PDA

View Full Version : Minimum time control for standard rating



Rincewind
21-04-2004, 12:01 PM
What should be the minimum time control for a game to qualify to be rated in the standard (ie not rapid) rating list.

Currently 60 minutes is the minimum. This is in line with when the FIDE rules on Rapid chess cease. IE You must begin scoring the game, etc. However, that does not mean the game necessarily needs to be rated in the standard rating list.

In another thread Rob from WA proposed that 90 minutes be the cut off, and I'm inclined to agree with him. 60 minutes feels too fast to me and more like a rapid game than normal chess.

Equivalents to 90 minutes using Fischer clock increments could be used too. Using the benchmark of initial + 60 * increment then a time control of 60 minutes + 30 seconds per incement would also be able to be rated. This time control is also good for directors as there is never a situation when a player can legally stop scoring the game and draw claims under 10.2 are avoided.

In the poll just choose the single period equivalent of what you believe. EG if you think 90 minutes OR 60 + 30s/move should be the minimum then select 90 minutes. Only choose 60 minutes if you think game in 60 minutes (the present minimum) is what it should be.

Ian Rout
21-04-2004, 12:26 PM
Not wanting to be a troublemaker, but I think it's an issue of the style as well as the length. A game with a guillotine finish, even with a relatively long starting period, still comes down to lightning, or potentially does. Even though 60+10sec is on the short side, it cuts out the nonsense of meaningless shuffling with vigorous clock thumping, or arbiters forming judgements on insufficient losing chances. I am relaxed about rating 60m+10 as "normal" given the practicality that at 90m its difficult to play seven rounds in a weekend.

arosar
21-04-2004, 12:35 PM
How's about 40mins+30secs [from 1]? I vote for that.

AR

Garvinator
21-04-2004, 12:42 PM
How's about 40mins+30secs [from 1]? I vote for that.

AR
this poll has been run before, do you remember that barry?

Rincewind
21-04-2004, 01:11 PM
this poll has been run before, do you remember that barry?

That was a while ago and there is a new bunch of people on here now. Be interesting to gauge current thinking.

Garvinator
21-04-2004, 01:33 PM
That was a while ago and there is a new bunch of people on here now. Be interesting to gauge current thinking.
thought you would have countered with, no that was for best time control for weekenders, not for standard rated games ;) :whistle:

Bill Gletsos
21-04-2004, 01:55 PM
How's about 40mins+30secs [from 1]? I vote for that.

AR
According to FIDE thats considered to be equivalent to G/70.

Kevin Bonham
21-04-2004, 03:07 PM
I play a lot of G60. I would say that the difference between G60 and G90 is much greater than the difference between G90 and G120 or maybe even slower, but not as great as the difference between G60 and G30 or even between G60 and say G45. In a G60 you can only deeply tactically examine one or two critical positions in the game, an awful lot gets left on the table but still the blunders are nothing like they are in a rapid. I'm still a bit uneasy about G30/+30 being mainlist-ratable but there hasn't been a huge outcry over this lately and I'm not sure how much that limit is used (G40/+30 seems a bit more favoured).

The time translations are always a bit rubbery. FIDE's 60 moves is an overstatement for most games but an understatement at others. I played a training game at G60/+30 under tournament conditions last week and it went for nearly three and a half hours, which is longer than the longest G90. I'd be interested to know if the tendency for both players to run the clock down to next to nothing is more or less pronounced in the add-ons than in flat games of comparable length.

Rincewind
21-04-2004, 03:10 PM
How's about 40mins+30secs [from 1]? I vote for that.

Sorry AR but I could allow for every combination of possibilities. You'll have to choose between 60 or 90 this time.

Rincewind
21-04-2004, 03:14 PM
thought you would have countered with, no that was for best time control for weekenders, not for standard rated games ;) :whistle:

That's true but I think one would influence the other. If we only rated G90 games (or equivalents) then we would see less G60 and more G90 weekenders I think.

Bill Gletsos
21-04-2004, 03:18 PM
That's true but I think one would influence the other. If we only rated G90 games (or equivalents) then we would see less G60 and more G90 weekenders I think.
I seriously doubt it. See my comments in the underrated junior thread.

