this week's ladder looks like this now:


Can you see the problem?
I can see 11 players, and 10 are UNKNOWN for me. I am on line every evening (CET) since July 2009.
To get 3000 ladder points you have to win at least 12-13 games in a row! I don't think it's possible for new players. Some cheating must be present in the above result.

If we allow players to cheat on ladder, we'll lose ALL new players who won't be able to collect a single gold for free. Please do something!


My proposal is again:
make 3 ladders:
1st - the Elite crew: total prizes= 800 gold. For players above 1300 rating and 100 wins. 20 players awarded
2nd - newbies: total prizes = 300 gold. For players below 50 wins, regardless of the rating. 10 players awarded.
3rd - the others: total prizes= 400 gold. 10 players awarded.

total in prizes: 1500 gold, like now. total awarded: 40 players like now.

When the ladder starts, you will be assigned to a league. You can play ANYBODY in the same league, but nobody from outside!!!! (so, with 1301 point you can face someone with 1700 in Elites, or with 800 you can play against 1600 in newbies).

Don't allow multiple accounts in ladders. Check the IP and disable Similar button for the alts if a player is already in the leagues.

I wonder if/how many of the people topping the ladder each week are alts.

The point system seems strange. I'm not sure why it would match anyone against someone they get 0 points for beating, but I've had it happen. If your rating is low enough, you just need to play enough games and win occasionally to top the chart.

That's an interesting idea, it would probably help discourage "cheating" although since there really is not a rule against it, that word is not technically appropriate. Unless I am mistaken, and there is a rule against it, which I would like to see to so I have something to show to people who I catch doing it, as Ive seen about 2 people doing it this week alone, 2 of them are in the top 5 of your list.

-Noob

I'd like to strongly agree with PredatorORC as well as others in this thread on Resignations. I not only see players resign because they haven't pulled their dream hand -- they also resign because they haven't pulled their multis (or colluding friends).

The first step to fixing the ladder is to count a resignation upon meeting your opponent equally to a 20 glory victory -- for both the winner and the loser. As others have noted, sim is NOT about choosing your own opponent. Also, according to this Ladder Changes post from Jed, this should already be implemented (but isn't).

Now that "fix" will exacerbate the other problem with the ladder -- namely people who either have multiple accounts and play themselves or, people who have friends who take turns losing each week. Several suggestions to combat this:

1) Apply a steeper discount to repeated games -- instead of dividing by 2^n, divide by 4^n.

2) Keep track of the number of unique players one ladders against during a week -- apply a bonus to the ladder rating as this number increases. If this unfairly penalizes players with a higher rating, consider weighting this based on the number of "similar" players who have participated in the ladder that week.

3) Keep track of each players *lowest* actual rating during the week. Incorporate this number into the ladder rating to discourage people from artificially keeping their rating down. For example, replace the current ladder rating with a hidden variable tempLadderRating and make the actual ladder rating equal to the square root of tempLadderRating times lowestRatingOfWeek. It is possible some kind of scaling will be needed on the lowestRatingOfWeek variable.

4) Internally track the total number of moves (or rounds) per game -- consider monitoring people who consistently are high on the ladder and have a very low number of moves/game or rounds/game or unique opponents. Or, if monitoring is undesirable, at least publish this information so those people can be ridiculed.

5) I also like PredatORC's specific ideas. However, I might add that wins should NOT include SarsAI -- I'd go further to say SarsAI totals should not be what I see when I mouse over people. Those wins are worth little more than the ones people get with multiple accounts.

Thanks,
KP

Wouldn't that effectively make it easier for "cheaters" to make more points in less time?
They would just get in a game with their friend or themselves or whatever and just resign, saving time.




This would be kind of pointless, because people would just use different alts every week instead of the same one every time, making this a waste of time and effort.


Also, I'd say a few of your ideas would make people not want to play as much for fun. I play a lot of challenge games with fun decks or against better players just for fun because i don't really care about rating...but if people were worrying about their rating all the time because it makes them get less gold, they probably wouldn't play as much for fun, and that would be a sad day

Just a few thoughts

-Noob

A rating should reflect how good you are. The problem here is that there is a financial incentive to deflate your rating -- which doesn't make things more fun for anyone except, perhaps, the cheaters. If you want to play fun games, play unrated games.

As for the other comment, it isn't much harder to play two moves and resign (or have the designated loser play their first building off the flux well and then pass 10 times). On the other hand, it upsets me when I'm waiting for a similar player and a string of people challenge and resign immediately with either a weak excuse or no comment at all.

I couldn't agree more
a rating should reflect how good you are, however that is just not the case in this game.
Even people who do not purposely lower scores still have much lower scores then would reflect their capabilities, and vice versa.

And yeah, i know the feeling there...simply stating that it might create as much of a problem as it solves. I just don't think that it is a problem solved as easily as making resignations equal to playing a whole game.

