Community

This thread is for discussion of this week's The Week That Was, which goes live Friday morning on magicthegathering.com.

I think the idea is to get more people to EVENTS at stores and not just tournaments. But yeah, players going for those pro slots won't play in events with newbies... Because one newbie with a Legacy "griefer" deck of commons and uncommons can wreck their "rating" just by tying.. And in some cases you can even WIN and lose ranking because your "competition" won against a tougher opponent pairing.

That kind of thing is silly, and it's bad for the game not to have the best players teaching the game.  This will also reward the FNM guys that will never get past mid-pack.. But they show up EVERY WEEK at the store to play. Their points will count for something too in a different pool than the guys that only show up to win stuff.

This article was pointless. This addressed none of the important issues, mainly what for the guys with actual jobs who can BUY the 1000 something dollar decks and can only play once a week, and how going 4-5 and 3 american gps beats winning a ****ing european one! ADDRESS THE ****ING PROBLEMS, DON'T PROVIDE FLUFF!!!

Larabee: The one-sentence description of Planeswalker Points is: "Playing is good. Winning is even better."

um...

Forsythe: "...the way you keep people invested is through a constant stream of positive reinforcement."

Now this one's particularly telling: it's about keeping everyone "invested".

I've participated in a grand total of one (prerelease) tournament ever since Psychatog (and Mudhole) rotated out of Standard, and there's still nothing to convince me to come out of that snap.

Just wanted to point out there every PTQ/states/regionals I played prior to 1999 has an incorrect event multiplier. hopefully this will be fixed en masse as I submitted one appeal but there are 30 or so events and I'm too laxy to submit for each of them.

I have 2 problems to the new system.

1. Participation points based on size.  This fact chokes smaller stores horribly.  Players interested in qualifying are suddently leaving the smaller stores for bigger ones because larger size means more rounds for more points and more participation points.

2. The FNM modifier is disproportionate.  If you work friday evenings your chances are tiny.  You have PTQs and that's it.  This exascerbates problem 1 with the biggest FNM getting all of the business.  My stores in my area play nice and each take a different day to host a standard tournament.  The non-friday days enjoyed the 16k weight where the FNM had only 8k weight, but you get other incentives like foils.  The Friday store seems to be king now.

I've had a few days to ponder these changes since they were announced, and while I'm not happy about them I'm going to learn to live with them. I am one of the so-called "casual-competitive" players. I enjoy bigger events like PTQs, GPs, and Nationals and will travel to them when it is reasonable. I also have a full-time job and a relationship, so I have to choose the events I play in carefully to maximize my enjoyment of the game without disrupting these other important elements of my life.

The fundamental change that is occuring here is that you are replacing a rating system with an achievement system. While ELO was not perfect for Magic, it was a very clear indicator of how well someone performed over time. There may have been peaks and valleys, but everyone with a high rating EARNED it in some way by playing Magic well. Those with a high number of Planeswalker Points may also end up being good players, but the system rewards different things. Primarily, it can reward players who spend all their free time playing Magic, and all their expendable income on buying your products and travelling to your highest multiplier tournaments each week.

My greatest concern in all of this is that Wizards as a company is now promoting these potentially unhealthy social habits by dangling the biggest rewards they can offer in front of its players as an incentive. Aaron's MMO analogy is spot-on, and make no mistake, the high-level competitive Magic scene will undoubtedly become the type of grindfest that is the biggest negative association most people have with those types of games. Pro Tour Magic will no longer be about gathering together the best in the game, but about gathering together those for whom Magic is Life, with little room for anything else life has to offer.

I mostly agree with you, moonblaze.  However, even though the PP system rewards the players who go to high multiplier tournaments each week, it rewards those people that WIN those same tournaments even more.  So I don't think that your going to have what happens with the Baseball Hall of Fame happen to Pro Magic... the cream will (I think) still rise to the top and players that consistently win large tournaments on a regular basis will still outpace people who only rack up participation points.

Personally, I'm a casual player (FNM's and Pre-Releases) who has wanted to make the jump to playing in a few tournaments like GP's, but I didn't think that it would be worth the trip only to get trounced by experienced tournament players.  But the idea that I do get some reward for playing in a big event like a GP even if I don't change the world with my latest rougue combo deck does make a difference to me.  Mostly because of this announcement, I plan on going to the next GP in my area, which is in a couple of months.

As for the issue of players who have spent all their money on putting together top tier (expensive) decks, Magic has been like that as long as I've been playing. Even when I first broke in (Onslaught) I droped a fair hunk of change on my awesome zombies and clerics.  My final suspicion, and this is the other reason that the PP system has me going back to the tournament scene is that the PP system provides an easy way to create a replacement for the old MTG Player Rewards program. But nothing in this announcement mentions the rewards program, so for now that's just a feeling that I have.


So I may be in the minority, but for me at least, the PP system has done what it set out to do.

I'm going to say something that'll probably make you skim this post and not get the point, so I'm just asking you to read the second sentence, too, before dismissing me out of hand.

"Incentivizing gameplay" is bad, okay?  It's one thing to take away reasons not to play the game, and of course it's great if the game is fun and people want to play to enjoy themselves.  But giving people a reason to play the game even if it isn't fun is going to hurt the game in the long run.

You actually SAID you stole this from MMOs; have you actually played WoW (which is what I assume was the specific game that meant)?  Grinding is NOT FUN.  It's a CHORE.  WoW's little system of turning everything into positive reinforcement took away the ability to enjoy the activities involved.  Leveling is a chore you do in order to be able to play in dungeons - except once you get to that point, playing in dungeons is a chore you have to do in order to be able to play in raids.  And then raids turn into a big grind in order to prepare for the next tier of content.  Everything is positive reinforcement all time, and NONE of it is fun as a direct result.  The only place in WoW that DOESN'T give you all positive reinforcement all the time is PvP, and hey, guess what?  PvP is what keeps long-term players in the game.

Constant positive reinforcement will not keep players interested in playing on a long-term basis.  In fact, the opposite is true: it will hurt people's ability to stay interested, because constant positive reinforcement is boring.  This is a demonstrated fact that you can see just by glancing over the demographics of what activities long-term players engage in, in WoW.

I feel that your changes are very short sighted, and while you mean well, it will start to alienate many players who cannot grind every weekend. The biggest allure for travelling to any GP is that top16 invite. To hang out with a few friends here and there and test and practice your matchups in order to get that elusive PT invite, because lets face it, the amount of variance in PTQing as you guys have pointed out can be very punishing. Top16ing a GP was a perfect balance between skill and the variance of playing magic.

While on this note, what happens to top50 at a PT, do these players also get an invite to the next PT? What about pro player club levels, are these now gone?

I'm not 100% sure I agree grinding is not fun, having been a minesweeper / tetris / bubble blast addict for many a year. I do agree this will not be good for players that have less time than others (which I guess is the principle flaw in other such games like WoW anyway).

On the other hand, as a casual player I find it kind of cool that I now am a certain level (I won't boast which as it won't impress anyone); my main beef before I looked at this thread was that I want these ratings to show up on the forum pages - would it be so hard to link a DCI number to each forum account? Could help us all to decide which posts to take seriously  

Sep 8, 2011 -- 11:30PM, Akym wrote:

my main beef before I looked at this thread was that I want these ratings to show up on the forum pages - would it be so hard to link a DCI number to each forum account? Could help us all to decide which posts to take seriously  

I'm a casual player. I play only at a kitchen table. DCI ranked players are a margin of all Magic players. I don't think that ranked players posts are more valid than kitchen table players. I may play Magic every day at my home and spend more money for my collection than FNM participan.

We have been paying a lot of attention to what people in the gaming world have been doing in recent years, and across the board—whether it is MMOs or social games—the way you keep people invested is through a constant stream of positive reinforcement.

Now I've heard everything.  Magic is now trying to be more like Farmville.

Would be nice to see BDM actually present some questions about potential flaws in this system.  I understand he is writing for Wizards but seeing him throw all these softball questions about the format while ignoring some of the elephants in the room is a pretty big bummer for what I consider to be one of the better writers/commentators in Magic.  I also noticed they threw out a couple Pros names in support of the system, but failed to mention that others such as Finkel and PVDR have been less than enthused with it.  For me the new system has just made competitive magic even less legitimate as it now promotes grinding over playing against other skilled opponents. 

As for the whole issues of people sitting on ratings, what was preventing them for creating a "competitive season" with ELO?  I'm not totally familiar with the equations used in ELO but it seems like they could have had a lifetime ELO and a competitive ELO that reset every 4 months, ensuring that people would have a reason to go out and play.

I like the system overall.

There's one major problem with it though: The FNM x3 multiplier. The three times bonus severely damages many magic communities and is also very unfair for those that can't play every friday due to work, long distances to FNM or the Shabbat.

