lichess.org
Donate

If a player runs out of time, they should lose the game, always!

No, Wulfsiege, the game was NOT drawn by the concept of the "Insufficient Material" draw rule.

It was drawn by this site's *false* implementation of it.
The Insufficient Material rule says nothing about having only a king and a knight, or a king and a bishop.
It says that there is not enough material for a checkmate to happen by either side.

So it was not a positional win when the game timed out; it was a material win that got counted as a draw.

If you really are not sure about this I suggest you look up the FIDE definition of "Insufficient Material", because yours is obviously wrong.

The easy fix is to never count a game as a draw on timeout except a) the enemy has nothing left but the King and you timed out b) it's King+Bishop vs. King+Rook, because those are the only 2 regular material force balances that can constitute "only one side" being unable to checkmate, while the other still can.

It's equally as fair as this site's incomplete support for the "Insufficient Material" draw rule. That uses a list of basic material schemes! So why shouldn't draw on timeout? It's fair and balanced; your argument is not.

Any counterexamples to where it fails?
Show me.

Otherwise, stop being an arrogant ass and start giving valid examples.
Because it's really you who's dragging this out under the simple notion of having an "original" opinion that is not going to hold through the whole way.

Most people here already have caught on that I am a moderator here, perhaps not yourself, so if you think that "I'm" dragging this out and "proving you right", any attitude issues won't spontaneously win you any debate against staff.
Mephostophilis is completely right
Join the Darkside and xianGay-o is completely dumb!
Enjoy the rest of the day
the good player must be careful not to tender for this
Well, I started this thread with the title it has because I really favoured the always-lose system.
And I still think it's least bad because it's intuitive and there's no need to care about all the possible positions.

But I see that Wulfsige strongly dislikes it for whatever reasons, so maybe we have to find another compromise.

Mephostophilis, I'm not completely against your list idea (actually it wasn't your idea, but rather Dougthehead's idea, see post #9 of the other thread) as long as it avoids "false positives"; meaning I don't want that any game gets rated as draw due to the list although there's a win possible.

My proposal for the list would be the following:

(assume that the player who ran out of time is named "NoTime" and the other player is named "HaveTime"; I'm sorry for the parentheses, I hope they help understanding)

>>>>
Timeout means loss, except

(HaveTime has nothing but a king), or

(HaveTime has nothing but a king and one knight), and (NoTime has nothing but a king and (zero or more queens)), or

(HaveTime has nothing but a king and one bishop), and (NoTime has nothing but a king and (zero or more out of queens, rooks, and (bishops that are of the same square colour as HaveTime's bishop))), or

(HaveTime has nothing but a king and (two or more same-square-coloured bishops)), and (NoTime has nothing but a king and (zero or one out of queens, and rooks), and (zero or more bishops that are of the same square colour as HaveTime's bishops)),

then it means draw.
<<<<

I tried to set up this list in a way that there's never a draw if checkmate could be possible. However, I'm not sure about this.
So if anyone can find a checkmate position which would be declared a draw by the above list, please post it.

Of course, the above list is not complete: Sometimes a draw will be declared loss by the list although no checkmate is possible. The only thing that *should* be secure is that no loss will be declared draw by the list although there's a checkmate possible.
As I said: I'm not sure about this, please post improvements.

Anyway, I still dislike the list idea, because I'm never feeling sure whether they really don't contain positions where a checkmate would be possible.
I'm only willing to agree to such a list if there's really no consensus for the (in my opinion better) always-lose system.
We already use a reasonably complete list for Insufficient Material automatic draws, but even that list is not quite as thorough as the one you just posted.

So I feel really that there is no cause for concern.

It is an overall improvement to the rules with no drawbacks, if you think about it.

Positions you gave like that full army vs. King and Queen where the Queen will inevitably be captured on the next move and so that side cannot win, are not currently covered anyway. Supporting them should *not* be the top priority.
Hooray!

I'm the first one to find a mistake in my list! :-(

Well, not really a mistake, rather an inaccuracy. I found another material combination where we can certainly know that there's no checkmate possible.

When I posted the above list, I intentionally excluded cases where HaveTime has multiple same-square-coloured bishops, while NoTime has multiple queens/rooks.
I did so, because I had the following checkmate position in mind:

White: Kd3, Bc2, Be2
Black: Kd1, Rc1, Re1

White checkmated Black?
No!

There's simply no possible last move before the checkmate. You can't do a double check with two bishops. At first, I wasn't sure about that because I've already seen a double check with two rooks. Yes, it's possible, you can do a double check with rooks (but not with bishops)!

So, we can add it to the list. And this "adding to the list" has even the nice side effect of making the list shorter, because some cases can be combined together.

The updated list would read:

>>>>
Timeout means loss, except

(HaveTime has nothing but a king), or

(HaveTime has nothing but a king and one knight), and (NoTime has nothing but a king and (zero or more queens)), or

(HaveTime has nothing but a king and (one or more same-square-coloured bishops)), and (NoTime has nothing but a king and (zero or more out of queens, rooks, and (bishops that are of the same square colour as HaveTime's bishops))),

then it means draw.
<<<<

Whether I like the list idea or not? Well, not sure anymore, I'm just about to change my mind... Let's see whether other opinions appear here in the forum!
Meph, I said probably about 10 times earlier in this thread that the Lichess obviously only takes material and not position into account because obviously there's no way to have an engine in the background checking positional mate. This was thi's response 2 pages ago. I was simply attempting 100 more times to explain this statement because Anon kept being confused. Now you seem confused. Why on earth are we trying to find a solution to something that requires either having a background engine, cycling through endgame databases, or arbitration. I really don't understand how the coding can in any way fix the false draw situations (which probably happens only in about 1 in ten thousand forced draws to begin with) rather than simply make assumptions about material draws (which is still a step better than "always lose on time" which is implemented in plenty of other places). But obviously you guys have come up with some half-baked scheme to fix it, so I'm done attempting to discuss it. Whatever is decided in the end will likely never even have an effect on a single game I play, so I don't care anymore.

Good luck, and goodbye.
No, you're confused because you keep responding to relevant facts, with irrelevant sidetracks.

If the material detection was correct it would not have said that King+Knight vs. King+Pawn or whatever was "insufficient material".

So that proves that the material detection scheme is flawed.

Your response is, "thibault doesn't take positional play into consideration for determining draws, so your argument is invalid."
More irrelevant facts. Just because something is a fact does not mean it matters.
So, no, you're confused and you wish to remain that way because you don't really want to see anyone else's point of view; you just want everyone to see yours. :)

And what was that about...?
"requires either having a background engine, cycling through endgame databases, or arbitration."

Excuse me but did you not know that this "arbitration" is the exact same type of listing that the Insufficient Material draw rule uses to draw games? It uses a list of material-based schemes; we're doing the exact same for the other draw rule of time-out.

You have serious mental issues if you think that just because thibault has his name on the "arbitration" for Insufficient Material while we have our names on the exact same "arbitration" for time-out draws.

And some major respect issues because many of us unlike you have been contributing to the Lichess site in various ways (translations, bug fixes, Wiki), while you, a player less than 1100 rating, think that your understanding of a position that happens "1 in ten thousand forced draws" is better than everyone else's.

dougthehead's was far from the first thread complaining about a drawn game that he should have won, that timeout rule falsely declared as a draw.
That's why we're fixing it!
Because it's been happening time and time again, while you express your stupid paranoia about us "getting our names credited" over something and know nothing about the history going on in this forum.

This topic has been archived and can no longer be replied to.