Planeswalker Points and You
Editor’s Note: This article is a result of collaboration between those signing at the bottom, and as such should be treated as a group effort, rather than being written by any one individual.
“Playing is good. Winning is even better.” – Mike Turian
“It looks like we are in for a very interesting lesson on incentives.” – Patrick Chapin
“If you’re not careful, your whole life can become a ****ing grind.” – Our Hero, Rounders
Mike Turian gave you the facts, or at least the facts we know. Now it’s time to look at what happens as a result – a smaller, weaker Pro Tour, bereft of many of the best players. Great news if you’re a Hall of Famer like us, but not so good if you’re anyone else, or even worse, live outside of North America.
The best guide for analyzing what will happen in the future is to look to the past. Magic had an ELO rating system that it used for many of the uses to which Planeswalker Points will now be put. The old system had the following incentives:
1) Your rating qualified you for benefits at thresholds, by far the biggest being PT invitations, but GP byes also being important. Thus, some players had strong incentive to care about rating, while most did not.
2) Ratings were zero sum, but their marginal value varied wildly, so often everyone was better off if one player won and the other lost.
3) If you compete in high level events, your rating will trend towards a high set point. If you compete in lower level events, your rating will trend towards a lower set point, since the win expectation curve was wrong and Wizards never decided to stop and fix it. Your goal was to play against the best possible opponents and avoid poor ones.
4) If you play matches where you are especially likely to win, you gain points. If you play matches where you are less likely than usual to win, for whatever reason, you lose points. Because Wizards used the same ratings curves as chess, where unlike magic someone really can be 90%+ to win a match, this meant players needed to avoid playing against players ranked significantly below them.
5) Playing causes your rating to fluctuate, whereas not playing makes it remain stable.
Players in different situations responded to this in different ways. Most of the time any given player would be indifferent to rating and would act as if ratings did not exist, other than to have something fun to look at every so often. Perhaps the status involved would help get that player motivated, but nothing more. For them it was all in good fun.
As players ratings climbed, they started to become eligible to get byes at Grand Prix tournaments. Byes are highly valuable, especially in multiples, so this could be highly motivating. As they climbed higher, players were close to qualifying for Pro Tours, and that was a whole different level. Once a player got close to that, they would radically alter behavior. Here are some common alterations:
1) Player would always play to maximize win percentage in a tournament. Fun decks would get discarded in favor of those that were marginally more effective. Experiments would be run in playtesting, not when points were at stake.
2) Tournaments where the player could not be at his best, for whatever reason, would often be skipped.
3) Tournaments where the player would face poorly rated opposition would be skipped.
4) If the player had an unusually high chance of winning due to the matchups in his bracket or the strength of his draft or sealed deck, the player would stay in the tournament even if otherwise the player would drop.
5) If the player had an unusually low chance of winning, for similar reasons in reverse, player would drop, even if undefeated. They would also avoid playing in smaller local tournaments against weaker competition.
6) Player would seek out tournaments where points could be obtained. In best case this was high-K events like PTQs or GPs. Sometimes this would include places where player would not be allowed to lose. Fraud occurred, from friendly help up to entirely fake tournaments and players.
7) Once the threshold necessary was achieved, even mid-tournament, player would sit on rating and refuse to play.
Thus, there was a rating-conscious class of players who tried to use the ratings to break into or stay in the Pro Tour Player’s Club (or “Gravy Train”), a group in the middle that kept track but didn’t sweat it, and a larger class who didn’t care.
Some of these effects were clearly good, but many of them are obviously stupid and bad. A player choosing not to play in order to maintain rating is bad for everyone, in Magic as in chess. I know back in the day there were chess tournaments I didn’t enter due to not feeling my best and wanting to keep my rating up – chess ratings are Serious Business – and that was when the rating meant nothing. I’ve also been known to sit on rating for months on end when I had to. Many Pros essentially gave up casual tournaments in order to maintain their ratings.
A lot of this was pretty terrible. Tournament attendance at Neutral Ground was often substantially lower than it would otherwise have been due to this problem. A WotC employee I will not name put it best when he simply said “Sitting on ratings is ****ing stupid.”
So what does this lesson teach us?
1) People respond to incentives.
2) People respond to incentives.
3) No matter how much you think people respond to incentives, and how many times you think you’ve taken this rule into account, they respond to them more.