I personally have been playing since revised, and attend my local tournament on tuesdays. We have about 50 people every tuesday and a thriving community. We couldn't gather nearly as many if we're forced to move this tournament to friday (due to the x3 multiplier, wizards virtually forces us).

If I play my tuesday night tournament and go 6-0 and some random dude goes to FNM and go 3-3 he will earn more points than me. That means I have to compete at FNM's to even have a remote chance to get national invite and GP byes.

Further more it diminishes the importance of the GPT's and PTQ's that have almost the same importance as FNM's now. I seriously don't understand why you wouldn't make these tournament more important and special by having no multiplier on FNM's or a bigger multiplier on them.

I'm not one of those players that say I'll quit magic forever because of this change. I do however think that I'll play far less competetive events, because I know have lost my incentive to raise my rating, and eventually get invites to nationals, byes to GP's and such. I don't play FNM's because I can't play on fridays, so I can't get enough point to compete with those that do. So now it's virtually impossible for me to get enough points even though I do fairly well in all the other tournament I participate in...

The weirdest thing about all this is that I think the system would be fixed, and you would make people alot happier, if you did a single very simple thing: Remove the FNM multiplier. Aren't the FNM championship enough reward to those that attend FNM's???

I despise how MMOs work. But Aaron reassured on twitter if grinders suddenly outpoint skilled players, the system will be ajusted.

Sep 8, 2011 -- 10:18PM, Tap4Mana wrote:

I mostly agree with you, moonblaze.  However, even though the PP system rewards the players who go to high multiplier tournaments each week, it rewards those people that WIN those same tournaments even more.  So I don't think that your going to have what happens with the Baseball Hall of Fame happen to Pro Magic... the cream will (I think) still rise to the top and players that consistently win large tournaments on a regular basis will still outpace people who only rack up participation points.

We'll have to see. As said before, someone who consistently wins european GPs can still be beaten by someone who just goes to every american one.

I am mostly a casual player and don't really care about any ratings but posts like Bondafong's made me wonder: Is there anything keeping locale stores from having their FNM on another day of the week? I mean other than the F in the name FNM. 

BDM: When you look at this program a year from now, what will be the hallmark of this being a success? If it goes smoothly without any hiccups?

"I think that we don't need to prove that Magic is a skill game anymore. We have fifteen years of data that proves that."
- Aaron Forsythe arguing against ELO

While there may be some validity to this point I tend to disagree since removing the ELO system completely and replacing it with the Planeswalker Points still undoubtedly sends the signal that Magic is in fact a game where the amount onf time you spend playing it gets rewarded rather than your actual skill level. I think it is critically important to the game that it keeps some semblance of an actual skill based game at least in its rating system. One of the great things about Magic after all is that it is fun to play at any level but DOES lend itself to actual competitive play very well.

I am firmly of the opinion however that the best players should reap the rewards of their playskill and receive byes or invites to the Pro Tour. This is a metric that the new system utterly fails at since only those who play frequently in high level events have a legitimate shot at the benefits.

What I find most hypocritical is that the Planeswalker Points will be used as a ranking metric when it is completely apparent that the thing that is being measured by them is not in fact playskill but rather the sheer volume of tournaments played. I find this very troubling as it telegraphs a completely wrong idea of what competitive play is about. The more impressive feat should always be going 15-0 in tournaments rather than playing 5-10 in three times as many tournaments. Under the new system these two "achievements" are met with the same reward. This is very obviously wrong and demonstrates that the PPs are in no way a justifiable measure for skill when they HAVE TO be if used as a ranking system. Number 1 in a ranking should always be the best because of winning most in relation to his/her losses.

Something else that is very bothersome is that the system is utterly US-centric. There are most tournaments available in the US and thus it is likely that most of the benefits will be handed out to US players on the simple grounds that they have more chances to play in more tournaments. Europe has greatly skilled players as has been made obvious on the PT circuit since basically forever, and distributing byes and PT invites via the worldwide rankings puts these players at a disadvantage since they don't even have the chance to rack up the same number of points as players in the US with the SCG open series and similar events. I don't see what the problem is with regionalizing the byes in a similar fashion as has been hinted at with the PT invites. It just seems way more fair to distribute the byes and incidentally more of the PT invites by means of the more regional (that is continental) rankings since players in those regions have more of the same preconditions and can thus be measured more fairly by the same scale.

I am all for the idea of making losing less of an issue but this can certainly not be the way to go about it if one wants to represent Magic as any kind of competitive and skill testing game.

EDIT: Something else that just came to my mind is that more casual players usually don't show much, if any, interest in their ratings. Why "mess it up" for those people who actually care about/benefit from/are interested in the portrayal of Magic as a skill-based game via rating only to make it for those people who likely won't care one way or the other. The tournament scene did not have much of a grudge against the "old" rating system as far as I can tell and it did reflect much better what mattered to those people: How good am I at Magic in comparison to my peers.

OK - I... Agree with the principles behind the system (Taking into account ). Positive reinforcement (and gameification. The weirdest part of this system is that it's gameification applied to a league structure, essentially) for playing, not allowing it to be gamed by not playing (Yes, this system is gameable, but... Being gameable by playing lots of events is probably better than being gameable by not playing), etc. All good motivations for replacing the old system.

...I disagree with the implementation, however. Partially because I'm a carless Brit who's about one and a half hours away by public transport from my nearest FNM (...And about two hours away from a place that would be ideal for grinding this system despite it being double the distance. Public transport is weird that way), partially because this will hurt smaller stores, especially in places where there are multiple stores hosting sanctioned events but... Mostly because as currently implemented this isn't a rating or ranking system. And I'm desperately trying to work out a way of rewarding play while being a ranking system without concluding 'have two systems running in parrallel'

Now, I'm no serious competitive player (only Prereleases), but I find the new system amusing and will probably check on it, whereas I never did with the ELO system. However, I do think some important points were raised. Most importantly, going 5-10 in three events is better than going 15-0 in one event. This is obviously a huge problem. So why not add another multiplier that takes into account the percentage of matches won? A person who went 5-10 won a third of his matches, so he gets a third of the points. A person who goes 15-0 won all of his matches, so he gets the full points.

The incentives remain the same, but the possibility and advantages of grinding are greatly reduced (one would have to go 5-10 in nine events to make up for one 15-0, which I think is pretty fair).

Sep 9, 2011 -- 1:31AM, GizenshaFox wrote:

I disagree with the implementation, however. Partially because I'm a carless Brit who's about one and a half hours away by public transport from my nearest FNM

You dont live in Lincolnshire as well do you?

Sep 9, 2011 -- 1:31AM, GizenshaFox wrote:

And I'm desperately trying to work out a way of rewarding play while being a ranking system without concluding 'have two systems running in parrallel'

We had that already, and they killed it.

We had a DCI Rating (that at least attempted) to reflect player skill.
We had Player Rewards that rewarded dedication and attendance, giving positive reinforcement regardless of win/loss records.

Players who turned up every week and sucked got Player Rewards points anyway, and got free cards in the post.

Players who turned up infrequently but did well got DCI points and a higher ranking.

Players who did both got both.

---

My FLGS runs a casual Hobby League every weekend (Fri 1PM - Sunday 4PM). Each weekend you can play up to 2 games against each other player. You get 1 point for a loss, and 2 points for a win. Someone who plays a lot does better than someone who plays less often, and a "loser" can beat a "winner". The League is a great way to break the ice with a new player, and it encourages people to play others outside their group of friends. Any game can count for the League, even if it is part of a tournament (as long as both players agree), and it doesnt cost anything to enter the League.

At the end of the month the store owner totals up everyone's points. The player with the most points gets a free booster pack of their choice. On top of that, the owner uses the League rankings as a way of deciding who gets the WPN promos for the month (for example, if he gets 30 promos to hand out, the top 30 players get them). This incentivises people to turn up and play games, and rewards people even if they cannot win the whole League (and even if they cannot play every weekend, since some months there are only 30 entrants in the first place, so everyone gets the promo as long as they play).

At the store, the serious players know that the League is just a bit of fun - as long as they get their promo, they are happy regardless of where they finish. Besides tongue-in-cheek bragging rights, the rankings mean nothing, since they dont reflect how good a player you are - just how often you can make it to the store. Does that remind anyone of anything?

