Pace of Game

[savin]

We are looking for some feed back on Pace of Game. We are aware that not many people actively post on the board but for those that do we would like to have your feedback.

Pace of Game is a source on much confusion and aggrivation. We would like to address some of the issues but beffore making any changes we need some feedback. lets start by explaining how Pace of game is meant to work:

We currently have 4 pace's of game defined, as follows:

Many Moves Per Day (Rocket)
2-3 Moves per day (Airplane)
1 Move per day (Car)
1-2 moves per Week (Bicycle)

However, the Rocket and Airplane sppeeds are only possible when the two players live in similar time zones. Basically to make more than 1 move per day the two players need to be awake and making moves at the site at the same time.

Given this, it almost doesn't make sense to have these two Paces of Play since in many cases its just impossible.

The next confusion is over what Pace of Play even means. For instance, 1 move per day does not mean each player should make 7 moves per week. It means that a player will typically reply to a move within 24 hours. That actually means 3-4 moves per week since there are two players.

Going forward we would like to make Pace of Play more understandable and more useful. Our current thoughts are to have 5 paces of play as follows:

1) Very Fast: On average the player will respond to their opponent's move within 24 hours, but may respond faster if both players are in the same time zone. This type of game is only valid with 3-day and 7-day forfeit settings.

2) Fast: On average the player will respond to their opponent's move within 36 hours. This is equivalent to normally replying within one day but sometimes taking 2-3 days to reply. This type of game is only valid with a 7-day forfeit setting.

3) Normal: On average the player will respond to their opponents move within 60 hours. This is equivalent to saying the player typically replies within 2 days but occassionally may take 3-5 days to reply. This type of game is only valid with 7-day and 14-day forfeit settings.

4) Slow: On average the player will respond to their opponent's move within 96 hours. This is equivalent to saying the player typically moves with in 3 days but may occasionally take up to 7 days to reply. This type of game is only valid with a 14-day forfeit setting.

5) Very Slow: On average the player replies to their opponent's move within 1 week. This is equivalent to saying the player typically replies within 5 days but may occassionally take up to 14 days to reply. This type of game is only valid with 14-day or 30-day forfeit settings.

Why do we think we need to adjust the definition of Pace? A good example is the current enhanced site tournaments that are set to a Pace of 1 day per move. Many players have told us they can't enter because its not possible to maintain a pace of 1 move per day when the opponent is in a different time zone. But by 1 move per day we meant within 24 hours of the opponent's move. And we really would have preferred a setting of Normal as defined above but in the current Pace options we move from 1 day to 1-2 moves perr week so we chose the closest to what we wanted.

What are your opinions on Pace options?

By the way, we can't even think of enforcing the current setting since all but one of them is impossible if the players are in very different time zones.

[Johnny5]

Hi Savin

Would it not be easier to only have three paces of play i.e. slow, medium and fast?

The reason being is that I have had many games with people where they have said that they will make 1 move per day and only seem to make a move the day before the forefit timeout.

People's commitments can change through the course of a match but if they only have three options for the pace of the match then they can guess an average pace easier and their opponents are less likely to be annoyed with them!

[savin]

Johnny5,

From a tournament perspective we would only ever use Normal, Slow and very Slow for the reasons you stated. But for some players they may wish to suggest a faster pace.

Very Slow is about the official Pace of the ICCF sponsored tournaments. We could certainly drop to three, but I suspect we would have some people complain. the good news is that with these definaition we could eventually add strict time controls.

[sisyfos]

I don't have any real problems with today's pace setting, although I do appreciate your suggestions and find them better suited for CC primarily since the 'impossible fast' settings are replaced by more sensible ones.

However, I have a problem with the short forfeit times - namely 3 days but also 7 days forfeiting.

The reason I find these forfeit times too short is because CC games tend to last anywhere from three months to well over a year - and most people I know cannot promise to be able to make a move within the short forfeit times over such a long time frame. I know we have the vacation days to catch those unforeseeable situations, but it's not a very clever solution.

Rather I'd like to suggest another approach to pace and forfeits; a solution common within tourney CC and which is both flexible for players and tourney directors alike and also has the benefit of setting a maximum duration of a game.

