Suggestion: change the wording for 'annulled' to something like 'unranked'

It would be totally unfair if games that are part of a correspondence timeout series would be annulled. Fortunately they are not. The winner of course gets that point in the tournament. However, the current wording seems to imply otherwise.

An aside: This wording also seems to indicate that gaining ranking points is the main reward for winning a game and getting a point for the tournement is just a side effect. From my perspective it’s the other way around. The tournement point is the main reward/achievement. The ranking is merely a measure for match-making.


The change of status has to be notified in some way… (Which disapear from annulled to unranked).

Annulled and unranked mean two different things. To turn one into the other simply confuses the meaning of language. If a score cheat wins a ranked game, the game is annulled, not turned into an unranked game. If the cheated game were merely changed to unranked, the display in the history would imply that the win was legitimate but unranked. Similarly, non-games under six moves would appear to be legitimate unranked games. Aside from the moral distinction, there is a practical one. A person’s history can be quickly vetted by scanning for scored games that were annulled.

Until recently, correspondence games subject to the serial timeout rule were simply left as is in the history display and the game tab, but no rating points were awarded. Going back to that status quo would resolve your objection with no violence to the language.

However, I want to take the opportunity to suggest a better solution: ABOLISH THE SERIAL TIMEOUT RULE.

The Forums had quite a lot of controversy over this rule 4 or 5 years ago, which brought to light a substantial amount of abuse of the rule. The abuse consists of deliberately timing out of a lost game and then consecutively timing out of more lost games in order to avoid losing rank due to those losses.

The intent of the rule, as you probably know, was to mitigate the effect of a catastrophic emergency such as dying, being incapacitated, having one’s computer stolen or wrecked, etc. At the time of the controversy, I argued in favor of the rule, because its purpose of preserving the integrity of the ranking system seemed good, and because it seemed the compassionate thing to do. I now think the rule should be abolished because of four considerations.

First, legitimate cases triggered by a real emergency are rare and/or the number of games involved are comparatively small. In other words, even in the extremely rare case of a person with hundreds of games in progress, the hit to the ranking system would be small. I base this on my observations as a mod (when Eugene and I handled all or nearly all the abuse cases) and afterward as an observer of lots of games.

Second, I strongly suspect that most mods have no experience in recognizing the abuse. When I reported the clearest abuse I ever saw, where the player had serial timeouts in clumps 5 times in 4 months, two mods did not recognize it. If that case does not meet the standard of proof, then no case will, which means cheats on OGS have a free rein to abuse the rule.

Third, virtually all cases of abuse are reported to mods (when they are reported at all) as escaping (i.e., deliberate timeout). However, escaping from correspondence games is no longer a reportable offense. Therefore, there is no way to detect abuse unless you happen to watch games as I do and accidentally stumble over an abuse case (e.g., when looking at someone’s history). Again, this gives abusers free rein to cheat.

Fourth, since escaping is no longer prohibited in correspondence games, it seems philosophically inconsistent to prohibit it in cases of abuse of the serial timeout rule.

Solution: Abolish the rule. It serves no good purpose, and its enforcement is practically impossible.

1 Like

You are right of course. My suggestion was only meant for games hit by the serial timeout rule. I ignored that there are other cases of annullment as you point out. Although: what is happening in tournment games in those cases? I assume the other player gets the point? If so, I’d still think that annulment is a strange term, as for me (non native speaker) it seems to imply that there is no result at all.

Edit: Also I said ‘something like’ in the title for a reason: games can be unranked to start with, and ‘not ranked because of reason X’ is something different of course. So yeah, I don’t have a good suggestion for a term yet.

The rule actually damages the ranking system more than its absence would, due to the abuse. Getting rid of it would answer your concern.

Yes, I saw your suggestion. I just don’t have a strong opinion about it either way, that’s why I didn’t comment on it.


Another possibility:

And call those games 1/2^n-annulled? :wink:


Re: abolishing the rule, there was a recent discussion on GitHub too

Personally, I’m not concerned about the effect to the greater rank system. However, it seems like anyone who returns after a mass timeout would have a seriously deflated rank if we abolish the rule.

I join @Conrad_Melville suggestion.
i have seen people exploiting this feature to manipulate their rating too.

the min of timeout is pretty low but i suspect it’s hard to change because of the delay conflicting with the running of the rating system.

we could keep the procedure of annulation for mass timeout by request only to a mod which could help players in real distress. And help the good work of the rating system too.
If annulation is undoable for some reason (delay conflicting) then another way would be to reevaluate the rating of a player reintegrating OGS, by request too. OGS could then think of a calculation to minimize the sinking of the rating to apply to the concerned player.

Anyway the system shouldn’t give opportunities to get overranked by volontary annulation. Or to get some vacation on the detriment of your opponents.