Rincewind
21-04-2004, 03:45 PM
I seriously doubt it. See my comments in the underrated junior thread.

I think the main casualties would be the the larger 2 dayers who could not afford to drop to 6 or 5 rounds. Still the standard rating list's loss is the rapid rating list's gain.

The question is where is the right place for the division of the two lists. G60 feels too fast for me and more like a G30 game than a G90 game. If people are avoiding G60 events because of it then changing the rating division will help the problem in two ways.

(1) There may be more G90 events. Maybe not many more, but unlikely less.

(2) People won't avoid the events that stick with G60 because it will not impact their standard rating.

Bill Gletsos
21-04-2004, 03:53 PM
I think the main casualties would be the the larger 2 dayers who could not afford to drop to 6 or 5 rounds. Still the standard rating list's loss is the rapid rating list's gain.

The question is where is the right place for the division of the two lists. G60 feels too fast for me and more like a G30 game than a G90 game. If people are avoiding G60 events because of it then changing the rating division will help the problem in two ways.

(1) There may be more G90 events. Maybe not many more, but unlikely less.

(2) People won't avoid the events that stick with G60 because it will not impact their standard rating.
It seems we are debating this in two threads.
So be it.

Dropping rounds just to play G/90 is unlikely to increase player numbers.
Hence the number of normal rated games will drop.

The tournaments that stay at G/60 may not end up all getting rated.

Hence you end up with:
1) Less rated games in the normal system, hence less reliable ratings.
2) All though the rapid system may benefit, it is unlikely to be very significant.

Of course what I find interesting in all this is that although we may all like to pontificate on this, there has never been any sort of groundswell resistance to ther rating of G/60 in the normal system and its been around for at least 10 years.

Bill Gletsos
21-04-2004, 03:54 PM
Of course what I find interesting in all this is that although we may all like to pontificate on this, there has never been any sort of groundswell resistance to ther rating of G/60 in the normal system and its been around for at least 10 years.
Perhaps of course all this means is that we are Borg.

Rincewind
21-04-2004, 04:10 PM
It seems we are debating this in two threads.
So be it.

I'll try to reply in here to save that other one for DR and MS.


Dropping rounds just to play G/90 is unlikely to increase player numbers.
Hence the number of normal rated games will drop.

I don't know anyone who is avoid G60 games but it seems likely that there may be some. I can't think of too many people who would play in a G60+10 (FIDE) and not in a G60+30 (Fischer) so I think numbers would probably go up. Probably only marginally though.


The tournaments that stay at G/60 may not end up all getting rated.

I think they would as weekenders need ratings to work organisers are rating currently. The other advantage is that organisers would save on rating fees and thereby be able to give bigger prizes and in turn atrract more players.


Hence you end up with:
1) Less rated games in the normal system, hence less reliable ratings.
2) All though the rapid system may benefit, it is unlikely to be very significant.

As I said in the other thread, lowering the SD of time controls would have a positive effect on reliability. Also more players nd more game woould reduce the impact of less game overall.

I would hope the rapid rating list would be a big winner in this. There is some resistence at my club to rate the club rapid. I consistently argue in favour of keeping it rated as I believe the rapid rating list should be supported (even though most of us don't play in any other rapid events). But I feel most guys don't think the rapid rating is very credible or important, which is a shame.


Of course what I find interesting in all this is that although we may all like to pontificate on this, there has never been any sort of groundswell resistance to ther rating of G/60 in the normal system and its been around for at least 10 years.

I don't think I'm potificating just exploring possibilities. I think the rating landscape has changed substantially in the last 10 years though and so the issue is worth debating. There could be a silent majority of anti-G60 players out there quietly taking up bridge. ;)

Oepty
21-04-2004, 04:13 PM
I voted for 60mins, but by that I meant incremental time limits that go past 60 mins in total for 60 moves. I agree 60 moves is to long, but as it is what is generally used I don't think it is a huge problem to continue using it.
Scott

Rincewind
21-04-2004, 04:42 PM
Those that dont like to play long games.