4) While for the most part ratings encouraged players to play more rather than less Magic, the old system often resulted in people playing less Magic instead of more.
5) This was pretty f***ing stupid.
In short, the ELO system had some issues. In terms of actually choosing the best players it did a decent job, but it had rather bad distortionary effects on those who cared about it most. Luckily the primary way to reach the Pro Tour for most players was to qualify via a Grand Prix top 16 or a PTQ win, or by picking up Pro Points, with the ratings an afterthought in case they went on a hot streak without picking up titles. Thus the size of these distortions stayed comparatively small. Scott Larabee mentions that it can be so extreme as players taking a four month break for one bye at a Grand Prix, but I express skepticism that players would commonly choose to take such tradeoffs – and if they would, that only emphasizes points one, two and three above.
The ELO system could have been be fixed to work. Redrawing the curves to reflect Magic and not Chess win percentages deals with half the issues. The other half could be dealt with either through ratings decay, or preferably through letting people bank their invite for a future Pro Tour. Players could only ever have one banked invite at a time, and it would be used up if they attended a Pro Tour, even if they qualified through other means. This would have the nice effect of letting the Japanese player who qualified for PT Paris attend PT Nagoya instead. Rather than dwell too much on the road not taken, let’s move on and take a look at the new system.
Planeswalker Points (hereafter PWP)
The new system replaces ELO ratings with cumulative ratings that are periodically reset. There are lifetime ratings, but they don’t count for anything. In addition, it scraps the top 50 finish at Pro Tours and the top 16 finish at Grand Prix as methods of making the tour, instead folding those results in the ratings. This forces players to be far more conscious of the PWP than they were of ELO. ELO was optional. PWP is also optional, but your other options have changed.
The Quest for the Pro Tour
Let us assume a Magic player Alice who wants to qualify for the Pro Tour. Alice has the following methods available to her:
1) Win a PTQ, either online or in person.
2) Qualify through the accumulation of PWP.
3) Qualify through the PPC, assuming it or its replacement exists.
4) Be elected to the Hall of Fame.
That’s four tracks, and they have very different incentives. For the moment let’s assume that the Hall of Fame is out of her reach. If it happens, wonderful, but it’s not something one can count on.
Path 1: Win a PTQ
Alice’s path here is simple. Playing more PTQs is good. Winning a PTQ is what matters. This is what the old school road warriors did, going to a PTQ week after week hoping to win an envelope. Once winning is eliminated as a possibility, she will drop out or not as she prefers. Note that a Grand Prix now offers zero slots, so while on this path a PTQ is by far the best option. Alice would drop from the Grand Prix to play a PTQ instead. Also note that online PTQs give no Planeswalker Points. If she does especially well on this path and gets enough PWPs incidentally, she can transition to plan 2. Well, actually, she can’t, as we’ll see. I say this because science has shown that spoilers add to reader enjoyment.
Path 2: Accumulate PWPs
This will be a harsh path. Alice is going after one of a hundred slots; for the moment let’s assume that reserving a few slots for each region is an insurance policy but doesn’t change the list that much; realistically the North American and presumably Europeans were locks anyway, so at most fifteen slots can change hands and the Japanese slots are also mostly safe due to their GP warriors. If no one alters their behavior from what they did in the past, the threshold will be about 1700 points. However, that’s a fantasy land. People respond to incentives, as noted above, and the tournaments available have changed. This number will go up. The only question is how much. Keep in mind that Top 300 used to be 1400 or so points. That’s a lot of people on the bubble.
Alice’s behavior on this track is incentivized to maximize her point total until it reaches the threshold where her invite is secure. It would obviously be great to win a PTQ along the way, but Alice will still need enough points to get her byes if she wants to grind for the next Pro Tour, so there’s no escape. Byes outright count as wins, and byes are great at earning the right to play more Grand Prix rounds, so being without them in a season is a severe handicap. With a Grand Prix almost every week, winning Trials isn’t practical as an overall solution.
Each weekend Alice makes a choice. She can fly to a Grand Prix, go to a PTQ (if one is available), go online for a PTQ (if one is available) or go for some other event. The multipliers on other events are relatively poor, but large events like Star City Opens are reasonable choices if they stay large, especially with one on each day. A PTQ grants a good multiplier, and likely is superior, although going to another event the other day is even better. Online PTQs are right out, of course, since they’re valuable time wasted for no points and even bigger fields than before now that those not grinding have fewer choices.
