Re: Pete. (aka. Gaming the system-- who's to blame?)
- Thread starter Flaming_Spinach
- Start date
- Status
Cyrus
Iron Chef - Master Emeritus
Replies in bold. Always a good discussion, Ian.
Now, for a concise proposal for a rewards system of rating invite that mirrors the current system we have now. It's imperfect, but I suggest that you guys refrain from dissecting this amateur proposal; rather, just acknowledge things that could work to solve the current problem (and if you're P!P, then feel free to lift it if you like it :smile
.Brand name: PLAY! POINTS
Structure: issue invites to the Top X in Play! Points per zone (say, top 40 in North America)
Process of rewarding points:
Battle Roads: one (1) point per win
City Championships: four (4) points per win, two (2) per every second place, and one (1) for every top four finish. Could be reduced to discourage team play/event hopping
State Championships: eight (8) points per win, four (4) points per every second place, two (2) for every top four finish, and one (1) for every top eight finish.
Regional Championships: sixteen (16) points per win, eight (8) points per every second place, four (4) for every top four finish, two (2) for every top eight finish, and one (1) for every top sixteen finish.
National Championships: tentative. May need to be structured appropriately per rating zone.
So yeah, take it for what it's worth y'all.
Oh, and re Ice Cold: I only went to five Battle Roads that one season, so your suggested cap would have done absolutely nothing.
. Cool idea, though!
Last edited:
vegitalian
New Member
Actually John, Decipher (creators of SW CCG, LOTR TCG, Star Trek, Austin Powers, .hack//enemy, and a bunch of others) had a system just like that. They called it Premier Points, and it was used to gain invites for Worlds and Premier events (skipping Day 1/Grinders). It worked fairly well.
The main principle of the system is to encourage as much play as possible, increasing attendance at all of their events (no one would be sitting out of BRs due to ratings), and keeping the majority of players until the end of the tournament. I fully support this system - there are few downsides to it (if an appropriate balance can be made with number of points). They had ratings as well, but it was more of a way of comparing yourself to others, and they did not reset.
The main principle of the system is to encourage as much play as possible, increasing attendance at all of their events (no one would be sitting out of BRs due to ratings), and keeping the majority of players until the end of the tournament. I fully support this system - there are few downsides to it (if an appropriate balance can be made with number of points). They had ratings as well, but it was more of a way of comparing yourself to others, and they did not reset.
Replies in bold. Always a good discussion, Ian.
Now, for a concise proposal for a rewards system of rating invite that mirrors the current system we have now. It's imperfect, but I suggest that you guys refrain from dissecting this amateur proposal; rather, just acknowledge things that could work to solve the current problem (and if you're P!P, then feel free to lift it if you like it :smile
Brand name: PLAY! POINTS
Structure: issue invites to the Top X in Play! Points per zone (say, top 40 in North America)
Process of rewarding points:
Battle Roads: one (1) point per win
City Championships: four (4) points per win, two (2) per every second place, and one (1) for every top four finish. could be reduced to discourage team play.
State Championships: eight (8) points per win, four (4) points per every second place, two (2) for every top four finish, and one (1) for every top eight finish.
Regional Championships: sixteen (16) points per win, eight (8) points per every second place, four (4) for every top four finish, two (2) for every top eight finish, and one (1) for every top sixteen finish.
National Championships: tentative. May need to be structured appropriately per rating zone.
Personally I don't care because I just judge and play for fun. I am not a competitive player.
But just saying those who can't travel will cry foul over any pro point system. Without any chance of losing points players who can travel will go to as many events as possible. It is a great idea and I agree with it though.
Cyrus
Iron Chef - Master Emeritus
Thoy: as I suggest in the response to NoPoke, there are ways to alleviate (or even outright eliminate) the road trip factor. Even beyond a direct solution, POP-regulated scheduling does a great job at limiting people's ability to rack up points.
Back on the topic of byes: I had a conversation with a talented player today - a guy with at least one bye at this year's U.S. Nationals. And do you know what he was considering? Playing out his rounds instead of enjoying the prize!