The basic concept is to copy the timing method of normal OTB play but stretch it; e.g. a normal paced game gives each player 60 days to complete the first 40 moves; how the players plan to use this time is totally up to them, but they forfeit the game if they use up the 60 days before the 40th moves has been made. Remember that each player has 60 days for their first 40 moves making it a total of 120 days. Upon the 41st move another 60 days are given for the next 40 moves and the third time frame gives another 60 days for the remaining moves (regardless of how many) --- thus the maximum duration of the game is 360 days.

In other words; a game is divided into three frames of a predefined number of moves within an explicit time limit. The first two frames should last at least 60 moves in total and no frame can be shorter than 10 moves. The third frame is limited only by time (thus there's no move limit). The time limit cannot be set lower than at least 1 day per move and we might consider an upper limit as well (4 days maybe?). Any excess time from a precious frame is passed to the subsequent frame. The number of moves and days left of the frame is clearly stated on the pages.

These relatively simple guidelines give a wide range of different pace possibilities. A very fast example could be 20/20 + 40/40 + 40. An example of a game where the opening is rushed, the midgame is slow-play and the endgame is normal; 10/10 + 50/100 + 60.

[echecmat]

Hi Savin,

I like the proposition. It's much clearer than the current system. I also agree with the coupling of the forfeit time to the pace setting. I agree with Sisyfos that one cannot foresee what's going to happen in a year, but that's where the vacation system is for right?

Happy new year to all,

Echec-mat

[Blockhead]

Whatever system is finally adopted there will always be some who will find fault. This is hardly surprising given the necessary elasticity required to accommodate the needs and expectations of so many.

IMO the 5 speed compromise proposed by savin looks to be better (tighter!) than what we have at present.

sisyfos proposal to mimic chess clocks has some appeal but I'd expect a small number of miscreant players, perhaps out of spite, to simply sit on their hands when faced with a lost game (or those who so readily take offence during the game dialog). This enforced waiting period could take considerably longer than the present forfeit limits.

We all know of players who treat the forfeit time (minus one day!) as an acceptable rate of play ... their play. Given the data floating around Stan's game server, would it be possible to monitor/check all the ongoing accounts for such behaviour? Yellow/Red card them if & when appropriate!

"You can please some of the people all the time ..."

Best wishes for an enjoyable 2006 to all at SNC

[savin]

The suggestion from sisyfos is an excellent one. Stan and myself spent a significant amount of time thinking about how to implement chess clocks -- we in fact have a completed specification and even developed a working model. There are three main challenges:

1) Chess clocks might be very unfamilar to some of the players.
2) Time zone differrences can be vey unfair to one of the playerrs, especially in fast paced games.
3) In friendly games many players may what the option to "turn the clocks off" at any point in the game.

Number 2), time zone differences, are our biggest challenge. In a game where a player has 6 days to make eacxh move the time zone issue is not really a problem. But if we had a game with 1 day per move then the time zone issue becomes major. Consider one player in Europe who makes a move at 7:00pm in the evening. A person in NY replies to the move at 7:00pm in NYC -- 6 hours later. The European replies to the move at 7:00pm their time on the next day. Both players are making moves at the exact same time each day, but one player is getting charged with 6 hours per move, while the other is getting 18 hours a move. not very fair.

We came up with a solution to the problem, but it made the clocks work differently that they do in overr the board games. Our solution was to make every move in a game a "sealed move". The opponents clock does not start until they "open the move". If they don't open the move they don't know what the move was. In this model we would keep two clocks. One to track "elapsed time" and one to track "thinking time". In the previous example both players would probably only take 1 hour to make the move from a thinking time point of view. We would give both of them 1 day to open the move.

We may eventually implement the clocks we designed. But to be honest, the clocks would work best for games where a player had at least 3 days to "open the move". Its just very difficult to enforce much more than 2 days to open a move.

[rttchess]

Hi Savin,

I don't want to sound negative but I don't think changing the categories or names will make any difference. People play at the pace they play at. Unless it's enforced by forfeit (I think this would be tragic!), I don't think anything that you do will make a difference. Making these changes would probably not be the best use of your valuable (and limited) time. The current catgeories give an indication of the pace people want to play at (or think that they do). Most fair minded people probably don't accept rocket games if they know they normally play at bycycle pace.