Alice’s best bet, of course, is to go to a Grand Prix. At that Grand Prix she gets the biggest multiplier, a whopping eight, and lots of matches to play some of which are hopefully byes. Her hope is to make day two, and hopefully win cash and glory, but she won’t sweat it. If she can’t, she’ll finish up day one and play side events. They’ve eliminated the PTQ on Sunday, but that is fine because all side events at the GP have a 5x multiplier! That means that she can go around entering side events, collecting participation points and then dropping to go enter other side events and collect more participation points. Depending on what pace is required, she may or may not be able to afford incidentally attempting to win some of these tournaments, especially limited events where her deck proves excellent.
Week after week, she’ll go through this grind. If she plays well, she’ll win a higher percentage of her matches, scoring more points. Making day 2 of more GPs keeps her in high multipliers, which is good, but decreases participation points and matches played, which is bad; in many ways the side events are actually the better bet as she gets more participation points, more rounds and easier opponents. If there is punishment here it is light.
At the end of the season, Alice and all her competitors for those 100 slots will compare their performances. Those who went to the maximum number of Grands Prix will have a prohibitive advantage, as will those with three byes. Those who got to start the season with a Pro Tour will also have huge head starts, as no one is kicked out of day two so most will get four or five hundred points; this alone is a huge source of threshold inflation.
The key question will be: Are there 100 or more people willing to grind at this level?
If there are substantially more than 100 players willing to do this, then everyone else is shut out.
Think about the points Alice will get if she goes full blast in with three byes. Suppose she lives in the USA. She’ll be in Austin, Orlando, Lincoln, Baltimore, Seattle, Indianapolis, Nashville, Mexico City and Salt Lake City. At each event, she’ll win her first three matches (since they’re byes). She’ll then play the rest of the Grand Prix day 1, let’s say 6 more rounds, and then play day 2 or play side events. Let’s be skeptical and assume her average weekend is a 3-3 day 1 after byes, for 6-3 and no day 2, then 6-3 in side events; this Alice isn’t actually all that great at Magic. The weeks there is no GP, she would go to a PTQ as the next best option.
She now gets: 9 GPs at 6-3, for 6 wins and about 8 participation points each, or 24*8 = 192. She plays in 9 sets of side events, for 6 wins per set out of let’s say 6 events, three of which are drafts (1 participation each), and three of which are major events (let’s say average 4 participation each), all of which is times five. That yields 15 participation and 18 for wins for 32*5 = 160 additional points. Doing this nine times yields 352*9 = 3168. FNM also has a 3x multiplier, so if she’s truly dedicated she’ll grind that too.
Now let’s look at Brian Kibler, so we have someone to blame for all this. He’s awesome. He made top 16 of the PT going 12-4. He goes to five GPs, top eights three and wins one! The other two he makes top 16 and top 32. What a season.
You’d think that, wouldn’t you?
His PT was worth 504. At the GPs, he won 17, 16, 15, 13 and 12 matches, but after that he does drafts with friends because that took all weekend. So he gets 73 match wins, 5 sets of participation points (let’s say 8 each again, times the multiplier), and an 8x multiplier, which comes out to… 2576.
That’s right. Alice gets the slot over Brian. Brian thanks the Godslayer that he’s in the Hall of Fame.
If you think a little tinkering with modifiers will fix this, my response to you is: Really? Alice had a horrible season. She won barely more than half her matches, won no money, earned no glory. She also didn’t do a lot of mysterious winning once she could no longer win the tournament, if you know what I mean. Realistically, Alice will hardly ever do this poorly; if nothing else she’ll make a few day 2s of GPs. Brian had a ridiculously strong season, going to five Grands Prix and winning not just more than half but most of his matches. Brian is probably being called one of the hottest players on the tour right about now. Plus we all love Ascension on the iPad and his Counter-Cat Zoo variant was really neat.
Thanks to the elimination of the PTQ and FNM at GPs, Brian managed to keep it somewhat close and has a shot at getting one of the slots depending on everyone’s behavior, but he is far from secure. The original version of this thought experiment had different assumptions and had Alice win going away; the good news is that fixing a few mistakes puts us within striking distance of the right result, although the 5x multiplier at GPs gave Alice a bigger edge than she had in iteration two. If you increase the multiplier in a non-linear fashion by rewarding strong performance, you can bring skill somewhat back into the mix.
