3 player section

Hi! I am relatively a new TD (Club). For an upcoming scholastic tournament (later today), I have 8 players in Unrated/U500 and 3 players in Champ (800-1200 range). This is a G30/d0 rated tournament.

What is the recommended (do’s and don’ts) for any section with 3 players in pairing/tournament format?

Options to consider in my view (I prefer option 1, but like to know the pros/cons):

  1. Just do a round-robin for the 3 players (one round, so 2 games each or double header if time permits, so, each player has 4 games)? In this case, I presume, this will still be treated as a tournament (and not as match. Please confirm).
  2. Ask a player in Unrated (who have 500-600 provisional rating) to play up and use Quad?
  3. Combine the sections (big rating range is a concern)?

PS: I searched the forum for similar topic (involving 3 player section), but could not find any previous discussion.

Thanks for sharing your insights.

An unrated player needs four games to get a published rating, so only ask a willing unrated to play up if it will be a double-quad or if the eight-player section would have been paired as a three-round octagon anyway.

Thanks Jeff!

My understanding is that a player needs to play about 25 (or 26 to be specific) rated games to get published USCF rating (i.e. move up from Unrated). Is that right?

About the section - good idea, will ask if any unrated or U500 player is willing to play up.

26 games is needed to get an established US Chess rating. Having 4 to 25 games makes the rating provisional (affects things like eligibility for matches, prize limitations at some large tournaments, one of the qualifiers for a TD to attempt to move from Club to Local, etc.). Having zero to 3 games means a person does not have a published rating. I think 8 games is when the rating algorithm switches to the standard one.

Unless the player has all wins or all losses

A player with 8 or fewer games as of the start of the event (section) or with all wins or all losses is rated using the special formula, otherwise the regular formula is used.

As to the original question, it would probably help to know how many rounds are planned. It would not be treated as a match, even if the 3 players played a triple or quad Round Robin.

What about players rated 500-799? Is there a middle section for them?

Did the TLA say that players in the Champ section must be rated at least 800? If not, you could call for volunteers in the lower section(s) to transfer to the top section.

One trick when you have a small odd number (like 3 or 5 or maybe 7) is to have everybody play 2 games at once. With 3 players, A would have white vs B, B white vs C, and C white vs A. Arrange three tables so that the short sides of each table touch at the corners, forming an equilateral triangle in the center. Since everybody is playing 2 games you’d want to double the announced time control. Instead of G/30 make it G/60.

In the case of 3 players you’d go for a double round robin. Each player plays 2 games at once, and you do the whole thing twice, with colors reversed. Each player would thus play 4 games altogether.

Bill Smythe

I would like to hear somebody from the rating committee weigh in on the advisability of this. It would tend to detract from the play of somebody who needs to stay focused on a game (due to bouncing back and forth between two games) and would thus muddy up the ratings as a predictive factor as to how the players would perform in the future. The volatility of players with three digit ratings may drown out the additional noise of such muddying but that may not be true in a general case.

I think it’s a terrible idea to have inexperienced kids attempt to play multiple games at the same time. It’s not all that good an idea for experienced adult players, either.

There is, as far as I know, nothing in the current rules or ratings procedures that would prohibit players from playing more than one rated game at the same time. While they might not like the idea of playing multiple games at the same time, the Ratings Committee has no authority to alter the rules or ratings procedures, their purview is limited to the mathematics of the ratings system. And they can recommend changes to the formula, such as the annual review of the bonus threshold authorized by the Delegates years ago, but any changes are then approved by the Executive Board.

I have played multiple rated games at the same time in several occasions, though not since I turned 30. But no one should be forced to do so. And it is definitely an awful idea for novices.

As a relatively new Club TD with little experience, do not go with tricks and gimmicks that are offered. Even long experienced TDs should avoid these. Keep it simple. If you have a three player section and the rest of the sections have even numbers, then just run that section as a quad. Each player will get two rated games and a bye. If there is an extra player or a “houseman” available, you can plug the player into the quad. Over the players’ next events they will get enough games in to have a rating published. New players need four rated games to get a published rating. They don’t have to get it all done in one event. To get an “established” rating the player must play 25+ rated games.

I assume you are running a 4 round event. You would prefer not to have quad section, but combining sections might cause mismatches and still leave you with the problem of having to give out a bye every round. Do the least amount of damage in managing your sections. I am not a big fan of Game 30, d0 for a new TD as it introduces more potential problems for you to have to handle when time pressure arises in multiple games. Read the Rulebook for such situations just to see what you have gotten yourself into. Likely you will not have enough digital clocks, or even a clock at all on some games. In any event, be patient and kind with the players. They often have less experience than you do and need your guidance.

The decision of what time control to use really needs to reflect the conditions and likely players in the event.

If the organizer is providing the clocks, that might lead to different decisions. If few players are likely to have clocks, or if most of the clocks are likely to be analog clocks as opposed to delay/increment capable clocks, that’s important, too. And the players’ familiarity with setting and using clocks could also factor into it, especially for scholastic events.

I think it’s a fallacy that d0 allows a tighter round schedule than d5 or +5, especially in scholastic events, where few players will use all 30 minutes anyway.

It’s important to know that different rules apply in certain situations when the time control has neither increment or delay. These rules differences IMHO tend to make things harder for the TD.

I should add that I would never have players play two games at once unless all agree. Usually, though, once they find out they’ll have twice as much time (e.g. G/60 instead of G/30), they’ll go for it.

Also, I wouldn’t do it for kids under, say, 14.

Bill Smythe

It is impossible to tell who the OP is since, for some reason, he(?) was able to register as an affiliate. Nevertheless, it appears from the affiliate’s tournament history that the OP decided to combine sections. Also, it was a four round event.

Alex Relyea

RAMU GANESAN was the Chief TD, and a Club director. It is possible that he and OP are one and the same.

I’ve created a new topic, Posting in the forums using an affiliate ID in the US Chess Issues forum for discussion of whether it’s possible and permissible to post in the forums under an affiliate ID.

Exactly what are you referring to with the above statement?
Many years ago now, a fella came to the Dallas Chess Club. He had just won I think 6.5 of 7 in a Florida U1000 tournament (his first),
and then 7 of 7 in an Oklahoma u1000 tournament. He was undefeated playing in the DCC U1400 section, until he reached 24 games. He had a rating close to 2200, YET, had never played anyone
over 1400. His quest for the NM for life, and a lifetime floor of 2000
ended, however, when he lost to a good friend of min rated about 1250. So sad. Never saw the guy again.

Rob

He was referencing Jeff’s comment, about how many games it takes for the rating algorithm to switch to the standard calculation. With all wins or all losses, the algorithm cannot accurately compute the rating. That is evidenced by your anecdote, since that player was obviously not at 2200 strength and severely misrepresented by the system.

Before 2000 we used to average the results for all provisional players, even those who won all their games. A player who beats 2 1000 players and 2 2000 players got a 1900 rating, which made no sense.

If a player has all wins, then that player’s true strength is unclear, about all we know for sure it that the player should PROBABLY be rated higher than any of his/her opponents so far. I think we cap such a player at 2700 (corrected from 2800), which would mean that the highest rated player he/she has defeated was 2300 or higher. (Personally, I think that above 2000 we should probably use 200 points rather than 400 points, but that’s not my decision to make.)

Likewise, if a player has all losses, then that player should PROBABLY be rated lower than any of his/her opponents so far. The hard (100) and soft floors (up to 150) will limit this, though. I suspect that if we eliminated all floors we would have a number of players with negative ratings, possibly some as low as -500.

The largest number of initial losses of somebody I know personally is a person who achieved a 110P46 rating (the 46 losses from Nov 2014 to Mar 2016 included 10 tournaments of three of more games each, which took the absolute floor from 100 to 110). A win in April 2016 made a new absolute floor of 115 and was the new and established rating. Since then improvement has left the floor behind.