In summary, I think the current system is probably as good as any other we could come up with.

Keep up the good work.

Regards Ray

[centerman]

I would say that the majority of my games would fall into categories 2, 3, and 4. However, I normally do not like to keep my opponents waiting and always try to respond within a day or two regardless of forfeit timeout or pace of play specified. I do not believe I have had any game where pace of play was many moves per day (rocket) or 2-3 moves (airplane). Also, I have found that when I have many games (more than 10 is many for me) my pace of play quickens and I may make moves in several games in one day.

To address the time zone issue I agree that a 24 hour control is needed.
As Savin states: "But by 1 move per day we meant within 24 hours of the opponent's move." This is the way ICCF works. In fact, when playing ICCF games the whole 24 hours must elapse before time (a day) is deducted. To calculate this sort of control I believe would require one 24 hour clock and one counter field that stores the number of days in whole numbers. When 24 hours expire, 1 is subtracted from the day field and the 24 hour clock is reset. If both the day field and the 24 hour clock reach 0, time is expired. Subroutines can be used to control the addition of time in days as in sisyfos' examples listed above, meaning excess time would be carried over. Additionally, this method avoids one player being charged 6 hours and the other player being charged 18 hours as Savin pointed out, and also eliminates the need for 2 clocks to track elapsed time and thinking time and consequently the "sealed move" "open the move" concept.

[Comet]

I check the computer every few days and make moves on my pending games, and if opponents are online, I usually make several moves in the same game. Sometimes, it may be about a week before I sign into Stan's, and other times, I may sign on daily. When challenging others, I tend to select 30 day forfeit time and the slowest rate of play, since my opponent's rate doesn't matter to me.

[savin]

Comet,

One of the problems we find is when one player thinks the game should have multiple moves per day and the otther does not.

In your case since you choose the slowest pace and the largest forfeit the expectations are set. But as an example, lets say you accepted a challenge from someone with a 14 day forfeit and ROCKET pace. Your opponent is expecting multiple moves per day everyday. You move quickly when you ar eon, but may only log on once a week. Suddenly we have a problem. I'm not saying you have evr done this, I am just using as an example of how we get upset playerrs today.

This is why we are considering changing the current definations. Lots of people totally ignore the pace of play and just use the forfeit time out as a guide.

[reyn]

rttchess said:

Hi Savin,

I don't want to sound negative but I don't think changing the categories or names will make any difference. People play at the pace they play at. Unless it's enforced by forfeit.....

I'm afraid I'll have to agree with Ray on this one, human nature being what it is.

I think until such a time comes in the future whereby you can implement some kind of chess clock (as you mentioned) to enforce the pace somehow, best leave it as is.

And, as Ray stated, your time available here is limited.

[mugley]

I agree with Reyn and Ray on this, it doesn't matter what you call it there is a large number of people who just don't care what the expected pace is, as long as they move before the forfeit period is up, that is all they worry about, and they could care less how you feel about it. I've seen too many posts over the years on message boards here about that.

I understand things come up from time to time that will prevent people from moving like they would like, I've had periods where i found i wasn't going to be able to meet the expected pace consistently, but in those cases i have told everyone i was playing that if it was a problem to to them or they preferred not wait, then I would resign, draw or cancel if they wished.

I also understand that with the different time zones it can be deceiving as to what you should expect, nor have I ever considered myself strict about expecting a move per day or etc, but i do expect something reasonable close to it most of the time, for some one such as myself I prefer a fast paced game, its hard for me to get any continuity in a game when I only get a move or two a week and it really hurts my enjoyment of the game

sorry for the rant but it is a sore spot with me that people agree to a pace of a game they have no intention of even trying to keep, regardless of what the person who issued the challenge aid they would like and as you said Savin, "Lots of people totally ignore the pace of play and just use the forfeit time out as a guide. "

[torphichen]

If I may add my two-penneth worth, I certainly think the holiday period should refer to a game rather than person.

Ideally I would like to play with a maximum 2-3 days per move with 30 grace days for emergencies, holidays or times where an extra bit of thought is required.