Beginner Question-Guide to timings


#1

Hi,

I am new to Go and to OGS. I’m definitely at the losing your first 50 games as quickly as possible stage! If I wanted to create a game to challenge others, is there a guide to what the various timing options mean ie Fischer, Simple, Canadian etc? I don’t quite understand them.

Also is there a guide to how results against others affect your own ranking? One that isn’t too heavily mathematical if possible.

Thanks,
Paul


#2

Hello, and welcome :slight_smile:

regarding the time systems, if you really wish to understand it all you can check here: https://senseis.xmp.net/?TimeSystems

However, for the sake of simplicity, I think I can give you a rundown:
simple - You have a certain ammount of time to do each move (you can spend 1 minute on EACH move for example). Overstep once and you lose by timeout.

Byo Yomi - (arguably the most common here) you have some main time (let’s say 15 minutes, that keeps going down whenever it is your turn). Once you run out of your main time you have certain ammount of time to make each move (usually like 30s) and several “lifes”. If you make it in time, for your next move, you will have 30s (or whatever was set) again. If you overstep this time you lose one “life” lose your last “life” (last 30s) and you lose on timeout. This one is kind of complicated to explain, but after you see it once it will be clear :smiley:

Fisher - You have some normal main time which goes down gradually whenever it is your turn. However, every time you make a move this time increases by the set ammount up to the maximum set limit. This is often used for correspondance games where a simple timing would need extremely long main time, which is hard to predict and you also might want to make sure that if your opponent stops playing it will not take 2 months for him to time out…

absolute Quite simple - you have a certain ammount of time that goes down whenever it is your turn and there is nothing you can do to gain more. You run out of main time and you timeout. Carefull about this one, it is easily exploitable and not at all flexible.

Canadian - You have set ammount of time that goes down whenever it is your turn to make x number of moves (let’s say ten). If you manage to play ten moves within the time limit, the timer resets again.

Not possible I am afraid. One of the downside of the system used here is that it is kind of complicated for ordinary people. It works very well, if you trust it, but it is complicated.


#3

My personal advice: Don’t play games with the simple time setting. It is bugged and time will just run out on you at random intervals.


#4

I like Fisher time best for live games and simple w/24-72 hrs per move for correspondence. Haven’t encountered the bugs but I haven’t played that many games here either.


#5

Of course, when you win, your ranking improves and when you lose, your ranking worsens.

If you beat a stronger opponent, your ranking improves more than if you beat an equal opponent. If you beat a weaker opponent, your ranking improves less.

If you lose to a stronger opponent, your ranking worsens less than if you lose to an equal opponent. If you lose to a weaker opponent, your ranking worsens more.

(The logic is that you are expected to beat weaker opponents and lose to stronger opponents, so unexpected results should have more impact on the rankings.)


#6

I believe the ELO system is what weights the score between strong and weak players and the Glicko RD factors other things to make your rating more or less volatile. The idea is to move your rating toward accuracy as quickly as possible.

A high RD means that a win nets you more points and a loss loses big. A low RD means your rating is more stable and is more likely to be an accurate reflection of your position in any particular pool of rated players.

Winning or losing a lot of games in a row, or taking a long hiatus are some things that increase the RD


#7

Thanks for that. Its all a bit clearer now. The timing systems seems quite straightforward and the ranking system is similar to that on chess.com

Thanks,
Paul