~ Tim

This system does not work for me or someone like me. I'm married, have children, have a busy career and like to live a well rounded life of which MTG is an important part. I play in about two tournaments a month and judge about one. For someone like me this represents a significant amount of time dedicated to a hobby.
I am reasonably good at Magic and under the ELO system was ranked well within the top 100 in my state. Under the PP system I don't crack the top 600. The best players in my state under the ELO system now barely crack the top 100, while players who are everpresent speedbumps at local events litter the top 25. 
PP reward BAD players with NO LIVES. Invitations have been cheapened becasue they may now be bought as well as earned, and there are actually more opportunities to buy them than earn them. ELO may not have been perfect, but at least it rewarded skill and talent more than raw volume: PP do the opposite. I'll probably play less, and my excitement about the expanded Grand Prix's has vanished.

I don't even think their metric for success is the correct one, all Rainbows and Buttercups spin as it is.  They think the first-time tournament attendee who 0-5s was fine until he got home, looked up his rating, and then gave up on tournaments after seeing his 1581 rating?  But now he's going to go "I have 6 Meaningless Points!  Can't wait to lose again!"  I can't even fathom that level of thinking.

Here's what a success for this plan would actually look like.  A decent player who's never tried anything outside his home store one day discovers he has a bye at a nearby Grand Prix, and takes that as an encouraging sign to step up to the next level.  That would be a success for the new system.  But again this system doesn't reward skill so much as grinding, so he'll be behind the guys who drive to PTQs on weekends instead of playing locally even if he beats them when they meet.

In truth I have to agree that busy players who lose a lot won't be near the top, even if they are above infrequent-play-but-always-wins.  For 65 world invites the competition will be between "Oh my God how can anybody play that much" and "Oh my God how can anybody play that much who also wins slightly more often".

Sep 8, 2011 -- 11:30PM, Akym wrote:

On the other hand, as a casual player I find it kind of cool that I now am a certain level (I won't boast which as it won't impress anyone); my main beef before I looked at this thread was that I want these ratings to show up on the forum pages - would it be so hard to link a DCI number to each forum account? Could help us all to decide which posts to take seriously  

I have a DCI rating of 1650, and i have no idea if that is good or bad. But i know that it won't make me a top player nor a top-poster. 

My post count however means that people around here know of me, it indicates that i know something about the game.

 

I was thinking... if the pro only get's value playing more it would probably means they will play more and will probably win... so: if now they don't play and others players have space for winning... if with the PWP the tops become active... some places would end with just the same top 8 forever... pushing new and less skilled players away...

BDM: This sounds a lot like making the Top 8 of a multiple Pro  Tour Qualifiers could have the same result as actually winning one.

Definitely. This system rewards you the more you play. If you are  one of those players who makes the Top 8 a lot—if you always get  close—it is going to add up. Before it didn't add up; it was just yes or  no. Those are the people we want to reward.



Riiiiiiiight. So let's do the math:

The cutoff for the Top 100 on Planeswalker Points was 1701 in the last (unofficial) season between April and August. So let's say you're a very skilled PTQ player who is really serious about qualifying and travels to five PTQ's (which is a lot, especially now in Europe, where more than 50 percent of the PTQ's were taken away). Let's say he is so good that he Top 8's all of them and gets kicked out in the semi-finals (on average), so your record is like 5-1-1 and then 1-1 in the Top 8, so 6-2-1. Assuming around 100 players participate in each of these PTQ's, the total points received this way for one PTQ would amount to:

(6x3 + 1x1 + 4 participation points) x 5 = 115 points for Top8-ing one PTQ. So five PTQ's amount to 575 points. Even if you take into account that the 1701 points cutoff is above average since there were nationals in this season and some people in the top 100 will already have received their invitation in one or the other way, THIS IS STILL NOWHERE CLOSE TO THE THRESHOLD YOU'D HAVE TO REACH. Even if you participate in two GP's and do well in them, plus you participate weekly in your local FNM, you'd end up at 1.200 points or so AT BEST.

I rarely post in these forums, but this lying is really outrageous. Please stop bullsh***ing us.

I think we all know this isn't going to work.

They should have added this as an alternative ranking system, filling half the slots using the old method and half using the new method.

Some players with no life but magic will get in to tourneys, even if they are mediocre and most great players who have professional and personal lives that limit repetition will never see another invite. As others have pointed out, the players who tend to have the most money to spend on the game are also likely to be the ones who don't have time to play magic non-stop.

I think they have to stick with it for a year, but then I predict they will have to do something different and this will just be added to the list of embarrassing moments for the MTG developers.

I'm overall fine with this, though I not an aspiring pro, and until 2011 had been in few sanctioned events in 15 years. I do have two suggestions for refinement:

(1) Set the multiplier across all events equal to match wins minus match losses, with a minimum multiplier of x1 to preserve the "always beneficial to play" aspect. This would let a great achievement like a GP Top 8 be very hard to duplicate, points-wise, with several mediocre finishes---as it should be.  

(2) Leverage this data to provide more public metrics of player skill for those decrying the loss of Elo as a skill rating. Average GP/PT finish, lifetime and season-by-season number of PTQ T8s / GP/PT T16/32/64 (with # of events played in each case), that kind of thing. Give Spikes an opportunity to see how they measure up against the world or their region.

And, fellow players, let's not forget that this brings with it the ability to run twice as many Grand Prix, impossible in the old T16 invite system. Major upside. I might go to my first GP in 2012 because of the increased availability. 

Sep 9, 2011 -- 3:56AM, DrSylvan wrote:

(1) Set the multiplier across all events equal to match wins minus match losses, with a minimum multiplier of x1 to preserve the "always beneficial to play" aspect. This would let a great achievement like a GP Top 8 be very hard to duplicate, points-wise, with several mediocre finishes---as it should be.  

This sounds awesome. Please think about it, WotC.

The "grinding" problem could be at least partially avoided by changing the Competitive rating to some kind of weighted average of your best N events in the relevant period of time.  This way, players would never be penalized for losing, but would not be rewarded for playing large numbers of events and doing badly.  There would be some incentive to play in as many events as possible in order to try to get better results, but it would drop off rapidly.  This system would also give more of an advantage to players who do well at a few events over those who do badly at many events.

Actually, despite the fact that I still don't like the change to the rating system, even if I understand such things can be needed, one positive thing was noted. They're throwing in the plane tickets for the ratings spots, when before they didn't. So to get more, yes, people have to put in more - they have to play a lot. "Winning is even better" at least means they will still, to some extent, be bringing in the right people.

Sep 9, 2011 -- 1:49AM, Zindaras wrote:

why not add another multiplier that takes into account the percentage of matches won? A person who went 5-10 won a third of his matches, so he gets a third of the points. A person who goes 15-0 won all of his matches, so he gets the full points.

The incentives remain the same, but the possibility and advantages of grinding are greatly reduced (one would have to go 5-10 in nine events to make up for one 15-0, which I think is pretty fair).

This is possibly the best thing written on the subject of PWP so far. It doles out the "everyones a winner" incentives but greatly improves the rewards for successfull play and stops the system being gamed by those who have the means to attend events all day, every day.

The only 2 problems with this are a) it doesn't always give a nice round number of points (but magic has a rule about always rounding down unless told otherwise, doesn't it) and b) if a player drops do you use a multiplier based on his wins over games played or actual rounds in the tourney (this is easily solved by picking one and sticking to it; using the rounds in the tourney is even less gamable, though prolly have to treat day 2 of big events as seperate tourney to stop people not making the cut having their points shredded).

It's interesting that the new system works kind of like MTGO Leagues did (i.e. it really helps if you have lots of time to play), the article references MTGO as one factor in moving away from ELO,

...but MTGO record won't be captured in PWP, for some reason.

I could understand not including MTGO in Professional or Competitive points counts.  I don't understand why MTGO matches can't go directly into Lifetime points, though.  Lifetime basically doesn't matter for anything, and there are plenty of people like me who've played 1 or 2 paper prereleases, but years of MTGO.  Magic really is (or at least was) my life; it's strange that one way of playing doesn't count against Lifetime points.

I don't know exactly how I feel about this, but as someone who is and has always been one of the people in that little "casual-competitive" niche, where I work hard to win but I don't get much enjoyment out of just sleeving up a netdeck and am willing to accept losing a bit more to fight on what feel more like my terms...

"For newer players the experience was often like, "Welcome. You have  lost. You have spent all your money, and then when you go home and look  it up on the Internet we will prove to you that you are terrible. We  have quantified that you are worse than a player who has never played at  all."

I totally get this line. Any time I found myself in the middle of a losing jag and wondered if I ran the risk of dropping back below 1600 it really infuriated me. It felt like it was saying "Although you have invested ten years of your life in our game, and although much of that has been tournament play, we are perfectly fine with labeling you 'worse than a noob.'"

That was something I always disliked. On the one hand, yes, if I am on a losing streak, I don't have a problem with numbers reflecting that in some way. Yep, been losing. Yep, have to change up my decks, practice harder, fight to win if I wanna. I'm Spike enough to know that, and not to mind. But the idea that someone who'd never shown up before and bought a precon five minutes ago is "better than me" (ie. has the 1600 by default) because I've been losing, when she hasn't even been playing at all didn't seem right.

