20 games before rank might force me to surrender 1 game to reach finished 20 games, but I would still be incentivized to stall up to 4 games. 25 games might force me to surrender 2-3 games, but I would still be incentivized to stall up to 4 games.

Penalty points basically mean that the more you play, the better your rating. It means that one kind of bias (more points for more games) tries to balance another kind of bias (more points for stalling surrenders). It may work to some extent, but I will still be motivated to count how many points I lose by surrender and how many I win by getting another game finished. If surrender takes me 60 points and finished game gives me less than 60 points, I will not surrender but stall.

the more you play, the better your rating

No, rt ladder has that using variation. I'm talking only about penalty points for first 20 games. Assuming those penalty points would be enabled for first 15 games, you are stalling your games 8 and 9, so you'd be 150 points lower than your current rating... And you're right, it's one kind of bias against the other. before I recommended just forcing to finish first 15-20 games, however that might also have other problems, other players might block you from the ladder. With that system (it can of course be tweaked with regard to value) it'll motivate you to finish your first 20 games. And if the system forces you to do that, you are welcome to stall, after 20 games the advantage of stalling is greatly reduced due to lower variation and increased overall game count. I'm still a proponent of non-expiring games, however if people want a system of expiring games, I recommend the system I just mentioned.

Just as a side note stalling does not equal cheating. So while #1 is #1 (since you did not cheat), it doesn't mean that stalling isn't a bit shady or should not be ridiculed.

It is like someone not lying, but also not sharing the truth either.

You're basically incentivizing people to finish at least 20 games instead of just 15. Stalling 4 losses out of 20 games still gives me an advantage, however indeed a smaller one than stalling 4 losses out of 15 games. Furthermore, stalling 4 games out of 50 games gives me an advantage, however a smaller one than if I stall 4 games out of 20.

It's true that one's rating gets more accurate after 20 or 50 games than after 15 finished games, the downside being that you have to play quite much to get that accurate rating in the first place.

You could argue that finishing at least 50 (or 20) games to get an optimal rating isn't too much for a pro. They play more than a beginner anyway.

A beginner will get ranked after 15 games which is good, he doesn't want to play 50 games to get ranked. He will not be able to fight for first place rating anyway before he has finished 50 - 100 ladder games.

Basically, stalling 4 games still gets you points after 50 or 20 finished games but much less than after 15 finished games.

Yes, is a general response to your last post. I still don't get what is your point. My point is the earlier you stall, the bigger the difference is, thus lets create something that won't force people not to stall, but will diminish its effects. I'm not saying you have to make a 20 game cutoff, it'll work almost as fine if you do the same with 15 games.

Ah, now I understand your last post better. I thought your negative points would just increase minimal gamecount from 15 to 20, with people having 15-19 games finished getting ranked, but with decreased rating.

Now I get the impact those different negative points would have. 1st non-expried game would give -150 points if not finished yet, 2nd -140 points and so on. So it's a hybrid of the traditional ladder where finished games count no matter in which order they started, and a ladder where chronology plays a role.

It's an interesting idea. I can't think of all the implications of that in such a short time.

I could game the system a little by first playing the maximal amount of games, ie 5, and at once surrendering in every game that is not looking good for me. Then, after getting 5 good-looking games, I would quit the ladder until my surrendered games are expiring, thus starting with 5 won games under my belt.

Edited 5/30/2014 00:40:29

Yes, they're far more likely to lose the 10-15 games than the 1-5, the 5-10 can potentially be losses as well.

The 1-5 though, despite being easier to win, are the losses you'd be more likely to try to stall since they'd be more damaging.

"I could game the system a little by first playing the maximal amount of games, ie 5, and at once surrendering in every game that is not looking good for me. Then, after getting 5 good-looking games, I would quit the ladder until my surrendered games are expiring, thus starting with 5 won games under my belt."

That's always been an issue of sorts.

Currently I am rank 24 on the ladder.

If I left the ladder and waited for games to expire, my most recent 13 games would put me at a 2055 rating. Assuming I win my next two that'd be last 15 at 2100ish. Which obviously would allow me to leave the ladder and return around rank 5-6 instead of 20-25.

Edited 5/30/2014 01:28:39

With the exception of a certain piggy noone has been no1 for months without delaying to some degree. It doesnt matter if one or two die hards can game the system, it doesnt have to be totally airtight, there just needs to be incentive enough to stop the no1 spot only ever being held by delayers. Any of the suggestions so far would do that.

Btw there is still the problem that if my opponent decides to stall, then I lose points even if I have the game won. If you use chronological sorting, you have to tackle this problem no matter what.

@Piggy - Are you forgetting The Impaller, Zaeban and Sze? I'm fairly sure they didn't delay.

Edited 5/31/2014 10:35:52

And unknown, and Gui etc. Anyway, does anyone has sometihng constructive to say about my idea? Everyone is for, everyone is against? Some feedback is necessary here.

Its a good alternative at least, if Fizzer ever decides to actually change it though. If it stops stallers, I'm all for it.

I agree with all your consideration, but you forgot one thing, that IMHO, is the most important in the ladder as well the stallers thing: play less games possible, give you an important ahead.
I think the system should incentivate to play you more, like give you X more points for every game played after a certain number of games (like +10 points for every games after you played first 30). So in this way became too important play 50 or 60 games unexpired, and stall is useless, bc is better have 50-10 as record than be 15-0
LEt's make an example: a player with 50-10, probably would have like 2100 more or less.
so 60-30=30 * 10= +300 points = 2400 points, that is better or more or less the same of an actual 15-0 as rating

What do you think?
One easy tweak in the rating formula, that solve problems and incentive to play more.

The system cannot do that I think. With 3 months expiration if you want 50-60 games active, you have to be on warlight almost all the time, not everyone can/want to do that. I specifically designed that solution so that normal players that just want to play it slow don't get hurt. Also extra points for games is just not fair, even on rt ladder you don't get those points like that, the variation decrease is slowing down almost exponentially.

Extend the expiration date to six months and increase the number of games required to rank.

Sze, I think the penalty points idea is interesting but it seems like it's just patching up problems with the ranking system rather than actually fixing them.

"With the exception of a certain piggy noone has been no1 for months without delaying to some degree"

i mean in the last 3ish months noone but me has been no1 without delaying. Pulsius, beelzebub, pulsey and there was another guy who was below me in 2nd who loss deayed like hell whos name i forget.

My primary solution would be to not have any kind of expiration date, but it is frowned upon by many, thus I think it is the best solution if you don't want to change the core of the current system. So in a very roundabout way I'm saying you're right :)

leave the ladder as it is. I wouldnt have gotten #1 and such a high rating without stalling. There are plenty others who want #1 but aren't good enough to get it. I realize most of the people talking here have already achieved #1, but lets not ruin other's chances.

Edited 5/31/2014 14:35:18

What are the benefits of no expiration date?

Real ELO in the long term, overall reduced variation, stalling problem reduced to almost 0 after a few months etc.