...And can you guess why he was considering it? Because in the event that he did well at Nationals, but not well enough, there was a chance that he'd whiff the invite! And the reason for this is simple: the ELO system as it stands. So, ironically, you have two sides to the coin: you have the players who might abuse the byes and drop...And then you have the players who might give them up in exchange for a better chance at making Worlds. These may be minority cases, but it's all pure insanity regardless. ELO has some very big unintended consequences on competitive organized play.
So what's the solution to this glaring flaw? Aside from Play! Points, an ELO-centric answer would be eliminating the byes, and just give them perfect opponent's win % at the end of swiss. That way, you preserve the spirit of the prize (a leg up on the Nationals competition), yet at the same time give the players a chance to gain points, while eliminating the abuse of the system at the same time.
Back on the topic of byes: I had a conversation with a talented player today - a guy with at least one bye at this year's U.S. Nationals. And do you know what he was considering? Playing out his rounds instead of enjoying the prize!
...And can you guess why he was considering it? Because in the event that he did well at Nationals, but not well enough, there was a chance that he'd whiff the invite! And the reason for this is simple: the ELO system as it stands. So, ironically, you have two sides to the coin: you have the players who might abuse the byes and drop...And then you have the players who might give them up in exchange for a better chance at making Worlds. These may be minority cases, but it's all pure insanity regardless. ELO has some very big unintended consequences on competitive organized play.
So what's the solution to this glaring flaw? Aside from Play! Points, an ELO-centric answer would be eliminating the byes, and just give them perfect opponent's win % at the end of swiss. That way, you preserve the spirit of the prize (a leg up on the Nationals competition), yet at the same time give the players a chance to gain points, while eliminating the abuse of the system at the same time.
After some research it seems that even Chess, who the ELO system was invented for, doesn't really use it in a competetive capacity any more, they either use newer systems or modifications of ELO. The wikipedia said ELO is nice because players can calculate their own scores adding to transparency, but P!P doesn't operate transparently in that regard anyway, aside from seeing it happen after a tournament.
Flaming_Spinach
Feature Editor
"Play! Points"? I like it. Creative, fun. Rolls off the tongue nicely. :thumb:
I think the solution is some version of the ProPoints system -- a system where players gain points for making the topcut at a tournament, but get no change if they miss the cut. This is, as stated before, a reward-only system that never penalizes players for participating and getting unlucky.
"But OMG F_S! Reward-Only systems greatly favor those who can travel all over the place!
You are forgetting something: the system we have right now does the exact same thing! Someone who can play in 7 Cities, 2 States, and 1 Regionals will always have an advantage over someone who plays in 2 Cities, 1 States and 1 Regionals. I don't know if it is even possible to alleviate this problem outside of giving out all invites at a single tournament (ie. Nats), or at least on a single day (ie. Regionals).
The fix for this, I believe, is to let players attend as many tournaments as they wish, but only count the top X tournaments (5 or 8 or 10 or w/e. (I will use 5 as the example for the rest of this post)). If you make the cut in 5 tournaments (for example), the points you gain in each of those tournaments adds up to your final score. If you make the cut in 10, only the top 5 will count towards your score. Make the cut in 15? Still, only the best 5 count towards your score.
The most critical part of implementing something like this would be ensuring that the events are properly weighted. For example: does winning a City equal 2x, 2.5x, or 3x the points for a Battle Road? And how do States compare to Cities? Etc. Etc. That, I think, is the critical question.
Furthermore, how does one City compare to another? Does a CC with 10 players earn the same points as one with 40 players? I don't think it should, but the weighting of tournaments based on entrants is not a simple question. (I have some ideas for how this could be done, but I am not going to post them here because they would go over the heads of 99% of the people on the 'gym. Still, if you want to hear my ideas, P!P, you have my contact info.)
As an aside, (as stated above) most chess systems use a modified version of ELO. The "Glicko" system is the most popular AFAIK. Glicko uses complicated algorithms to assign players "uncertainty factors". If your uncertainty gets too high, the K-value of your matches increases. The effect is something like this: If you achieve a rating (say, 1800) and stop playing for a length of time (say, a year) the system assumes that your rating is no longer accurate, so gives you more points to win or lose to get your score back to an accurate representation of your skill as soon as possible.
Again, there are many chess rating systems. This is just one.
Take care.
I think the solution is some version of the ProPoints system -- a system where players gain points for making the topcut at a tournament, but get no change if they miss the cut. This is, as stated before, a reward-only system that never penalizes players for participating and getting unlucky.
"But OMG F_S! Reward-Only systems greatly favor those who can travel all over the place!
You are forgetting something: the system we have right now does the exact same thing! Someone who can play in 7 Cities, 2 States, and 1 Regionals will always have an advantage over someone who plays in 2 Cities, 1 States and 1 Regionals. I don't know if it is even possible to alleviate this problem outside of giving out all invites at a single tournament (ie. Nats), or at least on a single day (ie. Regionals).
The fix for this, I believe, is to let players attend as many tournaments as they wish, but only count the top X tournaments (5 or 8 or 10 or w/e. (I will use 5 as the example for the rest of this post)). If you make the cut in 5 tournaments (for example), the points you gain in each of those tournaments adds up to your final score. If you make the cut in 10, only the top 5 will count towards your score. Make the cut in 15? Still, only the best 5 count towards your score.
The most critical part of implementing something like this would be ensuring that the events are properly weighted. For example: does winning a City equal 2x, 2.5x, or 3x the points for a Battle Road? And how do States compare to Cities? Etc. Etc. That, I think, is the critical question.
Furthermore, how does one City compare to another? Does a CC with 10 players earn the same points as one with 40 players? I don't think it should, but the weighting of tournaments based on entrants is not a simple question. (I have some ideas for how this could be done, but I am not going to post them here because they would go over the heads of 99% of the people on the 'gym. Still, if you want to hear my ideas, P!P, you have my contact info.)
As an aside, (as stated above) most chess systems use a modified version of ELO. The "Glicko" system is the most popular AFAIK. Glicko uses complicated algorithms to assign players "uncertainty factors". If your uncertainty gets too high, the K-value of your matches increases. The effect is something like this: If you achieve a rating (say, 1800) and stop playing for a length of time (say, a year) the system assumes that your rating is no longer accurate, so gives you more points to win or lose to get your score back to an accurate representation of your skill as soon as possible.
Again, there are many chess rating systems. This is just one.
Take care.
What if you simply increase the K value of nationals?
Then, people will be less likely to sit out, because there's a higher chance that people will overtake them in points, while still allowing those with a clear lead to sit out of the event if they want - if they want to risk it.
Then, people will be less likely to sit out, because there's a higher chance that people will overtake them in points, while still allowing those with a clear lead to sit out of the event if they want - if they want to risk it.
Flaming_Spinach
Feature Editor
I felt this was important enough to go back and quote. If seeing how ELO damages the integrity (and spirit) of this game is not enough, a look at the way other games are run should present some fresh ideas and perspective.
We have players dropping after securing their place in the topcut at major tournaments (Regs, Nats, etc). In what other game would you see this?
I LOL'd
QFT
Jahikoi said:
That would probably not work.
The problem with Nationals as a rating-booster is that such a small number of players will actually end up positive on the day. The vast majority will go 7-2 or worse, and not gain any ground on those who believe themselves safe. Those who go 8-1 will gain points, but will make the cut. Unless they want to drop after making the cut, each of them will end up losing in the top cut (except one).
The result is that most everyone at Nationals will either experience no change to their ratings, or get a very small boost. Those that receive the most points will get into the Top8, where they get their automatic invite. Doubling the K value would not change how many players go positive on the day.
So, no. I don't see that working at all.
keniselvis
New Member
Mail the stipends AFTER the tourny. If people choose not to play to preserve their ranking...it's cool. They just do not receive the stipend. It's their choice.
The fact that you give out stipends for people to go to bigger tournaments BLOWS me away. I think that is so cool.
The fact that you give out stipends for people to go to bigger tournaments BLOWS me away. I think that is so cool.
Yes, if there was only 3 weekends of cities, that IMO would be the best solution, and then use P-P
Glumanda
Rulings Compendium Translator
What about a Play! Point system in which you can buy stuff in a Play! Point store?
In the Play! Point store you could buy things like:
Worlds invites
Travel awards for US Nationals or Worlds
First-round-Byes for Nationals
Exclusive Promo cards
etc. pp. (just be creative here)
Points will expire 24 months after you earned them.
That way, you could diminish team play because you excess points are worth something and you do not want to give points away. Players who can only play some tournaments would still have the chance to do something with these points or even save their points for two years to earn something cool. And: There is no incentive at all to start in round 2 or drop before the end of the tournament.
In the Play! Point store you could buy things like:
Worlds invites
Travel awards for US Nationals or Worlds
First-round-Byes for Nationals
Exclusive Promo cards
etc. pp. (just be creative here)
Points will expire 24 months after you earned them.
That way, you could diminish team play because you excess points are worth something and you do not want to give points away. Players who can only play some tournaments would still have the chance to do something with these points or even save their points for two years to earn something cool. And: There is no incentive at all to start in round 2 or drop before the end of the tournament.
Shino Bug Master
Front Page News Editor
You know, this idea has 3 very nice thigns going for it.
1) Collectors and PokeParnets, who come to play for fun or come to try and complete collections and just play for the sake of beign involved. They're at every event, earning ro stelaing points, not for the sake of the overall goal, but with this system, they get to do things with the points they earn.
2) It prevents the concept of overshooting. Whenever invites get ahnded out for the top 40 places, there is often a 100 point difference between number 1 and number 40. This makes it so that there is an applicable difference between the 2.
3) Prevents the double stuffing, where players win Invites throguh Nats or Worlds and then are in the Top 40 bracket and get a ghost invite that doesnt serve any purpose.
it would be difficult to decide the poitns for each though, and there would have to be other ways to earn an invite, or travel money to make the prizes make sense. I don't think buying byes makes much sense either, Play ! Points should encourage players to PLAY Pokemon, not skip rounds :lol: It'd be cool if there was Pokemno merchandise available though, like plushies, deck boxes, sleeves, mats etc. Maybe even bundles.
I'll throw my two cents in here.
So one of the Masters in our area (he's on these boards) is looking at a worlds invite for the first time ever...and is playing very conservatively to preserve these points.
He went something like 4-0 Drop in states, and was going to do the same thing in Regionals when he made the mistake of deciding to go ahead and play top cut...end result: Rating got hurt when he lost out in top 32 or 16 (forget which). His plan for Nats is to go something like 3-0 DROP (one of those is a bye).
His comment to me earlier this year was "The system benefits those who DO NOT PLAY."
Another example: Spring battle roads. There's a simple rule here. If you have the points to compete for Worlds, don't play it. Spring BR has ALWAYS been horrible attendance for exactly this reason. Now for my son, this was great, he jumped from 71 to around 53 with the Spring BR points...but no one who was situated with a worlds invite got hurt because they didn't play.
Again: the system benefits those who do not play.
Particularly with the early rotation (and I'm a huge fan of it) and the lack of understanding of what decks are going to play well at Nats, the "safe bet" is simply to play the easy rounds and then drop to preserve your points.
The problem is not just with those who earned byes (#1 States, #1/2 Regionals) or travel awards. By the time the first states tournament is completed worlds invites (with the current system) for SOME are pretty much locked in stone. At that point, the top 10-20 players stop playing.
One fix is to lower K values and play more rounds (I know, that's impossible), but doing so allows each individual game to matter less but long term play to matter more. You can get the same effect by raising the K value each round you play (to a maximum maybe?). Round 1 is worth 2 while round 4 is worth 6.
In any case if the system gives you more points for going 3-0/7 than 5-2/7 (3-0 out of 7 rounds vs 5-2 out of 7 rounds) then you are going to encourage people to drop or not play.
If the system allows you to enter a tournament in the top 20 players and leave the tournament still in the top 40 players then you are again encouraging people to drop or not play.
And I have to echo someone in the earlier posts talking about top cut costing points. It's absolutely insane that you can play well enough to attain the top X players in a tournament and go on to play final rounds and then end up HURTING YOURSELF by doing so. Why are top cut battles even K Rated? Top Cut are those playing for prizes...let the prizes be the reason to play top cut, not the K values. K Rate Swiss, not top cut. Here, again, you solve another problem: Players entering a tournament, looking at their match ups, and dropping even though they should be competing for the tournament win just because they don't want to hurt their rating.
Till this gets fixed, for many people it makes SENSE to take your bye's at Nats and not play anything more. You can't penalize people for doing what's best for their worlds invite chances...
If you want the top people in the US REALLY competing at US nationals, you have only 2 options:
1) Force them to compete by impacting their ability to get a worlds invite if they don't. (Extremely high K values, loss of points for not competing, etc.)
2) Protect their worlds invites to give them no reason NOT to compete.
The system IS the problem. You cannot expect someone who has an assured invite to Worlds to go to a tournament and take the chance of losing it just so you can have the top people in the world playing.
---------- Post added 07/06/2011 at 03:52 PM ----------
One more point that I think has been missed:
It's also not JUST about the rating system, it's also about the method by which we pair players (The Swiss System).
The farther along in the tournament the harder it's going to be to win games...that's how Swiss works...
Unfortunately this doesn't take into account the RATING of the person you're going up against, only the deck and matchups (meta) that currently exists at that point in time.
If I'm an 1850 player going into Nats right now with the format shift, and the uncertainty that's out there right now, you BET I'm going to be seriously looking at not playing. Comparatively little is known about the format at this point compared to what we knew last year at this time about what decks will be played and what would likely win.
Since it's ENTIRELY possible in the late rounds (this is the same always, but expecially so right now) that I won't be facing good PLAYERS just good DECKS (players with low ratings that happened to hit on a great deck concept or something that wins well against the meta), the best bet for my rating would be to drop before I have the "opportunity" to lose my shirt.
Rating rounds based upon current RECORD of both players (in that tournament, not overall record) instead of on their PLO rating would at least allow me to play in the 8-0 and not worry that I"m about to lose 32 points to a 1500 ranked player because he's having a good day.
The other point of the PLO system we're ignoring is how Pokemon is designed at it's heart. Pokemon is Rock-Paper-Scissors play. No matter what deck I run, no matter what my skill, no matter how good or bad I am, any player of any rating can beat me if they're playing the right deck...and when they do I'm heavily penalized.
Between the fact that we pair people based upon current tournament record but reward/penalize them based upon overall rating, and the fact that skill cannot completely overcome bad matchups, PLO encourages players simply not to play.
So one of the Masters in our area (he's on these boards) is looking at a worlds invite for the first time ever...and is playing very conservatively to preserve these points.
He went something like 4-0 Drop in states, and was going to do the same thing in Regionals when he made the mistake of deciding to go ahead and play top cut...end result: Rating got hurt when he lost out in top 32 or 16 (forget which). His plan for Nats is to go something like 3-0 DROP (one of those is a bye).
His comment to me earlier this year was "The system benefits those who DO NOT PLAY."
Another example: Spring battle roads. There's a simple rule here. If you have the points to compete for Worlds, don't play it. Spring BR has ALWAYS been horrible attendance for exactly this reason. Now for my son, this was great, he jumped from 71 to around 53 with the Spring BR points...but no one who was situated with a worlds invite got hurt because they didn't play.
Again: the system benefits those who do not play.
Particularly with the early rotation (and I'm a huge fan of it) and the lack of understanding of what decks are going to play well at Nats, the "safe bet" is simply to play the easy rounds and then drop to preserve your points.
The problem is not just with those who earned byes (#1 States, #1/2 Regionals) or travel awards. By the time the first states tournament is completed worlds invites (with the current system) for SOME are pretty much locked in stone. At that point, the top 10-20 players stop playing.
One fix is to lower K values and play more rounds (I know, that's impossible), but doing so allows each individual game to matter less but long term play to matter more. You can get the same effect by raising the K value each round you play (to a maximum maybe?). Round 1 is worth 2 while round 4 is worth 6.
