The ladder scoring now works differently. I edited the scoring above.

Each game you play with the same person in the same week counts for less ladder points.
You can't have negative ladder points.

You may want to include that those rank requirements are considered after rank calculation at the conclusion of the match. This means that you could be trying for ladder battles with "fight similar" and have them not count.

It would be nicer if some points were given if the requirements were met prior to match. Maybe they could be halved if they got pushed out after the match or a small flat bonus given / taken away.

I'd also love to see some sort of indication that a match would actually affect ladder ranking. I guess each player could do their own math, figuring out what the respective rankings might be if they won or lost and how many points were at stake, but the computer can do it much easier. Maybe pop up the potential gains for winning or the potential losses for losing.

I just realized that although it will never display a negative ladder score, you basically have a virtual negative score. After losing 3 ladder games, I've won my last two. My score is 10. I was told that this is because even though I've won the last two, I haven't dug myself out of the hole I dug with the first 3 losses. This means that I DO have a negative ladder score, I just can't see HOW negative it is..

If my score is actually negative, I'd prefer to know roughly how bad it is.

Just so I remember, right now ladder points are apparently calculated as:

(200+(opponent rating-your rating)) / 2^# rematches

So the first battle against someone 50 points higher than you would have at stake: (200+(1350-1300))/2^0 or 250 points. If you had a rematch at the same rating, it'd be (200+(1350-1300))/2^1) or 125 points. The second rematch would be 62.5 (rounded).

From what I have seen so far the usual scheduled ladder resets have been on Tuesdays, but I would like to be able to see is the specific time of the next ladder reset, and the time of the previous ladder reset.

It's tiny bits of information that could be useful to many, and I imagine that it is fairly easily added to the current showing of ladder ranking.

~Wicked~

Here's a quick snapshot of the ladder and why things might be better off if something changed.

Of the top 20 on the ladder, more than half (12) were created in the last 30 days.
Of the top 10 on the ladder, half also appeared atop the previous week's ladder.

You can see when an account was created by searching for members in the forum. I use the wild card (so I don't have to type the entire name) and it's interesting seeing how many names that appear to be duplicate accounts.

I can only see ratings for people online, but all of those top 10 currently online have ratings below 1300 (averaging around 1150), and ratings deviations of less than 50 (all 35s and 40s).

So we've got many of the same "new" people of relatively low rating, consistently beating people of "similar" rating, week after week, and with low ratings deviation their rating will stay low and they can continue to rack up ladder points.

Really new players with a low rating will be beaten soundly by them, and might be discouraged from playing at all. Suggesting that they might win gold in the ladder is rather disingenuous knowing there are many ringers out there.

If this is really a problem, how might this be fixed? Here are a couple of suggestions:

1) Raise the minimum ratings deviation. While this means that ratings will be easier to tank, they'll also rise faster. The time spent gaming the system will be shorter.

2) Change similarity to include deck strength. A deck with tons of odds, rares, and uncommons will often stomp something closer to a default deck, regardless of player rating. Any easy check would be to count the relative frequency of rarity, and come up with some measure there. A more accurate check would be to calculate deck cost based on market info and compare that way.

3) Actively discourage and possibly punish those creating accounts just to game the ladder. Secondary accounts aren't the issue, but nameless ladder gamers hardly add to the sense of community. Requesting that people link their numerous accounts in some way could help too.

I agree with every thing but point 2. There is no way to measure the strength of a deck based on the rarity of the cards or prices on market. In my experience the deck I use now that I consider better than the older ones have no rare cards at all. And if I am right the deck that angelatheist used in GC that took him to the 2nd place had only common and uncommon cards.

Rarity of cards I suggested because that's a simple lookup of a static table. Cost of cards is likely a truer measure but is a slightly more involved query.

I wouldnt be surprised if you had some expensive uncommons / commons in your current strongest deck. Of course very good players don't need the best deck to do well. Give a 1650 rated player like angelatheist a deck or let him create one with commons and uncommons, and he'll still perform like a 1650 rated player.

The problem is that a very lowly rated player with an overpowered deck will likely stomp a similarly rated player with a default deck or starter deck. With chess (what the glicko rating system was designed for) the players have the same starting pieces -- it's a level playing field. Player ratings can't capture deck changes.

What if players are matched also based on how many games they have played or won?
That solves the problem of true newbies getting owned by people of low rating but high skill and/or expensive cards.

That’s exactly what I mean. Even with a low priced and low rarity deck he still performs as a 1650 rating player. So rarity and card cost don’t have anything to do with the effectiveness of the deck and the player.