wallchart purposes; will his/her rating also determine his/her pairing number? If so, an unrated player could be playing a very high rated, if not the highest rated, player in the first round; that doesn’t sound right, but I don’t see any prohibition.
For unrated players, I don’t write any ratings on the wall chart. I just leave the rating blank or write “UNR”. But I might locate them at the 1399 position on the wall chart.
That’s fine, if you have any knowledge of the player on which to perceive his true strength.
So you’re still placing unrateds at the bottom of the chart, then. Unless they’re really the weakest players, this defeats the whole point of guesstimating ratings to begin with.
Don’t pair by pairing number. Pair by score and rating (including guesstimated rating).
Is that so bad? To me it feels about the same as pairing a 1400 against an Expert, which commonly happens in round 1 anyway.
Not necessarily. The rating guesstimate might change from round to round, depending on what you see in the earlier rounds. It wouldn’t be necessary to change the pairing number. (When pairing by hand, there would be no point. When pairing by computer, the program probably internally changes the pairing number on its own.)
I don’t think the program changes the rating or makes any estimate of a rating or pay any attention to the rating while it is determining the pairings.
When considering pairing significances, there are more than one type of unrated player:
The ultimate beginner that has not played Chess anywhere or at least to any kind of amount to increase their ability.
The experienced beginner that has played at a club or with friends before trying rated chess.
The intermediate to expert unrated that has played quite a bit, either in another country or venue or on the internet gaining him quality experience that would place him higher than a beginner.
Each needs to be treated a little differently for different tournament situations.
In the case of the complete beginner, putting him in the higher score group might not be best or the most accurate per his ability. Of course it certainly won’t hurt this beginner playing a higher rated opponent, but it won’t be the most accurate as far as pairing abilities.
The experienced beginner could actually learn more playing in the higher group, especially with a good post mortem acting like a lesson.
OF course the intermediate to expert unrated needs to be in the higher group because of ability and true performance.
Given a hard-number rating to an unrated has two effects:
Since they are now have a rating, they can be the lowest rated player in a score group, and so will be a strong candidate to be dropped. In particular, they can be given the bye, which is even a more serious problem, since you want the unrateds to get the full set of games if at all possible.
In evaluating swaps to fix colors, you now have a hard number to evaluate the difference between players.
I never said an unrated should be placed into a higher score group. On the contrary, one of my points was that an unrated is a poor candidate for being dropped (or raised) into a different score group.
Unrateds should be placed in their own score group (beginning with round 2, which is when score groups come into existence).
What I said was that unrateds could be placed anywhere within their score group, even in the top half if convenient. There is no reason to assume that an unrated belongs near the bottom of his score group.
Of course, in the zero-point group, if the unrated has been dropping his queen within the first 10 moves in every game, or falling for the 4-move checkmate, etc, then it might be reasonable to assume he should be at the bottom of the pairings.
The following, of course, is where we get the inclination to place unrateds at the bottom.
But then, we have this:
BUT, there is this…:
… and WHY is it pointed out that pairing numbers must be taken note of?
I think paragraph 5 (the TD TIP), could use a bit of work, for example, there’s the “[o]ne major difference” sentence, but it doesn’t clarify what that means the TD is to do. Is it saying that it is okay to consider the unrated player, who now has a rating - albeit assigned - as rated, and can be raised/dropped into a higher/lower group, or is it cautioning us against that?
You’ve found several examples of rules that need some serious work.
The admonition to use the pairing numbers instead of ratings is just a way of asking TDs, when assigning estimated ratings to unrateds, to be sure to actually use those ratings (rather than 0) to make pairings.
A better wording might be:
“When assigning a rating to any player (including unrateds) for pairing purposes, directors should make sure to use the assigned ratings, and not the actual ratings, when prioritizing the basic Swiss System rules.”
It seems to me that our main and key concentration should always as an utmost priority be the growth of our federation. Thus, we should view everything we do, in this light, and the
products such as our rules, rating system, etc, as tools towards
this goal. In that regard, unless someone does have a legitimate
foreign rating such as Canadian, FIDE, etc, or, in the case of the
poor performer or “sandbagger” for section purposes assigning
a rating more commensurate with their USCF rating history,
then we should endeavor in sections in which the UNR player
is participating, that they do not receive the byes. That we have
a priority in maximizing their initial tournament experiences.
Now this does bring into question as well the continued folly of
using published ratings for local events, esp, for scholastic, or youth oriented. Not a big deal really, just requires me to
publish "current ratings used’ for my events. We have the technology, lets use it.
Back in the days when I first started directing [pre-color TV?] we frequently put the unrateds about 1/4 to 1/3 of the way up the wallchart from the bottom. One reason this was done was to give the unrated player a chance for a higher rating out of his first tournament if they did have some skill. Now with computer pairings, you can still do this but you have to take Tom Doan’s comment into account. Don’t use “hard” ratings. This makes a big difference to the pairing program. Hard rated players ARE NOT UNRATED as far as the program is concerned. Check your program info to see how to give a soft rating. In SwissSys give the player a negative rating. He will be treated as unrated for the rules about an unrated but will be paired per the rating. -1530 is 1530 for pairing purposes other than as regards the unrated rules.
Please keep in mind that there is a difference between giving an unrated a soft rating and assigning a rating to a player who is either unrated or not.