In any case if the system gives you more points for going 3-0/7 than 5-2/7 (3-0 out of 7 rounds vs 5-2 out of 7 rounds) then you are going to encourage people to drop or not play.
If the system allows you to enter a tournament in the top 20 players and leave the tournament still in the top 40 players then you are again encouraging people to drop or not play.
And I have to echo someone in the earlier posts talking about top cut costing points. It's absolutely insane that you can play well enough to attain the top X players in a tournament and go on to play final rounds and then end up HURTING YOURSELF by doing so. Why are top cut battles even K Rated? Top Cut are those playing for prizes...let the prizes be the reason to play top cut, not the K values. K Rate Swiss, not top cut. Here, again, you solve another problem: Players entering a tournament, looking at their match ups, and dropping even though they should be competing for the tournament win just because they don't want to hurt their rating.
Till this gets fixed, for many people it makes SENSE to take your bye's at Nats and not play anything more. You can't penalize people for doing what's best for their worlds invite chances...
If you want the top people in the US REALLY competing at US nationals, you have only 2 options:
1) Force them to compete by impacting their ability to get a worlds invite if they don't. (Extremely high K values, loss of points for not competing, etc.)
2) Protect their worlds invites to give them no reason NOT to compete.
The system IS the problem. You cannot expect someone who has an assured invite to Worlds to go to a tournament and take the chance of losing it just so you can have the top people in the world playing.
---------- Post added 07/06/2011 at 03:52 PM ----------
One more point that I think has been missed:
It's also not JUST about the rating system, it's also about the method by which we pair players (The Swiss System).
The farther along in the tournament the harder it's going to be to win games...that's how Swiss works...
Unfortunately this doesn't take into account the RATING of the person you're going up against, only the deck and matchups (meta) that currently exists at that point in time.
If I'm an 1850 player going into Nats right now with the format shift, and the uncertainty that's out there right now, you BET I'm going to be seriously looking at not playing. Comparatively little is known about the format at this point compared to what we knew last year at this time about what decks will be played and what would likely win.
Since it's ENTIRELY possible in the late rounds (this is the same always, but expecially so right now) that I won't be facing good PLAYERS just good DECKS (players with low ratings that happened to hit on a great deck concept or something that wins well against the meta), the best bet for my rating would be to drop before I have the "opportunity" to lose my shirt.
Rating rounds based upon current RECORD of both players (in that tournament, not overall record) instead of on their PLO rating would at least allow me to play in the 8-0 and not worry that I"m about to lose 32 points to a 1500 ranked player because he's having a good day.
The other point of the PLO system we're ignoring is how Pokemon is designed at it's heart. Pokemon is Rock-Paper-Scissors play. No matter what deck I run, no matter what my skill, no matter how good or bad I am, any player of any rating can beat me if they're playing the right deck...and when they do I'm heavily penalized.
Between the fact that we pair people based upon current tournament record but reward/penalize them based upon overall rating, and the fact that skill cannot completely overcome bad matchups, PLO encourages players simply not to play.
NoPoke
Active Member
except that it doesn't .You have to have played and done well to get to the lofty position of being able to consider sitting out.
@Illydth
<3<3<3<3
Sums it up pretty nicely. I know that I wish I had decided to not play. I lost to a 1580 player in the first round (mewdow with godstart, what can u do?) and played against pointless (HA HA) people all day. In the end I got punished for playing topcut and now im 51st in ranking. I wish so hard I just wouldve dropped, its so frustrating >_>. I basically only played this season because I wanted to play worlds, the format was so stupid, but I wanted to make worlds. And after all the work I end up like this
<3<3<3<3
Sums it up pretty nicely. I know that I wish I had decided to not play. I lost to a 1580 player in the first round (mewdow with godstart, what can u do?) and played against pointless (HA HA) people all day. In the end I got punished for playing topcut and now im 51st in ranking. I wish so hard I just wouldve dropped, its so frustrating >_>. I basically only played this season because I wanted to play worlds, the format was so stupid, but I wanted to make worlds. And after all the work I end up like this
- Status