Is there a signficant number who would play G60+10 (fide) and not G60+30 (fischer) do you think?


I'm not convinced of that. Many see the rapid list as mainly a junior list.

Today, that is true. But we are not bound to that paradigm.


You would need to prove that the rating of players at G/90 is significantly different to their rating at G/60.

No, it is a question of degrees. Do you have the number of rated games broken down by time control?


True, however unless their is a distinct difference in a players strength between G/60 and g/90 then then the reduction in games will reduce reliability.

Perhaps but the increase in reliability of the rapid list is another factor. As it is the rapid list that would be used in G60 events and the standard rating would only be used with G90+ events.

Garvinator
21-04-2004, 05:37 PM
Perhaps of course all this means is that we are Borg.
does that mean there will be more Jeri Ryan's at tourneys, if so, im all for it :lol: :owned: ;)

JGB
21-04-2004, 05:40 PM
I think 60 minute games suit players with a strong opening knowledge the best. Players saving 20 - 30 minutes or more over the first 10-15 moves. stronger players with a better opening knowledge can use sharp continuations that require time for the oponent to analyse and win valuable time. But i guess thats why these players are 'stronger' in the first place?! :doh:

arosar
21-04-2004, 05:44 PM
Perhaps of course all this means is that we are Borg.

Speak for yourself. Some of us are Species 8472.

AR

Rincewind
21-04-2004, 06:18 PM
Virtually all NSW events are G60+10s and not just G60.

Also I assume those advocating G90 would not be satisfied with G60+10s or possibly even G60+30s

I am satisfied with G60+30 provided it was fischer increment from move 1. The G60+10 where the 10 only get's added on after the first period has finished is not much better than G60. In fact, I have a pet theory that it causes more losses on time than a guillotine G60.

Rincewind
21-04-2004, 06:22 PM
As are many club tournaments, etc.

BTW do you have an feel for the proportions of rated games broken down by time control?
Nope.

That's a pity. Would you consider running two list in parallel for a period of time to see if predictiveness of the standard list in G90+ envents is improved?

Alan Shore
22-04-2004, 03:03 AM
Those that dont like to play long games.

Me, me, I said it first! :cool:


Is there a signficant number who would play G60+10 (fide) and not G60+30 (fischer) do you think?

I would try to not touch anything longer than G60, I grudgingly accept the fact of life of 60+10 but would most likely avoid a 60+30, 40+30 and 30+30 as I have already demonstrated many times in other threads why these time controls do not work for weekenders.

(The above only apply to weekenders, I am perfectly happy for G90 type games in club comps, Championship tournaments, etc.).

Barry said on the other thread in answer to me something along the lines of tournaments on 'long weekends'. There aren't that many long weekends around but I'd actually prefer to play a greater number of rounds in allotted time rather than longer games. I have seen ones with the format of some games 60 some 90 (Redcliffe comes to mind).

As for the G90 in normal weekenders, this would see only 5 rounds play - an insufficient number to avoid a heavy cluttering of the leaderboard.

Rincewind
22-04-2004, 10:11 AM
I would try to not touch anything longer than G60, I grudgingly accept the fact of life of 60+10 but would most likely avoid a 60+30, 40+30 and 30+30 as I have already demonstrated many times in other threads why these time controls do not work for weekenders.

(The above only apply to weekenders, I am perfectly happy for G90 type games in club comps, Championship tournaments, etc.).

So you are happy to play G90 just not on weekends. On long weekends most tournaments are G90 anyway so there is no change there.


Barry said on the other thread in answer to me something along the lines of tournaments on 'long weekends'. There aren't that many long weekends around but I'd actually prefer to play a greater number of rounds in allotted time rather than longer games. I have seen ones with the format of some games 60 some 90 (Redcliffe comes to mind).

There is Australia Day, Easter, Anzac Day, Queen's Birthday and Labour Day. That's 5 weekenders a year in each state (say) 30 weekenders a year for the rating list.


As for the G90 in normal weekenders, this would see only 5 rounds play - an insufficient number to avoid a heavy cluttering of the leaderboard.

If the weekender is a large one where 5 is insufficient they can play 6 rounds. Alternatively they can play using G60 controls and rate their tournament on the rapid list.