The person with the under-1600 is putting herself out there. She's competing. She's losing, but she's not sitting on her rear end. And yet, under the old system, not sitting on her rear end meant being at risk of looking worse than those who are.

That bothered me. I don't know what I think of the new system and I don't know if it's better. But my initial reaction is that I am glad that my many, many years of participation, of willingness to get knocked around and come back fighting and hopefully to improve as a player even if I'm not yet winning again count for something. Seeing that I wasn't at the bottom of the levels made me smile.

Because I have been here, and no, I'm not a pro -- but yes, I compete, and no, I'm not going away.

Wizards, I don't know if this system will turn out to be good or bad. I honestly don't. But, whatever happens to it, however it changes or doesn't or dies completely: thank you.

ETA: I had decided to take a break from FNMs until Innistrad's release -- I'm so sick of Zendikarian nonsense I could puke. But this makes me think maybe I don't mind Birdies and angry lands quite so much, since showing up at all will count for something.

Sep 9, 2011 -- 5:48AM, Coyltonian wrote:

Sep 9, 2011 -- 1:49AM, Zindaras wrote:

why not add another multiplier that takes into account the percentage of matches won? A person who went 5-10 won a third of his matches, so he gets a third of the points. A person who goes 15-0 won all of his matches, so he gets the full points.

The incentives remain the same, but the possibility and advantages of grinding are greatly reduced (one would have to go 5-10 in nine events to make up for one 15-0, which I think is pretty fair).

This is possibly the best thing written on the subject of PWP so far. It doles out the "everyones a winner" incentives but greatly improves the rewards for successfull play and stops the system being gamed by those who have the means to attend events all day, every day.

The only 2 problems with this are a) it doesn't always give a nice round number of points (but magic has a rule about always rounding down unless told otherwise, doesn't it) and b) if a player drops do you use a multiplier based on his wins over games played or actual rounds in the tourney (this is easily solved by picking one and sticking to it; using the rounds in the tourney is even less gamable, though prolly have to treat day 2 of big events as seperate tourney to stop people not making the cut having their points shredded).

a) I would personally just use the regular rounding system (0,5 and more is 1, less than 0,5 is 0) but this shouldn't be a problem overall. Given the large amount of points that will be doled out anyway, the rounding is likely to be insignificant (and, otherwise, they just average out to being slightly positive over the long run, so I don't think anybody would be unduly harmed).

b) I would say that day 2 of the big events are separate events. Otherwise, it depends on what Wizards want. I personally think dropping is silly, but I only play Prereleases and it's more about playing than winning for me, so I would say that the total number of rounds would count (this also makes it harder to game the system, as you said), but maybe the actual Pros would find that horrible and like their drops too much.

In all fairness, we should wait and see how the new system plays out, but I do think that a performance multiplier would make people feel a lot better about themselves in the long run. And because Magic at this high a level is a game played over so many rounds, we still get a nice sliding scale (the difference betwen winning a Pro Tour and coming in second is one match on 20 or something, which makes a difference of 0.05 in the multiplier, certainly nothing huge). Except the system leans toward being exponential rather than linear, which rewards performance and participation.

Sep 9, 2011 -- 6:13AM, AlexaM wrote:

I totally get this line. Any time I found myself in the middle of a losing jag and wondered if I ran the risk of dropping back below 1600 it really infuriated me. It felt like it was saying "Although you have invested ten years of your life in our game, and although much of that has been tournament play, we are perfectly fine with labeling you 'worse than a noob.'"

OK I get that.  A player needs a minimum number of matches for a realistic rating (someone could calculate the number, but I'll guesstimate 10 tournaments).  And the fact that someone with zero matches is assumed to be 1600 instead of simply being "NA" is could be offputting.  But that's a communication or tweakage issue, because as you say it's the long-term player looking at hypotheticals who feels it, not the new player who goes "woo-hoo I'm 1600, time to brag!"

I believe Chess, despite having a similar method, starts people at 100, which is also the floor.  So it was fixable without scrapping it.  Now there is no rating, there's no way to know how your skill measures after a good or bad run even if you want to.  You know if you've played a lot or a little, but then you shouldn't need a number to tell you that.

Since I appreciate your point of view AlexaM, let me ask you this.  I don't know what kind of tournaments you favor, but suppose you get the idea to step up to a Grand Prix.  And you have a good run against some really tough opponents, finishing the day 6-2-1 and just missing Day 2.  You feel pretty good because that's a good record.  Afterwards you look up your Planeswalker points, and they tell you that you got 204 points for the day.

Is that satisfying to know?  Since you've checked your rating in the past you obviously cared to see how you measure up, but now there's no way to know.  You can't tell if some of those opponents you beat were really accomplished.  You can't tell if you've done better than when you were really hot in 2009.  All it tells you is that you played and scored points, but you knew that anyway.

I think they've lost something big by tossing out the skill rating.  It's not bad to have a player reward, and people should feel like they've accomplished something by playing.  But throwing out skill ratings really sends a message to me that Magic isn't a serious game anymore.

Sep 9, 2011 -- 5:48AM, Coyltonian wrote:

Sep 9, 2011 -- 1:49AM, Zindaras wrote:

why not add another multiplier that takes into account the percentage of matches won? A person who went 5-10 won a third of his matches, so he gets a third of the points. A person who goes 15-0 won all of his matches, so he gets the full points.

The incentives remain the same, but the possibility and advantages of grinding are greatly reduced (one would have to go 5-10 in nine events to make up for one 15-0, which I think is pretty fair).

This is possibly the best thing written on the subject of PWP so far. It doles out the "everyones a winner" incentives but greatly improves the rewards for successfull play and stops the system being gamed by those who have the means to attend events all day, every day.

The only 2 problems with this are a) it doesn't always give a nice round number of points (but magic has a rule about always rounding down unless told otherwise, doesn't it)

Oh, the solution to that is simple. Multiply all numbers involved by 100 or 1000 before you start, which if pinball machines and old arcade games are anything to go by, is something players will love. "3000 points for a win? 1000 points just for playing? Sure, that'll be multiplied by my percentage (but prevented from going to zilch) but... Awesome." (I mean, ok, some arcade games actually use the insignificant bit to record information such as 'how many continues', but for the most part all those zeroes were just there to make players feel better about paying money into the machine)

Sep 9, 2011 -- 5:48AM, Coyltonian wrote:

and b) if a player drops do you use a multiplier based on his wins over games played or actual rounds in the tourney (this is easily solved by picking one and sticking to it; using the rounds in the tourney is even less gamable, though prolly have to treat day 2 of big events as seperate tourney to stop people not making the cut having their points shredded).

Even if I made day two of a large event, I think I'd prefer my day one results in such a system to be 'safe' by treating the two days as seperate tournaments for scoring (Actually using percentage as a point multiplier I'd probably be willing to call this a ranking, albeit a really weird one (Hey! We can even help the positive reinforcement they're going for if we multiply by the percentage treated as an integer (minimum 1) rather than multiplying by the percentage so you only get 'all' your points if you get no losses. Which prevents the need to worry about rounding (other than rounding the win percentage) - Go 15-0, score multiplied by 100. Go 5-10, score multiplied by 33. Mathematically that should be the same as 300 for a win and multiply down but it probably feels better to the players who go mid-table)

...And suddenly the scores start closer reflecting an old-school arcade shmup than a sleek, modern, RPG, but, as I said earlier, I don't mind the philosophy behind the change (make it so that people can't game the system by not playing [Gaming the system by playing lots is probably better], just the implementation.)

Alternate modification - What about making the points the prizes of all tournaments rather than rewards for single matches? Say, 8 points available per player in a tournament (this still has the problem of penalising people for supporting smaller retailers but it's difficult to figure out how to fix that while taking the philosophy that larger tournaments are (on paper) harder and as such should be worth more), 50% of the prize pool goes to the first place, halfing all the way down until you get to 1, afterwhich everyone gets a point for turning up? Callibrate until you're giving the 'right' amount of people 'real' points. (4n 2n n 0.5n 0.25n... In an eight man draft that would be giving 1 point to sixth and below, 2 for fifth, 4 for fourth, 8 for third, 16 for second and 32 for first. For finding good prize pool distributions for both wide and deep options, you can probably find some on poker websites.) Apply additional multipliers from there for the events WotC want to emphasise (i.e. FNM, GP, etc)

Sep 9, 2011 -- 7:29AM, GizenshaFox wrote:

Sep 9, 2011 -- 5:48AM, Coyltonian wrote:

Sep 9, 2011 -- 1:49AM, Zindaras wrote:

why not add another multiplier that takes into account the percentage of matches won? A person who went 5-10 won a third of his matches, so he gets a third of the points. A person who goes 15-0 won all of his matches, so he gets the full points.