At its heart, however, the points race is a pure grind. Those who grind the most get the slots. Think of the formula as something like this:
Points = (Matches * Win % + 3 *Byes * 3 + Participation) * Multiplier
For a GP, you can substitute 7-9 for Participation. Therefore, everyone gets 2-3 match wins for showing up! Then you get 3 more from your byes! Match win percent on the Pro Tour is never going to reach 70%. For a Grand Prix if you exclude Byes you might hit that, but no reasonable player is under 50%, and there’s five free wins to start you off if you come bearing byes from last season. The deck is stacked in favor of whoever comes more often.
The good news is that there almost certainly will not be 100 grinders with Alice’s dedication and ability to travel. The bad news is that what matters will not be how much players are like Brian. It will be how much they are like Alice, as Brian’s 2576 represents the high end of what is possible without a PT top 8 and may or may not be good enough.
Also, if you’re not American, unless you’re willing to fly off-continent, you’re only competing for your regional slots. In terms of the 100, you were never in it.
3. Qualify with PPC points
We don’t even know if such a thing will ever be possible again. If it is, Alice will need to get her points from Grand Prix events since there are only three Pro Tours and with so many Grand Prix the grinders who were shooting for the hundred slots will eclipse her if she doesn’t put up good results. The good news is that now winning is heavily rewarded, so she’d better deliver the goods. If she does, she gets to stay, and rewards for winning scale enough that she can likely pick and choose her events, especially since she knows what threshold she needs, although as is true today the marginal value of moving up the ranks pushes her more and more towards a similar grind.
What will Alice do?
Alice knows that there are no half measures. If she starts down the PWP path, all work is lost if she doesn’t cross the finish line. If she starts down the PTQ path, what good will making the finals a few times do? She’ll still be well short, having played in smaller tournaments with less rounds and smaller multipliers and well behind the grinders, even if it happens to start the season. There’s effectively an all-pay auction for those 100 slots. In the first season with PWPs, Alice will have to take a guess as to what the threshold number of GPs is to make it in; her performance is good for plus or minus one or at most two, and a lack of byes might cost her an additional tournament, which she’ll factor in. After a round or two of that, it will likely become clear what the number is.
Over time, it will be clear about what the top 100 threshold will be. At that point, a number of players usually between 90 and 110 will decide to take measures necessary to pass that threshold after seeing their PT performance. Some seasons those who cut it close will miss, while in others there won’t be 100 applicants and a handful of people will get invites they didn’t expect, although they are likely to mostly be players who were otherwise qualified but wanted to compete and thus willing to score in no man’s land.
The result is three or four groups. First we have the Path 4 people like us, who have Hall of Fame invites. Second, we have the remaining Gravy Trainers, assuming such a thing exists, who qualify by doing well at tournaments, but the road involves being a Grand Prix warrior and remaining vigilant at all times lest one fall from grace. Third, we have the Grinders. These players will do everything in their power to maximize their PWP, whatever implications that has, taking flight after flight to Grand Prix after Grand Prix and gaming side events and other tournaments. The first season will be rough without three byes, the second one they’ll have them. They’ll buy their way into the Pro Tour with that long, hard slog, and they’ll never be able to stop lest they lose their byes and have to start over without them. Finally, there will be the PTQ players who comprise the majority of each Pro Tour, who will occasionally break through to a Pro Tour but have a very difficult time doing more than that. Even if one were to win a Pro Tour, they’d likely be back to PTQs within a year.
Consider two of the best stories of the recent Pro Tour: Brad Nelson and Paulo Vitor Damo Da Rosa.
Brad Nelson is from North Dakota. He plays Magic Online all day, every day. He qualifies for the Pro Tour! Everyone is excited. Brian David-Marshall writes up stories about the grinder of all grinders, ready for action. He gets to Honolulu, plays his heart out, and finishes 9th on tiebreaks.
There’s a good chance we never see him again.
He lives in North Dakota. He can’t play any events during the week other than at most FNM, and it’s a small one. He has no byes and no appearance fees, and minimal reasonable places to play a PTQ. He can’t afford to go on a Grand Prix binge, so the points from 9th place, which aren’t that many anyway, go to waste. We never get his ideas, his decks, his good times or his epic Player of the Year showdown.
Paulo’s story goes a similar way. Before breaking through to the Top 8, his first five events were top 50 finishes that allowed him to remain qualified. Without that lifeline, he is even more shut out than Brad. He can try to be one of the chosen five from Latin America, again starting from nothing, or he can keep grinding PTQs. We likely never get his story either.
We are killing the next generation of Pros in the crib. We are killing the current ones after the PPC runs out, with lightning speed if it isn’t replaced but rather quickly even if it is. Only members of the Hall of Fame and the grindiest of the grinders survive. The thing is the grinding champions will almost never be the best Magic players. Magic is one of the rare sports where its top players almost always have more lucrative options elsewhere. Too many top players leave competitive play as it is, when the time commitment becomes difficult to balance with life and a job. Well, the time commitment just went through the roof and the chances of making the Pro Tour went through the floor.
In other news, players at FNM will get more points for playing in larger groups, so every travel area will experience a shift of players away from its small stores towards its large ones. In some cases this will kill FNM at the small stores, since the moment they can’t reliably get 8 people it’s all over. Note that being large doesn’t on average make matches any harder or easier.
In a perfect world, Wizards would go back to a system without Planeswalker Points, tweak the ELO system, and maybe make all ways to qualify a little bit harder to deal with PT size inflation. However Wizards has too much invested – the new system can’t be scrapped. The details are far from fixed especially in the medium and long term, so what can we do to fix the system that might actually happen?
Here’s the top of a practical fix list:
1) Keep the Pro Players Club and Pro Point system. This is more important then everything else. The last thing we want are the Luis Scott Vargas’s of the world needing to grind into every Pro Tour. We need our top players at the biggest events, and they need the security that consistent strong finishes will keep them qualified. Also, unlike PWP, the PPC looks back over a full year, allowing for a hiccup or two without immediately falling off the tour, like Jon’s 251st at PT LA2000 en route to winning Nationals and Worlds that year.
2) Participation points need to be severely scaled back, at a minimum for side events. Right now you can get multiple free wins from entering a tournament, and enjoy the main event multiplier on them. Participation points flat out shouldn’t get the event multiplier at all, but if they must then at least take that away from side events. Originally I assumed this was going to be fixed, but it looks like it isn’t. All side events at Grands Prix will have a 5x multiplier, so we’ll get to see what happens when a bunch of people who are experts at gaming systems realize that their point maximizing option is to sign up for and drop from as many side events as possible. We’ll also get the answer to the question: “Can I sign up for a side event while playing in Day 2 of a Grand Prix despite the fact that I have no intent in actually playing in that event?”
3) Instead of 3 points per win and 1 point per draw, go to N squared points for N wins, with draws counting for a third of a win (and fractions being retained). This rewards long tournaments, real Magic play and outstanding performances organically, and is the big proposed fix to the system.
4) Adjust event multipliers, at least to raise the PT. If you don’t do #3, adjust radically.
5) Scale up the multipliers for top performances. There should be a bonus multiplier for all performances that currently receive Pro Points.
6) If for some reason #1 is unpalatable, announce as quickly as possible that there will be an heir to the Pro Players’ Club at least to the extent that some players will qualify through top performances over the course of an entire season. Work out a system that lets players know where they are at. Details and thresholds can be worked out later, but current regulatory uncertainty must be contained.
7) At a bare minimum, top 32 of the PT and the finalists of each GP get slots. The GP slots should pass down to avoid collusion in the top 8.
8 ) Use ELO for pairings! This provides strong incentive not to game the system too much to get PWPs as you’ll kill your ELO which is both relevant and now highly visible (source of high status), while providing a solid reward to those who perform well. This can at least be done on the GP and PT level. Make sure to include some amount of randomness to make sure players can’t game the system too much, and to avoid players having the same matchups each week.
Wizards may have done the math, although many past seasons give absurd results in any number of ways. They clearly failed to do the game theory and then rerun the math. Like many game designers before them they have failed to notice that one type of behavior has been rewarded so far beyond others as to make the others irrelevant. I’m all for playing being good and winning being even better, but this system can be summed up as “Playing is very good. Winning is slightly better.”
Make your plans accordingly.