All too often, threads pop up in which a player who's opponent has had time added back to their clock as a result of lag compensation accuses people of hacking the clocks and in some way "time cheating".

It seems to me that there is a very simple fix for controlling the way lag compensation is perceived:

• Rather than add time back onto an opponent's clock as soon as they've moved, bank the lag and simply delay the start of their clock on their next turn by that amount of time.
• If a player moves more quickly than the amount of lag they have banked, their clock simply wouldn't count down at all, and any remaning additional lag would be banked and carried forward to their next move.

I believe that this would completely eliminate clock's jumping and gaining time and phantom time-outs in which an opponent's clock runs down to zero and sits before they suddenly appear to get awarded more time, and as a result much of the confusion and frustration surrounding this behaviour.

All you;ve done is replaced one problem with another . Instead of clocks gaining time, people would complain that their opponents clocks werent running properly.

A good effort though.

Well, the fact that lag needs to be compensated for can't be eliminated, so no matter how you do it there's going to have to be some "clock manipulation" performed by chess.com.  As a result, clocks appearing to not run properly is inescapable, however how you compensate for that lag can have a big impact on how it is perceived by the users.

I'd put forth the theory that this mechanism is much less likely to be noticed as often, misperceived as badly, and also much easier to understand when explained as "lag compensation", as that's exactly how the clock would respond: it would lag.

Much simpler would be to allow screening for lag in seeks. At ICC, for instance, players can set a formula that excludes challenges from players whose lag exceeds a certain threshhold. They also have the option of entering the 1 0, 3 0, and 5 0 pools in which the server screens out those with lag above such threshholds. Very rarely has lag been an issue among my opponents at ICC, while it often seems the norm here and FICS.

Sure, but I think that these are two different issues.  Yours is a workaround that allows people to avoid those with bad lag.  Mine simply addresses the optics, and as a result the perception associated with lag when it does happen.  No reason both can't be implemented.

I prefer to relegate those with bad lag to second-class citizenship.

Others have addressed well enough why the current system seems preferable to your innovation. However, if yours had been implemented in the beginning, they would defend it against efforts to suggest the system we now have. People like what they are accostomed to. When the lag is severe, it will impact rapid games and throw off those that use the rhythm of the game more than clock watching to comprehend the passage of time.

Those with lag can play each other. If I see orange or red, I generally abort the game. OTOH, whan I play blitz on my iPad in a coffee shop, I notice that I often show orange. I'm happy when my opponent does not abort as I might have done.

Again, we are talking about two different issues.  This is no way addresses the issues at the root of lag, and how to avoid them if you have a relatively good connection.  It simply addresses the optics when lag is already present.

People clearly don't like what they are accustomed to, this can be evinced from the many threads complaining about (and often misunderstanding) the behaviour of the clocks when lag is present.  This recommendation doesn't purport to eliminate or allow people to avoid lag, please don't confuse it with something that does.

One problem with your system is that when my lagging opponent has one second remaining, but has "banked" 25 seconds due to lag early in the game, his clock will not wind down in the last few moves.

That's what already happens much of the time in my experience, except that occasionally it will go to zero before returning to one.

No, it won't.

How is that worse, though, than what happens now with the clock stalling at zero, only to get the time added back?

The benefits outweight the drawbacks, especially as the drawbacks are effectively a wash with other existing issues.

I'm not certain that it is worse, but I do think it is no better. It is an interesting idea.

You already stated that they could bank the time, hence my scenario:

I don't dispute that, but currently, in the same situation, the clock runs down to zero and sits there.  People think they've won, their opponent appears to have timed out, and for those that don't navigate away from the game and end up losing themselves, they get to see their opponent's clock jump back up, much to their confusion and frustration.

I agree there are problems. it would be nice if we could attract some smart people to your thread. Although I disagree with you, I welcome discussion on your proposal.

Would it work better for most users? I doubt it, but my own perception is obviously colored by a different priority.

  Sure, it's nothing more than a perception manipulation, but I like the idea.

I've had fantastic success with perception manipulation workarounds for software problems that were otherwise unsolvable.

lag