The incentives remain the same, but the possibility and advantages of grinding are greatly reduced (one would have to go 5-10 in nine events to make up for one 15-0, which I think is pretty fair).

This is possibly the best thing written on the subject of PWP so far. It doles out the "everyones a winner" incentives but greatly improves the rewards for successfull play and stops the system being gamed by those who have the means to attend events all day, every day.

The only 2 problems with this are a) it doesn't always give a nice round number of points (but magic has a rule about always rounding down unless told otherwise, doesn't it)

Oh, the solution to that is simple. Multiply all numbers involved by 100 or 1000 before you start, which if pinball machines and old arcade games are anything to go by, is something players will love. "3000 points for a win? 1000 points just for playing? Sure, that'll be multiplied by my percentage (but prevented from going to zilch) but... Awesome." (I mean, ok, some arcade games actually use the insignificant bit to record information such as 'how many continues', but for the most part all those zeroes were just there to make players feel better about paying money into the machine)

Couldn't this be radically simplified:

• Win: 3 points
• Draw: 2 points
• Lose: 1 point

Thus, every match has a total (base) value of four points.  Anyone who just shows up still accumulates points, but someone who is actually winning accumulates points roughly twice as fast.

Sep 9, 2011 -- 6:59AM, Amarsir wrote:

Sep 9, 2011 -- 6:13AM, AlexaM wrote:

I totally get this line. Any time I found myself in the middle of a losing jag and wondered if I ran the risk of dropping back below 1600 it really infuriated me. It felt like it was saying "Although you have invested ten years of your life in our game, and although much of that has been tournament play, we are perfectly fine with labeling you 'worse than a noob.'"

OK I get that.  A player needs a minimum number of matches for a realistic rating (someone could calculate the number, but I'll guesstimate 10 tournaments).  And the fact that someone with zero matches is assumed to be 1600 instead of simply being "NA" is could be offputting.  But that's a communication or tweakage issue, because as you say it's the long-term player looking at hypotheticals who feels it, not the new player who goes "woo-hoo I'm 1600, time to brag!"

I believe Chess, despite having a similar method, starts people at 100, which is also the floor.  So it was fixable without scrapping it.  Now there is no rating, there's no way to know how your skill measures after a good or bad run even if you want to.  You know if you've played a lot or a little, but then you shouldn't need a number to tell you that.

Since I appreciate your point of view AlexaM, let me ask you this.  I don't know what kind of tournaments you favor, but suppose you get the idea to step up to a Grand Prix.  And you have a good run against some really tough opponents, finishing the day 6-2-1 and just missing Day 2.  You feel pretty good because that's a good record.  Afterwards you look up your Planeswalker points, and they tell you that you got 204 points for the day.

Is that satisfying to know?  Since you've checked your rating in the past you obviously cared to see how you measure up, but now there's no way to know.  You can't tell if some of those opponents you beat were really accomplished.  You can't tell if you've done better than when you were really hot in 2009.  All it tells you is that you played and scored points, but you knew that anyway.

I think they've lost something big by tossing out the skill rating.  It's not bad to have a player reward, and people should feel like they've accomplished something by playing.  But throwing out skill ratings really sends a message to me that Magic isn't a serious game anymore.

Again, I'm not trying to say I think the system is without problems. But I'd say that I would like knowing that, because looking at the other statistics on my page, 204 is relatively a lot of points.

I do take your point though that it's no longer weighted based on how good the people I played are, and I see why that's not great. But I'm not a stats person so I couldn't tell you how to get that back in there without either going back to ELO and resurrecting the other problem, or add it in somehow and still keep the new system sense-making.

OT: I see in your sig that you have pauper tournaments on MTGO... can you message me with info? I might want to play in them.

This is the result from having the person with the PhD in Math (Richard Garfield) off working on set design rather than attending the meetings which took place where this abomination of an idea was conceived and developed.  No one with a reasonable understanding of statistics would have allowed this concept to have gotten past the "spitball" stage.

I understand statistics, I respect statistics, and have no tolerance for those who attempt to utilize them without the sufficient knowledge to use them with the care and consideration they require.  This model is critically flawed and if this decision isn't reversed before full implementation, be assured that many players, myself included, will be inclined to utilize those flaws.  Personally, I won't bother doing it for my benefit but purely to show you its existence so you will realize that this system is invalid for the purpose you have assigned it to perform.  That said, it'll quite possibly be to the benefit of myself and/or other players at the tournaments at which I participate.

Now, in order to balance out the tone of this message, here's some constructive advice:

The ELO system itself was fine.  Personally, I think it's the most effective method for accurately representing a player's ability and success.  It was the method of implementation that required adjustment because of how it was being used to award invitations and Grand Prix byes .  Two simple things* to correct it:

1)  Instead of using a snapshot of a particular day on which to determine GP byes and invites, use the period leading up to that cutoff date (90-day, 120-day, 13-week, or 17-week, whichever works best), using the highest DCI Rating each player achieved during that period.

2)  Require a minimum amount of activity during that period to receive the byes/invites.  Total up the K-values under which each sanctioned match was played and the player must achieve a minimum value - 400 or 600 seem reasonable - to receive the associated reward.

The issue of players "sitting on rating" would be solved.  No more reason to stop playing just because they've raised their rating above a specific threshold and they'd lose the benefits of that rating unless they play enough.

* - I posted this suggestion elsewhere and someone pointed out that a group of players could falsify a series of tournament results, artificially inflating each of their DCI Ratings in turn over a qualifying period.  The software used to detect suspicious and fraudulent tournament results may need to be programmed to scan for such anomalies to prevent such abuses.

I have posted this when I learned about the change, and others have posted it in this thread before me: I am a bit sad that "external rewards" / "achievements" have seeped into yet another place. This sort of thing can be considered exploiting a psyschological flaw that many people have in that they get "addicted" to this (see MMOs and many so-called "social games" for clear examples).

That said, I admit the older system had flaws and this is a solid attempt at addressing it. There are rather serious issues with the new system as well (it really makes regional disparities a more severe issue for one). Those have been pointed out by others that know more about it, so I don't need to repeat it - it is a bit disapointing that the article simply did not discuss to address the more serious issues.

Ivo 

Dear Wizards,

Please at least leave the ELO rating somewhere accesible to players.  I find it more informative than the new system and would like to be able to review it.

Sep 9, 2011 -- 6:59AM, Amarsir wrote:

Since I appreciate your point of view AlexaM, let me ask you this.  I don't know what kind of tournaments you favor, but suppose you get the idea to step up to a Grand Prix.  And you have a good run against some really tough opponents, finishing the day 6-2-1 and just missing Day 2.  You feel pretty good because that's a good record.  Afterwards you look up your Planeswalker points, and they tell you that you got 204 points for the day.

Is that satisfying to know?

Easily. Easily is that satisfiable.

I have a large FNM store. I tend to play a lot of limited at that store. The last time I went 3-0 at that store, I scored 36 points (Field of 56 players). 204 points is 5 times that size total. That is a huge amount. That's added a whole Month to my score for just one event. If I was playing at smaller shops, that would even be MORE satisfying.

I posted in the other topic, but I figured I'd chime in here too to balance out the naysayers.  I really like this new system for the reasons Aaron listed.  I've been around for a while now.  I like that my points reflect that while I'm not an active "paper-Magic" player now, I have been for years.  Respect the licids, lol.  

Now, maybe I like this system because I've never been a player with pro-tour aspirations.  I know intuitively that should require a lot of work.  If you have a full-time job and family, maybe you shouldn't be able to easily ride that 1800 you've been caressing like a precious jewel for years to any major tournament you want.  I don't know how PTQs work now, but it used to be that the overall winner of a PTQ would get an automatic invitation to that season's PT.  I like this system.  It requires a LOT of skill to reliably win a large tournament, and it's only one weekend for those who have other things going on in their lives. 

As a player who likes to attend a random event here and there, my metric for success has always been how well I did at that tournament, not how well my overall rating is.  If am a good player that day, I am rewarded with packs that could potentially earn me a small profit.  But you have good days and you have bad days.  As Aaron said, I don't want one bad day to ruin my rating for the next six months. 

Under the old system, if you want to grind up your rating, you have to netdeck what's winning.  As a budget-minded player, I don't want my lack of a playset of Jace the Wallet Sculptor to discourage me from playing until they rotate (or are banned). 

I'm looking forward to this new system because it will make the tournaments I do play in more fun.  It will encourage people to play and not stress about some meta-rating that determines their fate. 

