I’m going to claim that, philosophically, adjustment of rating is done in order to achieve an accurate estimate of strength rather than to punish the loser or reward the winner. When a player times out due to not playing turns for a days or weeks … that’s not really the same as losing a live game on time. I think the current practice of adjusting rank for timeout makes the OGS ratings less accurate.
Once I entered a few tournaments and my rating ballooned up 3 to 4 stones due to timeouts by much stronger players. Then they were effectively rated too low and I was rated too high.
(Meta: Just in case you’re wondering why I “liked” both posts: first I was easily convinced by saxmaam’s OP, then pbgarden showed me that I didn’t think enough about this, and I’m not taking away my first like now )
Hey, pbgarden. I guess this must be the reason that DGS has a setting in which you click a box to indicate that you do not want your rank to be adjusted for timeout win, but timeout losses result in an adjustment whether you want it or not. I realize that someone very attached to his current rank could simply timeout to avoid the adjustment.
Maybe it will be possible to establish some rules reguarding the adjudication of such games, as well games where the opponent is uncooperative during scoring. Perhaps the won-by-timeout side should be allowed give up the win, if they choose so, and games whose results are disputed can be passed on to moderators (or be moderated by multiple volunteer OGS players, to avoid overworking the moderators)
This sounds like adding more complexity to the site for what I suspect is not a huge problem in the end. If you’re worried about your rank inflating too much, you can certainly ask a mod to adjust it back down. Alternatively, you can let it adjust its way back down naturally after a couple games. It will all wash out in the end anyway - law of large numbers and whatnot.
Plus, that checkbox might have the direct opposite effect of enabling sandbaggers even accidentally. Example: “Oh, this guy’s gonna time out. I’m ahead by a bit but it’s still midgame so I don’t know how it would work out in the end. I’ll just not count this towards rank.” is a totally understandable sentiment, and yet it will still cause that player to start sandbagging a bit.
I just don’t see this is a big enough concern in aggregrate to warrant such a change. Certainly it happens on occasion, but I don’t see it happening as a rule.
Seems like it would allow for easy abuse of the feature by letting people purposely abandon games and let it time out once they realize they are in a losing position.
Both are right to some degree. I prefer a feature OGS used to have, with a “rating confidence” number. Your rating confidence starts at (I think) 10%, and goes down every day, and with each timeout, and goes up with every game you win (or is it every move you make?). Then, the lower your opponent’s “confidence” number, the less your own rank is affected.
It’s already frustrating enough when, in a strong position, an opponent times out after 18 moves and the game is cancelled. If we’ve played 180 moves and my winning position is disregarded simply because my opponent timed out, i don’t think it’s an exaggeration to say that would ruin the server to the point of finding somewhere else to play.
I think that the penalty for timeout should be higher.
When you sign up for a game, no matter what “mode”, you commit to completing it.
If you have decided you are going to lose, then resign, don’t just walk away leaving your opponent wondering what is going on.
There’s no reasonable excuse for why a correspondence game should be allowed to time out, in general. Of all the modes, it’s the easiest to somehow find a way to get your turn in.
I have to admit that I have timed out a few times, so I can’t be too judgemental. I avoid playing Blitz games, and even shorter live games, because I don’t like to hurry over a move, and I got caught out a couple of times. It’s easier to set up a correspondence game and then if you are both online at the same time, just play like it is a live game.
A couple of years ago the broadband connection was off in our area for nearly two months. I had to go down to the internet cafe in town to keep my games going, which was pretty difficult. Eventually I caved an bought a cheap smartphone so I could play anywhere, but I definitely play much worse on the small screen. The worst time for me was when the site was having a revamp, and the little black button telling you the number of games with moves outstanding dropped away and I got caught out a few times.
I know some members are currently travelling and have intermittent access to elactricity or internet.
I guess what I am trying to say is that there are valid reasons to time out. Also, people are human and make mistakes; I don’t mind that. But what really gets up my nose is that there are some people who think it is funny to sit on the “Play” page, accepting challenges as fast as they come up and immediately cancelling the game. YAAAAARG!!! I went through a whole spate of it, so much so that I decided to restrict rank to one above mine in all my open challenges. It seems to have worked, but the issue that has developed is I am only playing people who are better than me.
I am learning more, but losing repeatedly.
Upon reflection, maybe I actually need those time out wins.
If you sign up for a correspondence game, please sign up for time parameters that you can meet based on your circumstances.
Then meet them.
It’s really as simple as that.
“Oh, I’m losing, and now it’s too much of a hassle to go to an internet cafe, so I’ll just time out and it won’t affect my rank” does not cut it at all.
I think that a rank penalty for timing out is entirely appropriate. I think the “T” for timing out should last much longer - maybe a minimum of 6 months lets say - to warn people that you are a timer-outer, so they can chose not to play with you.
Like BHyden said, if OGS became a place where “timing out is OK”, I’d find somewhere else where it isn’t, because I think it isn’t OK. At all.
(And in addition to that, the basic premise is clearly flawed because it gives a person who is losing a way out without taking the consequences of losing. That is just obviously wrong.)
GaJ
(There are of course reasonable circumstances for timing out, like a health or natural disaster - a change of circumstance you couldn’t predict. That’s what moderators are for - I was more referring to “normal” course of events)
There is also vacation mode for when something unexpected happens and you have more important things to worry about. Just a couple weeks ago we were in hospital with my little girl’s high fever. She’s fine now but i switched to vacation mode just so i knew i could put it out of my mind and focus on the situation at hand.
I could agree with a system where timing out still ranks the winner but doesnt effect the loser. As someone else has said on here, the rank inflations by timeout are a very nice learning experience as you get crushed back down to your stable rank.
This would still be open to the “Im losing, just timeout” abuse, but as the winner still gets credit, who cares if the loser doesnt rank down? If they consistently timeout from losing positions, then they arent at the right level and wont improve. That is their problem
Now that the ‘get an email when you’re low on time’ feature is actually functional, I don’t think I’ll ever time out on a correspondence game again. ( I used to time out now and then from games i thought were pause on weekends, but weren’t.) That, combined with the ability to turn on vacation if you get too busy should leave corr timeouts as almost non existent unless the person is terribly scatterbrained or doing it on purpose. I don’t think the penalty for timeouts should be lessened in any way, and even could be increased.
)
