USA Rankings now Live

[Khyodee]

Check out the ranking site http://usa.malifaux-rankings.com

If you wish to submit tournaments just send them to us@malifaux-rankings.com and they will be added to the site within a couple days. I'm taking all events that happened in 2016 so long as they had at least 8+ players. Additionally the team section will be based on state unless I'm convinced otherwise.

 

[Victoria]

Forest Hall
Nick Hrenda
Bryan Fairfield
 

I know that these names can be added to team NE!

[LordZombie]

This count for all events including Henchmen Hardcore? That the format I use with my community the most.

[goth]

1 hour ago, LordZombie said:

This count for all events including Henchmen Hardcore? That the format I use with my community the most.

From the "info" section:

All games will be played with a master. No Henchmen led games

[Guslado]

Has Califaux been submitted yet?

[Veskit]

Email sent.

[tomjoad]

On 6/23/2016 at 11:55 AM, Guslado said:

Has Califaux been submitted yet?

I think it hasn't been; my name isn't on the list, anyway.

[Kolath]

Cool!

What is the intent of the "Teams" section? I'm not sure why players should be grouped by state... we don't really organize teams along state lines (case in point the DC-metro area spans DC, VA, and MD and we all frequently play together and team up at events).

[Tawg]

On 6/22/2016 at 10:38 PM, Khyodee said:

Check out the ranking site http://usa.malifaux-rankings.com

When I open that website tab, should it read "Malifaux UK Rankings"?

[Khyodee]

Yea it does, I'll see if Kai can change that since its basically a copy-pasted version of the UK ones.

[Khyodee]

I've not been sent the Califaux results for this year.

I'm open to other ideas other then states, but it seems like the easiest way to organize areas right now. I could also do things via regions, I would just need to know which states should be placed together.

[SevenThirtySeven]

Great work on this project! I'm sure this will be a great asset as the competitive scene for this community grows.

 

A quick question on the ranking algorithm: From a glance it appears that players earn points from where they place at events, and those points increase commensurately to the size of the tournament.  How easy/difficult would it be to also add a variable which modifies the final point total based on the quality of the opponents?  For example, I think it would be far more challenging to place 3rd at a 20-person event where they played against 2 or more top 50 players (or maybe beat a Top 20 player, etc), than a player who won a tournament with 15-20 people and didn't play anyone in the Top 50, even Top 75.  I've also seen players finish tournaments #2 or #3, but due to limiting most 1 day tournaments to 3 games, didn't have to play anyone at the top tables all day.  Adding this extra variable should be easy since player rankings are readily available?

On the downside, this might add an unneeded layer or complexity or create a points bias for established metas (such as the various East Coast metas).  

Just curious, thanks!

[Adran]

15 hours ago, SevenThirtySeven said:

Great work on this project! I'm sure this will be a great asset as the competitive scene for this community grows.

 

A quick question on the ranking algorithm: From a glance it appears that players earn points from where they place at events, and those points increase commensurately to the size of the tournament.  How easy/difficult would it be to also add a variable which modifies the final point total based on the quality of the opponents?  For example, I think it would be far more challenging to place 3rd at a 20-person event where they played against 2 or more top 50 players (or maybe beat a Top 20 player, etc), than a player who won a tournament with 15-20 people and didn't play anyone in the Top 50, even Top 75.  I've also seen players finish tournaments #2 or #3, but due to limiting most 1 day tournaments to 3 games, didn't have to play anyone at the top tables all day.  Adding this extra variable should be easy since player rankings are readily available?

On the downside, this might add an unneeded layer or complexity or create a points bias for established metas (such as the various East Coast metas).  

Just curious, thanks!

I would say very difficult. 

Assuming its the same as the UK, you need to send in an end placing at the moment. You don't need to send in a list of who played who. 

How do you judge the quality of the opponent? everyone opponent in the top 10 gives you 100% of the score, everyone in 11-20 gives you 95% and so forth? you then add a time requirement to the results because if I send in a march result and the ranking have already process a couple of april results that signifigantly changed somones ranking (Say the were unranked in March, but did really well at all 3 events, by the time you get the result from the first event, the ranking from events 2 and 3 would have been worked out. That could make a huge change to the effect on peoples scores forcing you to re-calculate everything. 