BTW, I think it would be nice if you would send out DCI cards that had your level on them!  :D 

Sep 9, 2011 -- 7:54AM, DedlyEdly70 wrote:

The ELO system itself was fine.  Personally, I think it's the most effective method for accurately representing a player's ability and success.  It was the method of implementation that required adjustment because of how it was being used to award invitations and Grand Prix byes .  Two simple things* to correct it:

1)  Instead of using a snapshot of a particular day on which to determine GP byes and invites, use the period leading up to that cutoff date (90-day, 120-day, 13-week, or 17-week, whichever works best), using the highest DCI Rating each player achieved during that period.

2)  Require a minimum amount of activity during that period to receive the byes/invites.  Total up the K-values under which each sanctioned match was played and the player must achieve a minimum value - 400 or 600 seem reasonable - to receive the associated reward.

The issue of players "sitting on rating" would be solved.  No more reason to stop playing just because they've raised their rating above a specific threshold and they'd lose the benefits of that rating unless they play enough.

Wow, DedlyEdly, well done! This is exactly what Wizards should have done. To employ some economics to go along with your excellent understanding of statistics, what Wizards had was an "incentives problem" which rewarded a very small segment of players (only those eligible or close to eligible for a Pro-Tour and maybe a few close to byes at a Grand Prix) for not playing. What they have created is another "incentives problem" which completley marginalizes players who are competitive but can't, for whatever reason (access to events, work and family responsibilities, etc.) play really frequently.

I understand why Wizards did this, to "encourage" players who are most dedicated to the game. The announcement of Planeswalker Points combined with this weekend's "Magic Celebration" is telling, they may as well hand out a free sample of crack and say "Now, the more crack you use, the more you can earn."

I worry about how this effects the legitimacy of the game and I worry about the moral and ethical considerations behind the decision.

Sep 9, 2011 -- 9:49AM, Paralistalon wrote:

I posted in the other topic, but I figured I'd chime in here too to balance out the naysayers. 

I dont think one poster could realistically hope to achieve that.

Sep 9, 2011 -- 9:49AM, Paralistalon wrote:

I'm looking forward to this new system because it will make the tournaments I do play in more fun.  It will encourage people to play and not stress about some meta-rating that determines their fate. 

So its good because making ratings meaningless stops players worrying about them?

Why not scrap the rating system all together and just hand invites out randomly, like a raffle? No-one has to worry about their results then!

No, the players who didnt care about their rating are unaffected - they still dont have to care. But the players who did care about them now have to play a postcode lottery to see how well they will do. Great.

~ Tim

Sep 9, 2011 -- 10:15AM, LMTRK wrote:

Sep 9, 2011 -- 9:49AM, Paralistalon wrote:

I posted in the other topic, but I figured I'd chime in here too to balance out the naysayers. 

I dont think one poster could realistically hope to achieve that.

Technically, there were two of us.

I'm not convinced that you can afford to try a random deck rather than a top-level deck - under ELO, the points threshhold for an invite was more or less fixed - if you drop below the threshhold, you have to play more matches to recover. Under the new points system, every event the leading players participates in raises the target by some amount - so, rather than a bad result meaning you have to play in a dozen more events to catch up, a bad result effectively knocks you out of contention entirely - you're in a red queen's race where you need to keep running as fast as you can just to stay where you are relative to everyone else...

Sep 9, 2011 -- 7:54AM, DedlyEdly70 wrote:

This is the result from having the person with the PhD in Math (Richard Garfield) off working on set design rather than attending the meetings which took place where this abomination of an idea was conceived and developed.  No one with a reasonable understanding of statistics would have allowed this concept to have gotten past the "spitball" stage.

I understand statistics, I respect statistics, and have no tolerance for those who attempt to utilize them without the sufficient knowledge to use them with the care and consideration they require.  This model is critically flawed and if this decision isn't reversed before full implementation, be assured that many players, myself included, will be inclined to utilize those flaws.  Personally, I won't bother doing it for my benefit but purely to show you its existence so you will realize that this system is invalid for the purpose you have assigned it to perform.  That said, it'll quite possibly be to the benefit of myself and/or other players at the tournaments at which I participate.

Now, in order to balance out the tone of this message, here's some constructive advice:

The ELO system itself was fine.  Personally, I think it's the most effective method for accurately representing a player's ability and success.  It was the method of implementation that required adjustment because of how it was being used to award invitations and Grand Prix byes .  Two simple things* to correct it:

1)  Instead of using a snapshot of a particular day on which to determine GP byes and invites, use the period leading up to that cutoff date (90-day, 120-day, 13-week, or 17-week, whichever works best), using the highest DCI Rating each player achieved during that period.

2)  Require a minimum amount of activity during that period to receive the byes/invites.  Total up the K-values under which each sanctioned match was played and the player must achieve a minimum value - 400 or 600 seem reasonable - to receive the associated reward.

The issue of players "sitting on rating" would be solved.  No more reason to stop playing just because they've raised their rating above a specific threshold and they'd lose the benefits of that rating unless they play enough.

* - I posted this suggestion elsewhere and someone pointed out that a group of players could falsify a series of tournament results, artificially inflating each of their DCI Ratings in turn over a qualifying period.  The software used to detect suspicious and fraudulent tournament results may need to be programmed to scan for such anomalies to prevent such abuses.

There is merit in the new points scheme, but that comes with a huge caveat.

First, this guy seems to know a thing or two. I understand that the people in charge of this decision are probably game designers, but [long string string of profanities and explicit description of your chosen diety's joyride on a circus vehicle] the new system is not fit for the purpose you are trying to put it to.

The problem wih players sitting on their ratings is not a reason to scrap Elo. The problem of players' incentives to play a game that they love is not a reason to scrap Elo. There is no reason to scrap Elo. You're throwing the baby out with the bathwater.

No, you're burning down the house with the baby in it because of some dirty bathwater.

See, Elo isn't about incentivising play. It is about accurately determining the skill level of a player. People care about that immensely. Planeswalker Points do not do this. At all. They don't even pretend to. Well, maybe you're pretending that they do. They don't.

Planeswalker points can help determine your invitations, and your byes, and what-have-you, but don't kill Elo ratings for Magic. That's...they're not even the same thing!

Wizards: "We've secretly replaced the top-tier Magic Players' skill rankings with a system to incentivise casual players to attend FNMs. Let's see if anyone notices!"

Sep 9, 2011 -- 7:29AM, GizenshaFox wrote:

Alternate modification - What about making the points the prizes of all tournaments rather than rewards for single matches? Say, 8 points available per player in a tournament (this still has the problem of penalising people for supporting smaller retailers but it's difficult to figure out how to fix that while taking the philosophy that larger tournaments are (on paper) harder and as such should be worth more), 50% of the prize pool goes to the first place, halfing all the way down until you get to 1, afterwhich everyone gets a point for turning up? Callibrate until you're giving the 'right' amount of people 'real' points. (4n 2n n 0.5n 0.25n... In an eight man draft that would be giving 1 point to sixth and below, 2 for fifth, 4 for fourth, 8 for third, 16 for second and 32 for first. For finding good prize pool distributions for both wide and deep options, you can probably find some on poker websites.) Apply additional multipliers from there for the events WotC want to emphasise (i.e. FNM, GP, etc)

Hey

Hey guys

Guys listen

Guys

Hey, I have an idea

Guys

What if

Listen

I have an idea

What if we based your rating on how skilled your opponents were in each match you played--and how well you did against them?

Sep 9, 2011 -- 10:47AM, branewalker wrote:

Wizards: "We've secretly replaced the top-tier Magic Players' skill rankings with a system to incentivise casual players to attend FNMs. Let's see if anyone notices!"

Freaking hilarious, well said.

I understand the desire to incentivize people to play more magic, and I understand that gamification is an important way to do that, but this doesn't achieve that goal.

Suppose you are a good player with Pro Tour aspirations, but they're a long shot. You know eventually you'd want to make a run at getting on the Tour, but can't dedicate the time or energy to it now. Before you had an incentive to play sanctioned matches to improve your rating. Every point you gain now is one you don't have to gain later. Under the new system, no amount of skilled, steady play over the years will help you during the big push.

This is also not a rating system (so don't ever call it that!) and it's not going to get the best players on the Pro Tour. As Aaron said on twitter, he doesn't care about measuring skill for determining who gets to play on the Pro Tour. That tweet read like a punch to the stomach, but I guess increased Organized Play participation is more important than the integrity of the Pro Tour.