Remember this thread is not about what is the best TC for a weekender. It is abount what should be the minimum time control for a game to be rated. Fast players should care that much since they can always play fast. It the the slow guys who object to having their time pressured blunders at G60 skew their "real" rating. ;)

Garvinator
22-04-2004, 10:12 AM
As for the G90 in normal weekenders, this would see only 5 rounds play - an insufficient number to avoid a heavy cluttering of the leaderboard.

sorry cant help myself here even though i know what response i receive from bruce ;)

5 rounds could work if you divided the field into different groups(say on rating allocation) or you used two round acceleration :whistle:

Rincewind
22-04-2004, 10:22 AM
I voted for 60mins, but by that I meant incremental time limits that go past 60 mins in total for 60 moves. I agree 60 moves is to long, but as it is what is generally used I don't think it is a huge problem to continue using it.

60 moves is long but is better than using 40. This is why.

Imagine you are playing a G90 game and you want to plan your time usage. The average number of moves for a game is 40 so you do plan to use 2 minutes 15 seconds per move? That would seem to make the most of all the time allotted. Of course you don't as when you get to move 40 you would have no time left and if the game is still going you would no doubt lose on time.

At G90 I would probably budget somewhere around 1 minute 30 seconds per move After 40 moves I should have around 30 minutes left which will hopefully be ample to finish the game. Note that this is around 90 minutes/60.

Of course the time budgets are not fixed at the start of the game and can change depending on how the game is going. However, you get the gist of the argument.

To perform a fair comparison you can't use the average number of moves for a game you probably need to use around 2 standard deviations more than that. I've no idea what the standard deviation for number of moves of a chess game is but to my mind 60 is reasonable and a round number which is easy to work with (since there are 60 seconds in a minute).

Rincewind
22-04-2004, 10:29 AM
sorry cant help myself here even though i know what response i receive from bruce ;)

5 rounds could work if you divided the field into different groups(say on rating allocation) or you used two round acceleration :whistle:

Or they could stick with G60 and use the rapid rating list.

Garvinator
22-04-2004, 10:44 AM
is anything really going to change cause of all this talk?

Rincewind
22-04-2004, 10:47 AM
is anything really going to change cause of all this talk?

What sort of attitude is that? You should be banned from the BB for spreading dissent. ;)

Garvinator
22-04-2004, 10:51 AM
What sort of attitude is that? You should be banned from the BB for spreading dissent. ;)
:hmm: im trying to work out how one person and one person only having an opinion is spreading dissent, to spread dissent, doesnt it have to move from one thing to another :whistle: that has not happened at this stage, so i have not spread anything :p

Rincewind
22-04-2004, 10:53 AM
:hmm: im trying to work out how one person and one person only having an opinion is spreading dissent, to spread dissent, doesnt it have to move from one thing to another :whistle: that has not happened at this stage, so i have not spread anything :p

Having an opinion is one thing. Posting a message suggesting that all this banter is ultimately futile is something else. :p :p

Garvinator
22-04-2004, 10:56 AM
Having an opinion is one thing. Posting a message suggesting that all this banter is ultimately futile is something else. :p :p
i notice though that you didnt disagree ;)

ursogr8
22-04-2004, 11:07 AM
What sort of attitude is that? You should be banned from the BB for spreading dissent. ;)

I agree with your ruling Barry. How long are you going to impose?

And just a suggestion, did you mean 'fomenting' dissent or 'spreading'.

Garvinator
22-04-2004, 11:20 AM
I agree with your ruling Barry. How long are you going to impose?

And just a suggestion, did you mean 'fomenting' dissent or 'spreading'.

what is fomenting? - genuine question

ursogr8
22-04-2004, 01:02 PM
what is fomenting? - genuine question

'a'a'

foment = to promote the growth or development of; instigate or foster (discord, rebellion, etc).

That 268 gap is going to disappear real quick when you are banned.

starter

Rincewind
22-04-2004, 10:01 PM
That's a pity. Would you consider running two list in parallel for a period of time to see if predictiveness of the standard list in G90+ envents is improved?