I'm also not completely sure that your person finishing second at a 3 day event managed to avoid the Top tables altogether. to finish second in a 3 round event you probably need to win all 3 games, which normally means you have to face people that are also currently winning all their games. Sure, every time there is a an "easiest" undefeated player, but they still had to beat an undefeated player. 

 

[SevenThirtySeven]

8 hours ago, Adran said:

I would say very difficult. 

Assuming its the same as the UK, you need to send in an end placing at the moment. You don't need to send in a list of who played who. 

How do you judge the quality of the opponent? everyone opponent in the top 10 gives you 100% of the score, everyone in 11-20 gives you 95% and so forth? you then add a time requirement to the results because if I send in a march result and the ranking have already process a couple of april results that signifigantly changed somones ranking (Say the were unranked in March, but did really well at all 3 events, by the time you get the result from the first event, the ranking from events 2 and 3 would have been worked out. That could make a huge change to the effect on peoples scores forcing you to re-calculate everything. 

I'm also not completely sure that your person finishing second at a 3 day event managed to avoid the Top tables altogether. to finish second in a 3 round event you probably need to win all 3 games, which normally means you have to face people that are also currently winning all their games. Sure, every time there is a an "easiest" undefeated player, but they still had to beat an undefeated player. 

 

Fair enough! Sorry, not trying to come off as being overly critical, I think the ranking system is a great addition to the game.  Just offering some thoughts.  There is prob some an easy code to write which could assist here, in case you are ever interested.  I have personally been to several events where folks finish #2 and #3 etc, and don't have to play any of the Top 3 ranked players at the tournament, which is why I bring this up.  Just want to give folks credit for quality games vs top opponents vs running up big differentials on newer players.  

[Adran]

I wasn't trying to be too critical either, but think that its not an easy implementation. And I don't think it really tells you what you want. I was trying to get you to think about what you want it to achieve, and get you to think of your own ideas how it might work. 

I've got nothing to do with the rankings, so I'm never going to want to change the code. I just can see people looking very confused when they are on 300 points  and suddenly drop to 290 because someone the played a week ago had the results from an event they played the week before added. I also think people would feel unfairly punished just because the random round 1 draw placed them against someone unranked, so that even if they won all their games at the event and topped it, someone else could walk away with more ranking points from the event. 

I know the current Ranking system has its faults, which is one of the reasons I don't think its the right way to judge the "quality" of your opponent. I know that one year I finished every event I played above 1 player, but he attended some events I didn't, and ended up with a better ranking score.  Which one of us was actually the harder opponent? 

Also lets say the best ranked player is a Dreamer player, but they are really bad at reckoning. (they always lose that strategy but win every other strategy). I face them in a reckoning game, should I really get a bonus to my score for winning that game? 

 

[SevenThirtySeven]

On 4/5/2017 at 3:57 AM, Adran said:

I wasn't trying to be too critical either, but think that its not an easy implementation. And I don't think it really tells you what you want. I was trying to get you to think about what you want it to achieve, and get you to think of your own ideas how it might work. 

I've got nothing to do with the rankings, so I'm never going to want to change the code. I just can see people looking very confused when they are on 300 points  and suddenly drop to 290 because someone the played a week ago had the results from an event they played the week before added. I also think people would feel unfairly punished just because the random round 1 draw placed them against someone unranked, so that even if they won all their games at the event and topped it, someone else could walk away with more ranking points from the event. 

I know the current Ranking system has its faults, which is one of the reasons I don't think its the right way to judge the "quality" of your opponent. I know that one year I finished every event I played above 1 player, but he attended some events I didn't, and ended up with a better ranking score.  Which one of us was actually the harder opponent? 

Also lets say the best ranked player is a Dreamer player, but they are really bad at reckoning. (they always lose that strategy but win every other strategy). I face them in a reckoning game, should I really get a bonus to my score for winning that game? 