Also the Elo ratings could have been fixed very easily to solve the things you claim were the problems (good players sitting on their ratings):
1) Change the win probability curve, especially at the extremes and reduce K values of the biggest events
2) Use the highest Elo rating over a period of time, rather than just a snapshot for invites. (This actually encourages play!)
and punish inactivity with either 
3) (hard) Rating decay. This is trivial to implement in a Bayesian rating system, as time passes the uncertainty increases, use the bottom end of the confidence interval.
or
4) (easy) Requiring a certain amount of play to get byes or ratings invites. You could either sum up the k-values of the matches played (so you have to risk something to keep your rating) or use some new metric that you come up with that measures playtime and incorporates wins crudely also. Maybe it could double as something that caters to lower rated players to incentivize them to feel good rather than bad about playing and losing. If you're going that route, I'd name these points something in-theme, like after Planeswalkers or something...

A final request: I know you're basically pot committed to PWPs at this point, but can you at least continue to calculate Elo ratings and leave up the website? I know I'm going to keep playing whenever I can and I'd like to see whether or not I'm getting better instead of the fact that I can't play as much as other people.

Sep 9, 2011 -- 11:00AM, branewalker wrote:

Sep 9, 2011 -- 7:29AM, GizenshaFox wrote:

Alternate modification - What about making the points the prizes of all tournaments rather than rewards for single matches? Say, 8 points available per player in a tournament (this still has the problem of penalising people for supporting smaller retailers but it's difficult to figure out how to fix that while taking the philosophy that larger tournaments are (on paper) harder and as such should be worth more), 50% of the prize pool goes to the first place, halfing all the way down until you get to 1, afterwhich everyone gets a point for turning up? Callibrate until you're giving the 'right' amount of people 'real' points. (4n 2n n 0.5n 0.25n... In an eight man draft that would be giving 1 point to sixth and below, 2 for fifth, 4 for fourth, 8 for third, 16 for second and 32 for first. For finding good prize pool distributions for both wide and deep options, you can probably find some on poker websites.) Apply additional multipliers from there for the events WotC want to emphasise (i.e. FNM, GP, etc)

Hey

Hey guys

Guys listen

Guys

Hey, I have an idea

Guys

What if

Listen

I have an idea

What if we based your rating on how skilled your opponents were in each match you played--and how well you did against them?

Sure, now come up with a system of that which doesn't incentivise people to sit on their ratings.

The ELO system is good at rating relative skill of players... However, when people are gaming the system to rate the game by not playing the game, which apparently was happening, then there's a problem. (Though I entirely agree that that shouldn't cause replacing the rating system with a level meter)

Sep 9, 2011 -- 11:56AM, GizenshaFox wrote:

Sure, now come up with a system of that which doesn't incentivise people to sit on their ratings.

The ELO system is good at rating relative skill of players... However, when people are gaming the system to rate the game by not playing the game, which apparently was happening, then there's a problem. (Though I entirely agree that that shouldn't cause replacing the rating system with a level meter)

Let me introduce you to this fellow, who wrote on the previous page:

Sep 9, 2011 -- 7:54AM, DedlyEdly70 wrote:

This is the result from having  the person with the PhD in Math (Richard Garfield) off working on set  design rather than attending the meetings which took place where this  abomination of an idea was conceived and developed.  No one with a  reasonable understanding of statistics would have allowed this concept  to have gotten past the "spitball" stage.

I understand  statistics, I respect statistics, and have no tolerance for those who  attempt to utilize them without the sufficient knowledge to use them  with the care and consideration they require.  This model is critically  flawed and if this decision isn't reversed before full implementation,  be assured that many players, myself included, will be inclined to  utilize those flaws.  Personally, I won't bother doing it for my benefit  but purely to show you its existence so you will realize that this  system is invalid for the purpose you have assigned it to perform.  That  said, it'll quite possibly be to the benefit of myself and/or other  players at the tournaments at which I participate.

Now, in order to balance out the tone of this message, here's some constructive advice:

The  ELO system itself was fine.  Personally, I think it's the most  effective method for accurately representing a player's ability and  success.  It was the method of implementation that required adjustment  because of how it was being used to award invitations and Grand Prix  byes .  Two simple things* to correct it:

1)  Instead of using a  snapshot of a particular day on which to determine GP byes and invites,  use the period leading up to that cutoff date (90-day, 120-day, 13-week,  or 17-week, whichever works best), using the highest DCI Rating each  player achieved during that period.

2)  Require a minimum amount  of activity during that period to receive the byes/invites.  Total up  the K-values under which each sanctioned match was played and the player  must achieve a minimum value - 400 or 600 seem reasonable - to receive  the associated reward.

The issue of players "sitting on rating"  would be solved.  No more reason to stop playing just because they've  raised their rating above a specific threshold and they'd lose the  benefits of that rating unless they play enough.

* - I posted  this suggestion elsewhere and someone pointed out that a group of  players could falsify a series of tournament results, artificially  inflating each of their DCI Ratings in turn over a qualifying period.   The software used to detect suspicious and fraudulent tournament results  may need to be programmed to scan for such anomalies to prevent such  abuses.

That about does it, and as you agree, avoids replacing the whole system. But nobody asked us. They, instead, apparently spent years implementing a new system that took only a few minutes to think up, and which basically trades fixing the one problem with the old system for all of its benefits.

Clearly a good idea backed by sound leadership.

To apkerm, branewalker, and quadibloc:  The compliments are greatly appreciated.  It's pretty much the only thing on which I have the self-confidence to accept such praise so reading such things truly means a lot to me.  Suffice to say that I don't feel the same about my Lifetime Planeswalker Points ... and I'm ranked 24th Worldwide for FNM!

Also, yes, that homage to the Folger's Crystals commercial is priceless!

How about this for a hoot:  I just followed the "Brian Kibler" link from the article.  His suggestion matches up with the first of the two points I made - using the highest rating over a qualifying period.  Seems to suggest that while they claim to have been listening, all that was actually occurring was that they heard someone was complaining without *listening* to the solution they were proposing.  It's worth reading - he makes other cogent points that are relevant to this discussion.  I still believe my second point is also essential as it requires players to put a sufficient number of points "at risk" over each period to qualify for those rewards.  As with so many things in life, the most effective to encourage the behaviour you want from others is done by applying both a carrot and a stick.

I do respect the opinion of the yeasayers even though I'm not on your side of this argument.  There are just too many factors against using such a system for determining rewards like byes and invites.  I've been at this game since 1993, including considerable time as a Tournament Organizer and game store employee.  Also, I've been playing games (and with numbers) since 1973 (age 3).  When I see invalid data - or in this case FORESEE invalid data - I bring it to the attention of those who have determined what will be collected and will be making decisions based upon it.  I say this from the point of view of someone who would be FAR more likely to receive such benefits under the PWP system than by DCI Rating.  However, I know how good a player I am and I'm not now - nor have I ever been - deserving of byes at a Grand Prix or a tournament invite.

From a purely statistical viewpoint, this new system encourages behaviour I consider unscrupulous.  However, the only options are to perform those acts or watch as those who are playing in locations where those things are being done accumulate points at a rate you are mathematically incapable of matching.  I'm not talking about anything against the Tournament Rules either, only things that will be done to chase those top 300, 2000, and 15000 spots for the Grand Prix Byes and those invitations to Nationals, Pro Tours, and the FNM World Championships.

Sep 9, 2011 -- 7:54AM, DedlyEdly70 wrote:

2)  Require a minimum amount of activity during that period to receive the byes/invites.  Total up the K-values under which each sanctioned match was played and the player must achieve a minimum value - 400 or 600 seem reasonable - to receive the associated reward.

That's a great idea.

I'd be willing to accept Wizards keeping Planeswalker points, and handing out some of the Pro Tour spots based on those, and others based on the Elo with a certain level of Planeswalker Points used to determine who was an active player.

However, for the threshold of an "active player" to be reasonable, that alone would not eliminate the problem of "sitting on rating", so a tweak or two to the Elo half of the system would be needed. In Chess, the tweak is easy; lower K when the game is a draw. In Magic, a more complicated scheme is needed: lower K when there's a big difference in ratings between the players.

You can't change the K value depending on the ratings of the players. That's not what a K value is. The K value isn't actually modified when you get a draw, you get .5 scoring points in the formula, instead of 1 for a win and 0 for a loss. You won't lose as many points when you draw rather than lose against a lower-rated player, but  

Elo calculates a win percentage based on rating and multiplies the K value by that to determine what percentage each player should win. If they win that percentage of the time, their ratings stay the same. If the difference in Elo rating is 400 though, the better player is expected to win over 90% of the time. That's just unreasonable though. No one can win 90% of their matches against an average player. Even LSV, who is #1 in constructed Elo rating right now, and deservedly so. If LSV were playing in FNMs  against 1600 rated players, he'd have to win 97% of his matches to keep his rating the same. I'm pretty sure that even if he were playing against one of those life-size cardboard Jace cutouts, he wouldn't be able to win 97% of the time. Against Chandra maybe, but sometimes Jace just gets there. 