Bill, did you answer this question and I missed it or are you still considering it?

I can understand you might be "not entirely enthusiastic" about the extra effort involved. Perhaps you could just track for 12 months the time controls that rated games are played under so we would have an idea as to the proportions involved. That wouldn't be too onerous, would it?

You could just request and record this from the state ratings officers, they'd be the ones sticking pins into small Barry Cox effigies. ;)

rob
23-04-2004, 06:20 AM
If the weekender is a large one where 5 is insufficient they can play 6 rounds. Alternatively they can play using G60 controls and rate their tournament on the rapid list.

Remember this thread is not about what is the best TC for a weekender. It is abount what should be the minimum time control for a game to be rated. Fast players should care that much since they can always play fast. It the the slow guys who object to having their time pressured blunders at G60 skew their "real" rating. ;)

I agree with your comments in the 1st paragraph. Personally if I were running a weekender with an anticipated large turn-out (too many for a 5 round swiss-extremely unlikely in WA) I'd split the tourney into two groups to mantain the 5 round slower TC.

Your 2nd paragraph seems to give support to my comments in the other thread :) and also MS is probably of similar opinion although he appears to prefer weighting tourneys according to the TC for a less significant skew (as I do for junior ratings in WA).

Bill's latest comments suggesting WA has a more varied TC for weekenders indicates to me (I know he said it in context to giving ppl a variety of TC's - something ppl already get from other tourneys) that a significant proportion of G60 events can cause a skew (don't expect an admission) :)

As Bill is almost invaribly correct about rating issues, I am very hesitant about the comments I have been making.

I shall return but not before Tuesday - no comments required on this statement thanks :)

PHAT
23-04-2004, 07:06 AM
I post this before reading the thread. I read the poll and straiht away thought is there any reason that a game could not be used in [b]both[.b] lists.
eg
30 to 90 into the rapid list
60 to inf into the normal list.

games between 60 and 90 would be included in both lists.

Rincewind
23-04-2004, 10:15 AM
I post this before reading the thread. I read the poll and straiht away thought is there any reason that a game could not be used in both lists.
eg
30 to 90 into the rapid list
60 to inf into the normal list.

games between 60 and 90 would be included in both lists.

That's an interesting idea which I might need to think about some more.

My first impression is that you are losing some reliability in (at least) one list but picking up a stabilising effect making it less likely that the rapid and standard ratings would diverge too greatly. Not sure of the benefits of the second outweigh the costs of the first. I'm interested in listening to your arguments though.

Bill Gletsos
23-04-2004, 10:41 AM
I post this before reading the thread. I read the poll and straiht away thought is there any reason that a game could not be used in [b]both[.b] lists.
eg
30 to 90 into the rapid list
60 to inf into the normal list.

games between 60 and 90 would be included in both lists.
Dear Mr Latham,
Perhaps you only suddenly thought of this after reading my post in the other thread. ;)

Regards,
Bill Clinton


The USCF rates anything between G/10 and G/29 in their rapid system.
They did rate anything greater than G/30 in their normal system.
They have recently however started to rate games of G/30 to G/60 in both rating systems.

Bill Gletsos
23-04-2004, 10:57 AM
That's an interesting idea which I might need to think about some more.

My first impression is that you are losing some reliability in (at least) one list but picking up a stabilising effect making it less likely that the rapid and standard ratings would diverge too greatly. Not sure of the benefits of the second outweigh the costs of the first. I'm interested in listening to your arguments though.
Actually I have been looking at rating G/60 games and G60+10s(after the 60mins has expired) games in both the normal and rapid list.
The problem I have encountered in my attempts to retroactively run some tests is determining which previous tournaments met this criteria.

This therefore ties in with you question above.

Bill, did you answer this question and I missed it or are you still considering it?Nope, I just overlooked it.

I can understand you might be "not entirely enthusiastic" about the extra effort involved. Perhaps you could just track for 12 months the time controls that rated games are played under so we would have an idea as to the proportions involved. That wouldn't be too onerous, would it?