 

Thanks, good points.  I hope it's clear from the deferential tone of the original post, my intent is just to try to find ways to improve the process.  It doesn't mean I am specifically sold on one idea on how to do this, so I appreciate the discussion.

I think locking in rankings at the time of the games would be the easiest and simplest ways to calculate. Those rankings aren't too hard to find and document.  Sure, it wouldn't mitigate errors like you mention (a player going on a great run of tournaments over a 3 month span), but I think it would still be better than a system agnostic of "strength of opponent" elements.

I also think the adjustments should be subtle, so that these will adjustment someone's ranking by 20-30 points over the course of a year, no 60-80.  If someone complains that they got an "easier" draw of opponents, they shouldn't, because they probably stomped those newer players and earned a win with a great differential - which would always outweigh strength of opponent adjustments.  

That's one of the big things I hope this could address - score inflation against new players.  I have witnessed many scenarios where people have been artificially propped up by 10-0 scores against brand new players in tournaments.  Malifaux is a hard game, and the learning curve is steep.  Even a player with average experience has a great advantage vs. a brand new one.  And like I said earlier, I think it would also be nice to reward players who either played very tough opponents all tourney (and prob didn't have the W/L ratio to place as a result) and reward players who placed while also playing the toughest opponents.  Given the data we have, I don't think that would an insurmountable change as long as TO were willing to write in current rankings of players in their tourney when they submit.  Players could probably help this process too. 

As for receiving bonuses for playing people who have a master who is bad at a certain strat.... maybe they should play a different master who is better for that strat in that game?  It's the beauty of this game system.

 

To turn the question back to you, you seem to be defending the status quo.  I think we both agree that no system would be perfect, and all we would search for is a varying degree of improvement.  Nothing I'm proposing would lead to a "perfect" system, but I'm just searching for a modicum of improvement over time.  I don't know of any system I've ever worked with that had a "perfect" fix, either professionally or in games.  To that point - do you think this system is good enough so that we shouldn't try to improve?  What improvements could me made that would be easier or harder?

 

 

[Adran]

I guess the easiest way to answer is state what I want from the rankings system. 

I want something that encourages people to play in events, because I  want the ability to play against more people.

I want a simple way to see how I am doing as a player. A way to judge myself against others, and something to allow me to banter with my mates. A reason for the event results to be published somewhere so I can see what happened.

I like the UK Masters as an event for the top players to play at, so a way to determine the "top" players is needed. 

 

 

The biggest advantage of the current system is that it is simple. I can play at an event, and as soon as I know my placing, I know how many points its worth to me. Every event is entirely stand alone. (Thus does not require 100% player buy in to the system). 
It requires little effort from the tournament organiser to do to submit the results. ( Its all done by volunteers, I do not want to make their job harder than  it has to be. TOs and the people running the rankings sites have done and do amazing things for the community but sometimes things get to much for them.)

There are several disadvantages in that it rewards players who can make more events, it rewards big fish in little ponds. 

 

Whilst you can change the system to reduce or possibly remove the disadvantages, I feel your suggestion does so at the cost of the advantage, and bring in different disadvantages.

The system has changed several times over the years, and there have been quite a few discussions, often on these boards as to the pros and cons. I'm sure there will be many more over time. 

 

I don't like Strength of opponent as a concept without a good way to judge it, and I think the rankings aren't a good enough way to judge it at the moment. 

 I've seen several players do "silly" things for charity and so forth and wouldn't want to discourage them. I've seen people play at events with silly aims, because they want to challenge themselves, and not treat every single game they play as them needing to be "their best". 

[SevenThirtySeven]

Your point on not wishing an extra burden on TOs is well taken.

As far as simplicity, I don't think most players know off of the top of their heads exactly how many points they accrue from placing at tournaments like you do. I'm also not sure if knowing the exact amount of points they accued right after an event is critical to them.  I'm not sure how real of an objection that is.

I think many of your other objections are not necessarily against a modification to the system but against ranking systems in general.

I guess we'll have to agree to disagree, good sir.