I would like to suggest some food for thought.

Current Scenario:
Currently, ladder SIMs are arranged for people with similar points, not taking into account the number of games they played at all. I would like to suggest that the number of wins reflect the true number of PvP wins, excluding wins against computer. I believe this would be a better measure of a player's true competencies, how many games have one won against another fellow player- instead of rankings.

My reasoning:
Rankings are altered by a mathematical formulae, and after seeing many players who have well over 700 wins but below 1000 in rankings. I would admit that I would not concentrate as much in non-ladder games. Well, think of it this way, the system rewards those with lower ranking points, it is not a wonder to think why many people would be motivated to keep low ranking points. I would personally not sim anyone with much more wins than me no matter how high/low thier points are.

Solution:
We need to differentiate between normal games and ladder games. If normal games affect ladder (in terms of ratings) we can be sure that there would be an infinite number of ways to swiggle through the 'rules'. I understand that ladder is the means for poorer players to have an even playing field to get the weekly gold. Hence, I propose that we have a separate permanent ladder ranking, making it fairer, such that the higher you go, the harder to get gold, which is the initial intention of spreading the wealth to new players. Weekly remunerations can then be based of ladder improvement.

Drawbacks
I can forsee that this system is also prone to abuse, should friends cooperate to ladder working. However, the payoff would be much less than the current system. Also, this would prevent point tanking to retain fake ranking points for ladder purposes.

Zephsk: Those suggestions are all still quite easy to work around, basically i can sum it up in one word- alts.As long as players have alternate accounts, they'd always be able to have an account with a low win count and low ladder rating in your permanent ladder rating scenario. Not only that but when you say That is'nt completely accurate, because even if everyone played by the rules in your scenario, the people who buy cards have a leg up on people with the same wins, ratings, and skill that did not buy cards , and that is just a fact of the game.


edit- just fixed some mistakes

I think the allegations of ladder fixing are overstated.
- Just because you've never seen a player before doesn't make it an alternate account. This game draws players from all over the world. Even if you play every day, you may not ever be on-line at the same time as someone 9-10 time zones away. But they are still there. Personally, I've played legitimate games against at least 4 of the names on the list in the last few weeks. Won some, lost some. That’s how it goes – both in a single game and in the larger ladder contest. There is always next game, next week.
- Re-read how the ladder works and how it is scored. You get sim matches based on ranking, with uncertainty setting limits on the range of rating you match against. The ladder points system is complicated, but read my analysis of it here. Actually, when it comes to earning ladder points, the contestants’ ratings matter far less than their uncertainties. The ladder is set up to reward people for wins against players of similar rating – it does not reward you for beating someone with a high rating (relative to your own). You can’t GET a ladder match against a person with a much higher rating than you – that person is not considered similar. It rewards you for beating someone with a similar rating and a low uncertainty. Thus, wins against players with high uncertainty will get you little or no ladder points. Actually, that may be the source of Nighthawk's complaint. The way the ladder currently works, it is very easy to get sim matches against new players and you will get very few (if any) ladder points against them. Of course, the flip side is new players are usually pretty easy to beat so you shouldn't get much reward. Sucks if you consistently get ladder matches against them, but there is an element of randomness in this. I think that is unavoidable.
- Remember that multiple wins against the same account get fewer points than the 1st win. It is possible to get 3000 ladder points with as few as 8 wins and no losses, provided both participants in the game have a low uncertainty (under 75-ish). This means to get 3000 ladder points by cheating one would need at least 8 friends and/or alternate accounts with accounts that a very low uncertainty. If the accounts have medium or high uncertainties, more accounts are needed. I guess this means it takes considerable time, effort, and co-operation to cheat the ladder. Someone would need to spend a lot of time generating alternate accounts and then driving down the account’s uncertainty but if a person has not much of a life I suppose it is possible. Of course, a player would probably be better served in the long run to hone his/her skills and win the right way. I suspect most people will realize that pretty quickly if they try to cheat.
- Regarding resignations, I started that thread and I especially hate resignations after mulligans. New players resigning when they see they are up against an experienced player… I have mixed feelings about that. On one hand, getting your ass kicked repeatedly is quite a learning experience and will improve your skills in the long run. On the other hand, getting your ass kicked repeatedly is rather demoralizing and I can understand a new player getting tired of it and wanting to avoid it.
- Finally, remember no system is perfect. The developers have limited time and resources to improve the game. The ladder system is not perfect, but it is pretty robust and relatively hard to cheat. Remember that any new system will have loop holes, too. Look at it from week to week. Every week there are a few names which are near the top, but there are also new names. You can call these new names alternate accounts and cheaters, but it could also just mean the game is growing. Especially with the excitement from the release of Borderlands, I would like to think the 2nd reason is the real reason.