You could just request and record this from the state ratings officers, they'd be the ones sticking pins into small Barry Cox effigies. ;)
Yes I'd prefer just to track the information and play with the data.
I think you are correct that it wont make the rating officers overly happy. They have enough trouble sometimes getting information out of those who submit tournaments for rating.

arosar
23-04-2004, 11:04 AM
Dear Mr Latham,
Perhaps you only suddenly thought of this after reading my post in the other thread.

Regards,
Bill Clinton

Haha...I'll pay that one. Very good Bill. I tell you what, this Latham, or at least his minions, are friggin shootin' themselves in the foot. How stupid are these khrunts?

AR

PHAT
23-04-2004, 11:06 AM
I'm interested in listening to your arguments though.

ummmm, errrrr, I actually had no arguments and therefore have not lashed my soul to that flag. Nevertheless, you are forcing me to think .................................................. .................................................. ...................................... .................................................. .................................................. ....................................... .................................................. .................................................. ....................................... .................................................. .................................................. ....................................... .................................................. .................................................. ....................................... .................................................. .................................................. ....................................... .................................................. .................................................. ....................................... .

The figment that drifted between my ears was that it is not a prifitable use of data to limit the catagory size too much. It seems to be a concesus that more data is better - so let us cast the TC net a little wider for both lists. By necesity the two nets will over lap.

It has also occured to me as I tap away here that by allowing for catagory over lap, we could indeed have more than jus two catagories - rapid and normal.
Eg.
blitz = <5
rapid = <30
social = <60
club = 30 to 90
open = 60 to 120
normal/championship = >90

All this would require that most rated tourns would have to be entred into two catagories.

BUT, wow, think of the info that would generate. Players could see what TC eachother are better/worse at and adjust their play accordingly

PHAT
23-04-2004, 11:08 AM
Dear Mr Latham,
Perhaps you only suddenly thought of this after reading my post in the other thread. ;)

Regards,
Bill Clinton

No, great minds think alike. :owned:

... although might concider that fools seldom differ. :hand:

PHAT
23-04-2004, 11:13 AM
Haha...I'll pay that one. Very good Bill. I tell you what, this Latham, or at least his minions, are friggin shootin' themselves in the foot.

AR

I hope their rifle is of decent power - the bullet might also go through head they are stepping on - Howard's.

Rincewind
23-04-2004, 01:00 PM
BUT, wow, think of the info that would generate. Players could see what TC eachother are better/worse at and adjust their play accordingly

Mmmm, intersting. I guess it would depend on the quantity of data involved. Obviously if all active players played 20 games in every time control there would be little need to overlap categories. But with less data coming through I guess the loss of accuracy in the TC department might be compensated by the greater accuracy of rating performance.

However the number of lists you mentioned in the last post would seem to be going too much to the extreme, thereby diluting your data again and losing the benefits gained by overlapping categories.

I wonder would organisers of tournaments in the 60-90 range have to pay the rating fee for standard and rapid? I can see that being popular. ;)

Bill Gletsos
23-04-2004, 02:25 PM
I hope their rifle is of decent power - the bullet might also go through head they are stepping on - Howard's.
Actually it could be the case of the bullet going thru Latham's foot into his skull because his foot was in his mouth. :whistle:

Bill Gletsos
23-04-2004, 02:26 PM
No, great minds think alike. :owned:

... although might concider that fools seldom differ. :hand:
Both of those options are a definite matter of concern. ;)

Alan Shore
23-04-2004, 11:24 PM
The USCF rates anything between G/10 and G/29 in their rapid system.
They did rate anything greater than G/30 in their normal system.
They have recently however started to rate games of G/30 to G/60 in both rating systems.

I never thought I'd see the day when the Americans would actually have a system I approve of but I like it. :clap:

Ian Rout
24-04-2004, 10:22 AM
On the old board I started a thread about what we can learn from the US. My thesis was that America's chess underachievement was due to their basketballisation (or "dumbing down" as some media organs would term it) of chess, and that it would be wise to study the Americans with a view to avoiding following their lead. This looks like a prime example.

Of course we don't have the ability to paper over the cracks in the way that America has, namely by importing crateloads of ex-Soviet and East European GMs, so it is more critical here.