Fix the curve, and you're most of the way to fixing the ratings sitting problem. The number of players actually on the margin is pretty small. 

Copied from another forum, here are some suggestions that could hopefully improve the system which was announced:

Transform the ‘ Planeswalker levels’ into a lifetime achievement system-  replacing the previous player rewards system. Mail out rewards to  players for achieving levels. At higher levels perhaps this is free  entry to PTQs/GPs etc. to allure players to play more and/or higher  level events. Perhaps even give a PT invite upon achieving the highest  levels. Perhaps roll out rewards for higher levels gradually and  grandfather older players to receive the new rewards in order to keep  costs lower.

Limit FNM Points.  Limits are necessary to ensure that players/stores do not fraud the  system. Limits are crucial to have confidence that we won’t spend all of  our time and effort only to discover that we didn’t know/have access to  the cheats to achieve success. Limits let us know that by showing up,  we’re maxing out our possible FNM points for the week. 

Allow worldwide stores to send in an ‘FNM’ tournament once per week,  but  allow some countries and/or stores to hold FNM on a non-Friday  with  WOTC approval. WOTC can approve changes for countries which are  not  conducive to playing on Fridays or stores with hardships.

Reinstate GP invites-  At least a couple, and more based on attendance.  This keeps the dream  alive for MTG Online players or Spikes whose  responsibilities take  frequent precedence over Magic. For many, the invite is the  biggest  allure to attending a GP- especially later in the season when  they  figure out that they can’t keep up with Planeswalker Points.

Give out a few random invites to players who achieve a certain level of Planeswalker Points in a 4 month season.  That would keep players incentivized throughout the season to attend  more tournaments. There could be a 2 or 4 week cutoff before the PT to  announce it- and how exciting would it be to have another announcement  of that sort to make eh? If they want to incentivize attendance, as they  obviously do, they could even weight the odds based on number of events  attended. That way, we could achieve the level, but still want to play  more for better odds.

To make play skill more relevant, multiply a player’s point total by their win %.  This limits exploitation of the system by player’s attempted abuse of  attendance. I think it’s pretty clear that WOTC wants to move away from  win % mattering though, since it potentially leads to sitting on  ratings. 

And a couple of other WOTC-related comments:

Limit the speculation:  If we are expected to take a system seriously, then the system being  implemented should be perceived as being robust. Mike Turian gave us  many important details, but left us with many questions and the promise  of more to come. When they make announcements like this and leave many  details open to speculation, they give off the appearance of taking  decisions like these frivolously. It seems like important changes like  this should be better thought out and/or more completely presented prior  to release.

Stop moving the goal lines:  The requirements to achieve success in Matic can be costly both in time  and money, but historically we see that the benefits received from  success are not what was promised the year before. If players are to  invest what is now being asked of us, for promises of future glory,  there should be an understanding that the carpet won’t be swept out from  under us again next year. There are very few sacred cows in this game,  but fulfilling promises SHOULD be one of them and WOTC has pretty  consistently shown that there are other things more important to them.  When the perception is that money is more important, any changes will  likely be taken quite negatively. 

Hopefully I'm overreacting  and Wizards will make the appropriate changes as necessary to have a  successful system. As it stands now, the system is ripe for abuse, makes  skill comparisons difficult, is discriminatory based on location and/or  lifestyle and is generally deceiving in a variety of ways to people who  don't understand it. As a result of this as well as the manner of  announcement,  the perception given off is potentially that WOTC is  making these changes frivolously and greedily. At the least, this  devalues any sense of legitimacy about the new program.

And to add one more comment: More hardball questions from your 'reporters' would be nice, as it would help give the appearance that your writers really do care about our best interests rather than our perceiving that they're simply revealing the party line. I know you want to put a positive spin and get people excited about the changes, but too far down that road leads to the feeling that underneath the shiny coating, WOTC could really care less about how these announcements truly impact players.

When I wrote to WOTC customer service telling them I was concerned and got a response that looked like a form letter and basically stated 'oh but perhaps you don't really understand the system. Here's a bunch of links for you to learn more about it,' I do believe that I understand it and was clear in my letter about the issues that I had with it. When none of those were addressed, and my letter seemingly resolved/in the trash, I was left with the same feelings- patronized, discouraged and unimportant, which is the opposite of what the article indicates you're trying to do.

Sep 9, 2011 -- 1:47PM, menard1164 wrote:

You can't change the K value depending on the ratings of the players. That's not what a K value is. The K value isn't actually modified when you get a draw, you get .5 scoring points in the formula, instead of 1 for a win and 0 for a loss. You won't lose as many points when you draw rather than lose against a lower-rated player, but  

Elo calculates a win percentage based on rating and multiplies the K value by that to determine what percentage each player should win. If they win that percentage of the time, their ratings stay the same. If the difference in Elo rating is 400 though, the better player is expected to win over 90% of the time. That's just unreasonable though. No one can win 90% of their matches against an average player. Even LSV, who is #1 in constructed Elo rating right now, and deservedly so. If LSV were playing in FNMs  against 1600 rated players, he'd have to win 97% of his matches to keep his rating the same. I'm pretty sure that even if he were playing against one of those life-size cardboard Jace cutouts, he wouldn't be able to win 97% of the time. Against Chandra maybe, but sometimes Jace just gets there. 

Fix the curve, and you're most of the way to fixing the ratings sitting problem. The number of players actually on the margin is pretty small. 

Short response here (for a change).  The adjustment required to do what you're suggesting is to take increase the "400" value in the ELO formula.  It's that value which generates those extreme "expected win percentages" you've mentioned.

Not good for the game. AT LEAST keep the ratings along with this kids' system!

Can you imagine Kai Budde wearing a shirt proclaiming, "I'm a 65th Level Arch Mage!" I can't.

I think we are pretty much screwed over either way...changing this did nothing but add a FNM championship tournament somewhere.  If you noticed they started the competitive season on 29 Aug 2011...I went into the system to see where I stood after a week ( I only played 1 tournament with 14 points awarded) so I filtered for Washington state where I live to see how I compared to others and friends that play and walla what do I see!!!



It's been one week and for the most part, we can grind our you know what off till Christmas and still never catch Jesse.  Why because Wizards started this stupid new system the week of PTQ Philly instead of the Monday after so we could all start off with a clean slate and grind it out to see how well we might be able to do...

I give up...I'm 50 and still drag race and work so I don't have time to grind 24/7 but my kids are older and so I have time to play a few times a week, but I could never catch Jesse even if I played 7 days a week and went undefeated...

At first I thought sweet, we all have a chance to start over and boom, my bubble was burst...we were fooled again...Wizards only caters to the pros, but the ones that spend all the money get screwed again and a majority of smaller stores will get hosed by this too.  Guess it might be time to spend my money elsewhere again...

This actually gave me incentive to not play.  If you can't keep up your rating at an FNM, then you don't deserve that 1 or 2 round bye.  I routinely smash everyone at FNM, or occasionally lose the 1 match to keep my rating above 1900 or eke it out to the 1950 mark... then I do the ptq thing and that's where the rating really fluctuates.  Now I have no incentive to play at FNM or other tournaments because there is no way I can keep up with other people grinding, even if I win every single match.  The only events I will ever play at now are PTQs and Grand Prix.  So wizards, you got the short term money boost you wanted, but I think this is really going to hurt you in the long run.

Despite all the bad things the new system can bring, I can't help but appreciate it.

1) I suck
2) I play frequently

And quite frankly, seeing my low rating was getting pretty depressing.

May I apply a different perspective?

1) I do not care
2) I show up at Wednesday's "Open Play"

And quite frankly, I have no idea what my rating is (DCI#[6]1322781).  Something broke on the DCI site in 2002, and I haven't bothered Wizards to fix it.

When I played competitively, I regularly got my ass kicked by that Matt Place, at The Battlezone in Overland Park, Kansas (and a shop whose name I can't remember, in Independence, Missouri).  In the years since then, I just stopped giving a damn.

Two short things:

Aaron Forsythe]Rating is supposed to be a measure of how good you are

Then don't call PWPTS a rating, because that's just not what they do.

And: The old system was at least intended to treat everyone equally by giving them a comparable rating across the entire player population. Now, that was flawed, too, and it's good that WotC is fixing the system. But PWPTS neither give you a measure of skill compared to other players nor does it give every player the chance to compete equally.

I am very much interested to see how the Top 100 actually turn out. Our prediction is that all Top 65 players will be American - if that becomes true, the rest of the world might as well not exist. Then the PT will move to the US for the entire year, and everybody else gets GPs as large cash tournaments and that'll be that (except for the player's club, which is the only saving grace because that, at least, gives every PT player the same c wrote: