Pokémon TCG: Sword and Shield—Brilliant Stars

Issues Facing the TCG

Status
Not open for further replies.
Cyrus said:
Ian: either for or against, what's your assessment of a gain-only pro point system that resets every season?
Click to expand...


LOL I can't answer that specific question of for or against. :)

I will say

  • that I do not believe that our present elo system is working well as it stands.
  • that the fundamental flaw in a pro points system is that it expects, believes, even requires that opportunity to play is uniformly distributed.
 
Last edited:
Don't other games that use pro points only allow you to play in a few tournaments for each cycle (albeit having a lot MORE sets of tournaments)?

Ian's right on the money... pro points would greaten the gap between the players who can go to 15 cities and those who go to 4. I think there HAS to be something to lose in the current tournament format. Otherwise, going to a lot of CC's would mean you're bound to make top cuts eventually, and there's really no doubt that you'd want to go to as many CC's as possible.

What if they kept the current system for Swiss, and then additional pro points for top cuts with NO loss of points?

They'd be 2 separate scores... your Ranking would be based on the sum of your ELO score (which must maintain a 1600 average between all players) PLUS extra points for top cuts. It just sorta sucks that choosing to play in top cut gives you a statistically 50/50 chance of LOSING something... you shouldn't be punished for doing better.
 
one of the challenges for a working elo system is that it needs data to process. Without quite a large number of matches its output is far more like a pro-points reward based system than a player rating system
 
How about a system like ELO that then divides the points gained by the amount of matches played in specific tournament sets?
For example, say I played 3 States and gained a total of 90 points. I'd end up gaining 30 points.
If Guy B played 2 States and gained 70 points, he'd gain 35 points.

This would have to done based on each cyclus, to make BR's still attractive. The series that has less tourneys (AKA Regionals/Nats) would still be the big moneyshot if you preform well.
 
elo is already an averaging system. Your proposal would do two things: discourage play if you do well in the first tournament, further increase the emphasis on nationals.
 
To encourage play, and discourage dropping, what about a system where players can only GAIN points in the Top Cut? Basically ELO, except the loser in a Top Cut loses nothing, and the winner still gains the normal amount of points. It seems flawed to make a Top Cut, then lose points for losing Round 1.

To use a sports analogy, would you think less of a team that made the playoffs, then lost first round than a team that simply missed the playoffs? I wouldn't.
 
How about a hybridized system of ELO and ProPoints?

Here's what I mean. ELO can be used for the Swiss rounds only, and ProPoints can be used for the Top Cut. At the end of the season, you use some combination of ELO and PP to determine the players that earn their Ratings Invite.

The main thing that I dislike about ProPoints is that it completely shuts out those that don't live in a large city where P!P is prevalent and readily available. My players have exactly two decently sized tournaments every year within 2 hours of their home League. Everything else, they have to travel for. This isn't because our area is lacking a TO willing to run more events (I'd run something every weekend if I could). It's because the area itself is smaller, and thus garners less attention than the larger cities. ProPoints-only would mean that my players would never be able to earn their invites through ratings. At least with ELO, they have half a shot given consistent high performances at the local events they participate in and the 2-3 big events (States, Regionals, and Nationals) that they can beg into. Another problem is that it's difficult to quickly verify ProPoints without multiple sources of information.

The main benefit with ProPoints is that it is purely a reward system. As such, the players that can attend and perform well at more tournaments have a better chance at getting their Points. "Performing well" in this case means Top Cut qualifications and victories, almost regardless of Swiss Round activity. It would encourage smaller events to become larger to entice a larger Top Cut, and thus more points.

The main problem with ELO as a purely reward system is that it discourages people from competing at tournaments. Not from playing, but from competing. As stated with Fulop's case, it's possible to show up late, get that R1 loss, take 2-3 quick wins from "lower" competition, and then drop to protect the rating.

The main advantage to ELO is that it's Zero-Sum. This makes it very easy to verify whether or not the information is correct. If everybody in the system averages to 1600, the system is working correctly. It also encourages players to compete at higher level events (though at the expense of lower level events).

If there is some sort of hybridized system between ProPoints and ELO that maintains the P!P standard of all events (except Worlds) being open to the public, I would jump on that in an instant. But, as an either/or type of deal, I much prefer ELO to ProPoints. I just feel that ELO provides fewer downsides than ProPoints.
 
