Minimum rating floor in the online rating systems

Running Chess Tournaments
21

It is doubtful anyone who might qualify for a floating minimum rating is aware of them.

Do they help promote additional play? I don’t know how to measure that.

There seem to be two types of new players, those who bottom out and might qualify for a minimum floating floor and those whose initial rating is too high and comes down a bit (100-200 points) before stabilizing and then, with the improvement we see in most new players who stick with it, goes up further. Many of the former don’t play in enough events to reach that ‘improvement’ stage.

22

If they aren’t being applied to the OL ratings, then we have a “natural experiment” whether anyone actually cares about the ratings (not about process/principle, but the actual fiddling with very low ratings). Note that if we did not have an absolute floor at 100, the players would end up sorting into different ratings, where a 0-7 in a K-3 event might be rated -300, while a 0.5-6.5 might be -170 and a 1-6 +50. Giving a 101 to the 0-7, a 103 to the 0.5-6.5 and 105 to the 1-6 is purely artificial.

23

A while back I tried to see what would happen on the test server if I eliminated minimum ratings, but there were so many places where a minimum rating was checked that I never got a clean run.

24

Under the current rules for handling provisional ratings, it looks like a race to the bottom when you change the minimum rating. The following are the percentiles for established ratings for different settings of the absolute floor.

Minimum 5% 10% 25% 50% -1500 -132.75 84.55 424.60 868.49 -1000 -119.39 92.05 427.22 869.18 -500 -23.56 150.29 453.52 880.01 -200 101.25 241.43 506.80 902.72 100 271.94 385.54 606.98 957.30

I was surprised, but there are a couple of people who manage -1500 established ratings if -1500 is the floor. That’s basically meaning that you almost certainly lose to people who can’t beat anyone but you. You have to admire their gumption if not their skill. One thing to note is that the existence of the really, really low ratings has almost no effect on the bulk of the established ratings. You really don’t see much effect until floor+400 starts to get up into the 10%-ile or higher probably because the absolute floor applies in all passes.

25

Or perhaps feel sorry for them, if for some reason there is an expectation they will continue to play chess even after showing no aptitude for the game.

26

My understanding is that the Ratings Committee is recommending not extending the floating minimum floors to online ratings systems. The documentation would be modified.

27

Interesting. What is their rationale?

28

I don’t participate in RC discussions, so I don’t know their rationale.

29

If one reads the thread, that should be fairly obvious.

30

The cases I know of display mild strategic ability while being susceptible to severe tactical mistakes.

31

I have noticed this as I am organizing quite a few online events, like NINE last week alone. It is strange to see a person who is rated
below 150 online to have his/her rating DECREASE. Not something you would see with the same rating reg, OTB

32

I’ve seen it quite legitimately happen to an adult’s OTB regular and quick ratings (two different events), but it is unusual.

33

My understanding is that floating minimum ratings floors (under 150) for online ratings will not be implemented, per recommendation of the Ratings Committee.

34

So are there different “fields” or programs used depending on age for OTB U150??

35

No. The only thing which is age-related is the initial guess rating for a complete novice. However, as a result of that, it’s very rare for an adult to get a rating<150. (20 year old newbie who has 1500L, 1200L, 1000L will get a 600/3 rating; 6 year old with the same record will be 300/3; and 5-8 year olds tend to play schedules with much lower opponent ratings than that examples).

36

I like to use the example of a complete beginner playing in the open section of the World Open.

Alex Relyea

37

Which, with the inevitable 9 losses (or maybe 8 losses and a bye) gives 1000/9 for the 20 year old and 300/9 for the 6 year old.

38

Typical of the ratings committee, with more emphasis on purity of ratings whatever the cost, including perhaps hundreds of beginners not continuing to play US chess events, and the discouragement of their parents. Look the auto increases for U150 for OTB, as well as floors for OTB, make as much sense for online chess. One of these days, one of these days the ratings committee will get the memo that a KEY desire in the ratings process is the maintaining and increasing of the US chess membership base.

39

The problem with open sections are insane parents who enter their young new chess playing children in which they
have zero chance at providing any competitive games for their opponents, Or, in truly receiving much of a benefit themselves. I remember trying to encourage a dad to register his novice son in a U section, rather than the Open,
and his dad furiously pointed out that his son has played 13 games in two major tournaments open sections in the last year. (lost all 13 but played them).

40

Some organizers have a Premier section instead of an Open section. Since it isn’t open it can have a minimum rating to register.