Ness said:
To encourage play, and discourage dropping, what about a system where players can only GAIN points in the Top Cut? Basically ELO, except the loser in a Top Cut loses nothing, and the winner still gains the normal amount of points. It seems flawed to make a Top Cut, then lose points for losing Round 1.
Click to expand...

I think that's a cool idea! My only concern is in tournaments like Nationals where the possibility of Top 64/32/16/8/4/2 means 6 straight rounds of point gain. Of course, if you make it to top 4 you get an invite anyways (or is it top 8?). It would be super important to attend Nationals as it would end up being the stairway to Worlds more often than not with it's zero downsides once you hit top cut.

Throwing this idea out. Perhaps only top cut games can count for full points and swiss rounds count as 1/2 points. If you go to Nationals and lose the first three rounds and drop, it doesn't sting as much. If you go to Cities and go X-0 drop before Top Cut, you don't get as many points, as we'd like to promote players playing the entirety of the event and not just drop once they "get their points". Also, it would give incentives for making the top cut since you get more points for doing well in the top cut so consistent top cutters will have an advantage over those who X-0 drop in swiss.
 
Elo really isnt. The guy who gets to play like 15 CC's will always have a sick advantage to the guy who can only play 4.
 
Yoyo and NoPoke: some of the things specific to other games that work for them may not work for us, and so that might require adjustment. As I said from the outset, P!P would have to decide on the specifics, as the buck ultimately stops with them.

The gain-in-top cut-only suggestion by Ness is more or less similar to what I have, only it allows for some added movement. I've proposed the same idea in the past, and figured out that while its advantage over a pro point system is to award people for playing out the swiss, there are still many of the leftover incentives for a player to drop (e.g., drops to preserve rating). But If ELO is to be kept for next season, then I definitely support a switch like this - we "don't" need a 1600 average for ratings as much as we need a system that accurately awards 5th for 5th, 4th for 4th, 3rd for 3rd, etc. Pro points would do that, but so would this.

For those arguing that pro points automatically shut out people who live in small areas/cities...ELO already does that. The fact of the matter is that if you're in an an isolated area (certain states in the Heartland, the panhandle of Texas, New Mexico, and maybe Albany come to mind), you _still_ have to travel far and wide for even a shot at an invite. I don't think that this would change for the worse under a pro point system, but if you guys have evidence from other games to suggest otherwise, please show me.
 
Last edited:
So how do you reward fifth for fifth? That is a simple question that very quickly opens up a whole different can of [del]worms[/del] difficulties. Attendance matters, venue time might constraint the cut, the difference between fifth and fourth or sixth place might be in op-op-win% and not in match record. elo avoids all those pitfalls.

elo could be adjusted to account for our game that isn't pure skill. (A bayesian approach) A beneficial side effect of that approach is that higher rated players don't suffer the way they do right now when the inevitable loss occurs to a much lower rated player.

match play as standard helps. draws too.

Using provisional ratings acceleration helps deal with the reset artificially treating all players as identical. increasing the number of games considered by extending the rating year to every two years could be done once compromises associated with age ups had been settled.

There is a great deal of mileage left in elo. But I don't expect any of the above to actually happen, mostly because I think we would have seen it before now if it was going to happen.
 
Scipio said:
Elo really isnt. The guy who gets to play like 15 CC's will always have a sick advantage to the guy who can only play 4.
Click to expand...

Only if they win. If they don't, there is no advantage (and there could be significant disadvantage)
 
What's the bayesian approach? You might have described it and how it could be useful in pokemon somewhere else on this board, if you have can you link to it?
 
using 'pro points' for top-cutting events as the main determination for worlds invites sounds like an open invitation for travelling players to attend historically 'weaker' or lower-attended area events in order to take those points from local players...even moreso than 'just' the currently available ELO points.

jmho
'mom
 
^ Which is probably why several folks have suggested using a combo of both ELO and Pro Points.

I'm totally on board with the idea of ELO for Swiss and Pro for TC.

... and honestly I would rather get rid of people dropping or not playing in events like Nats to get World Invites than worry about strong players spending money to visit weaker areas.
 
Diaz said:
What's the bayesian approach? You might have described it and how it could be useful in pokemon somewhere else on this board, if you have can you link to it?
Click to expand...

I can't remember the exact post, but it's basically a luck-modifier for ELO that can temper the amount of points gained/lost between players with a wide point differential. NoPoke has the simplest version of the Bayesian Modifier to ELO that I've ever seen, but I can't remember where that post exists...
 
NoPoke, the "it would've happened" argument doesn't seem to fly simply because P!P's been willing to try new things since they took over from WotC. From '03-'06, ratings effectively had no value, bar some trip-less invites done via rankings (and even those weren't so significant, since invites were very inclusive back then). And since the first Hawaii season, the significance of ratings has been mixed up every year. In short, I think P!P is very willing to try something they haven't before - especially something that could stand to correct one of the worst flaws in tournaments today.

IMO, ELO cannot hope to correct endemic dropping/tournament manipulation we have in Pokemon right now, so you either have to have: A ) the reformation or elimination of ELO; or B ) the reformation of the prize/invite structure. One of these needs to be set in place by the beginning of next season, or else we'll just continue with this endless precession of X-0 drops, X-1 drops, and showing up to tournaments late to soften matchups.

Since reformation of the prize/invite structure is unrealistic/a shaky proposition, I say we just scrap the system that encourages top players to punt in order to win invites. Maybe this could be achieved just fine with pro points + ELO, but I would have ELO be secondary.

I'd love to hear a P!P person's perspective on the advantages/disadvantages of such a system, but I'm not sure if they'd be able to speak in a capacity.

P.S. like V_G, I don't see a problem with a system that'd lead to an occasional metagame curb stomp if it meant the elimination of dropping for incentive. At least someone is earning it by playing out the event (i.e., William Hung laying the smack-down on Iowa all those years ago for a Gym Challenge). Besides, a pro point system could allow for attendance to factor in, so that point might even be moot.

P.S.S. Austino's "concession" point is largely moot for the same reasons listed above. Plus, it's extremely conditional to even pull that off at high tier events like States/Regionals/Nationals.
 
Last edited:
My 'it would have happened' statement is not a criticism but a recognition that for change to happen you need more than just a desire or willingness to change. You have to also have available capacity to take on the change. Capacity to deal with the complications and misunderstandings.

Right now I think all of P!Ps spare capacity is allocated so adjustments/tweaks to elo are just no longer on the todo list.

----------

elo is a play at risk system, pro points is a play without risk system. There are always problems when players have nothing to lose, just as when they have too much to lose.

dropping is not of itself a problem. Dropping as a response to how tournaments are structured is. There should be no need to add further incentive to players than the reasons that they play in the first place: fun, friends, competition, the oportunity to win or improve. I don't accept any arguement that the top players need additional incentives.
 
Last edited:
I still prefer the old point system:

3 = Win
1 = Tie
0 = Loss

This encourages players to play through the tournament and, in a sense, discourages dropping as you lose nothing. This also reduces the Stall/Slow Play situation as Ties are reinstated. Now, the only thing with Ties, it allows for more collusion and gaming of the system. But, it can also push those close matches.

Pro Points are interesting, however so many players enjoy seeing their Ranking, regardless of how close to 1600 it is.

I still prefer the old methods of States - Nationals Trip and Regionals / Nationals = Worlds Trip. Especially since they change the Dating system to where it is difficult for players to attend multiple Regionals.
 
The more I think about it, the better keeping ELO in some capacity seems worthwhile. Fulop's allusion to the "X number of points for ratings 1800 or above," "Y number for ratings 1900 or above," etc seems like a reasonable way to keep the system - however flawed it may be - relevant. Plus, Fish is right to say that people enjoy seeing their ratings/rankings fluctuate.

However, I'm tired of a system that puts EVERY tournament - not just some - at risk for illegitimacy due to incentive drops. A pro point-focused system would change that, and while an imperfect build of it may result in a little more collusion than had already existed, I'd rather see a once-in-blue-moon incident that isn't encouraged by the system to succeed (collusion) than a regular occurrence that is encouraged (dropping to preserve rating).

RE the old point system: I think draws should return, but if they do, then they need to accurately reflect the real ELO. By that, I mean draws should count as a loss in points for the higher-rated player - an aspect that wasn't present in the 2003-2004 season.
 
Last edited:
Status
Not open for further replies.
Back
Top