Elements PvP League - Suggestions and Feedback

[bucky1andonly]

in which the best player has a chance to lose to the worst player

if the best player refuses to play, then the worst player has 0% chance to beat the best player

ELO does not require games to have no luck.

maybe so, but chess requires skill more than luck
also ELO works best when players cannot pick and choose their opponent

I'm not sure you completely understand ELO, bucky.
As more games are played, a player's rating becomes further from 1500 until an equilibrium is reached. For the best players in the league, it takes a large amount of games to reach their equilibrium. This means that playing a lot of games is beneficial to the best players in the league. Perhaps players in CL are avoiding you because of reasons unrelated to your rating?

I understand just fine thanks.  There will never be equilibrium for any player and that is due to league play being 2 months (extended to 3 this time).  If the top player stops playing after reaching the minimum amount of games required, and they have a decent enough amount of points, they dont need to continue to play, but this means the people that MUST continue to duel risk losing points.  So in other words, people that lose often should just quit the league if they have no chance to catch up to the top player.  If the top player continued to play like everyone else MUST do, then at least they would have a chance at losing points.  If they continue to play and win, good for them, then they deserve the top spot.  But to get to the top and stop playing for 2 months, no, they not only dont deserve the top spot, they dont deserve to be in the league play.

[Jen-i]

Here's the hypothetical situation.

The theory behind ELO is that we are using an estimation of a players actual skill to determine ranking.

Player A could get lucky and win his first 30 matches - giving him let's say a score of 1700 - this score is a horrible inflation of the player's actual skill - they have just gotten horribly lucky. If Player A now chooses to stop playing matches their ranking will be based on a score that does not reflect their actual skill.

Player B got very unlucky and lost their first 20 matches - if Player B's actual skill is say 1750 it will take a significant number of games before their actual rating matches their actual skill - but given enough games they will reach this equilibrium. If however Player B's actual skill is 1600 it will take a significant lucky streak to capture 1st place.

ELO works best the more games you play - because statistically the more games you play the smaller the error in calculating your actual skill.

Bucky - you are correct that Player A has an artificially inflated score - however their is no way we can know any player's actual skill - but you are also right the more games we require players to play the better picture we get of what their actual skill is.

[bucky1andonly]

But the amount of actual skill needed for a game of elements is very little for majority of decks.  It comes down to skill of making decks, which doesn't say much considering how many people use the exact same deck.  ELO again, is best over a long period of time, there is no way to accurately determine someones actual skill level (luck based skill).  And to get more people active during league if they sign up, weekly minimum games would help and then penalize those that fail to reach the amount.

[RootRanger]

But the amount of actual skill needed for a game of elements is very little for majority of decks.  It comes down to skill of making decks, which doesn't say much considering how many people use the exact same deck.

Deckbuilding is skill. Success in highly competitive play requires original decks due to the league metagame and the inadequacy of the common decks. I don't see how you expect to win by using the same decks as the rest of the league.

[RootRanger]

I have two suggestions, and they are both related to something that I think the League Rules overlook: CL doesn't require any more skill than BL.

First, I think the Leagues should be called Upgraded League and Unupgraded League instead of Champion's League and Beginner's League. I think it is unreasonable to call some of the best unupgraded players in the game "Beginners". I also don't think it makes sense to call someone a "Champion" just because they prefer the upgraded metagame.

I also think that the prizes for CL and BL should be the same. Better players deserve better prizes, and the best CL players are not better than the best BL players.

Both of these changes make sense with CL players being no better than BL players, and I find that true. I think BL takes as much skill, if not more, as CL. Success in CL requires a lot of electrum and/or upgraded cards. Many of the highly skilled players do not have enough upgraded cards to succeed in CL. BL is different, however. Almost all of the highly skilled players, however, have enough cards and electrum to have the resources they need for BL. This makes BL open for a significantly larger amount of skilled players, which means the top players are definitely as good as the best players in CL. I also think BL involves as much, if not more skill, as CL, because success in BL requires creative deckbuilding skill. Here (http://elementscommunity.org/forum/index.php/topic,11028.0.html) is the unupgraded PvP deck archive. There are not many decks, and all of them, including mine, are far from being perfect. I wouldn't use any of them by themself, and I would only use a couple of them if I heavily modified them. Here (http://elementscommunity.org/forum/index.php/topic,11023.0.html) is the upgraded PvP deck archive. There are 20 decks, and some of them would do very well in a competitive environment. Although I have never entered CL, I have entered many upgraded tournaments, and I know what will work well. Success in BL requires as much deckbuilding skill as CL, and likely more.

I hope I've made it clear that the winners of BL are just as skilled as those of CL. I find the names and the prize gap unreasonable because of this.

[Jen-i]

I completely endorse and agree with your conclusions RootRanger.

I'd argue with you about your assumption that CL doesn't require as much deck building skill (but as neither of us have experience in both leagues that's just our opinion) - I still agree with your conclusions.

BL & CL should be two options we have for league play - not 2 grades of players

[bucky1andonly]

Still needs to be penalties for inactivity and rewards for activity.

Like I had said, tttt was 1st in CL, now 2nd, but hasn't played in almost a month.  So in that time, he had no chance at all to lose any games, effectively lowering his score.  And being at that rank, with his score, playing and winning against him is better than playing the person in last, because he loses score and you gain score, closing the gap faster.

And I don't agree with BL and CL having the same rewards.  If you spent all that time getting the upgraded cards and playing upgraded generally is more difficult, you deserve more.  If someone with all the upgraded cards wants to play in BL because they can pick off newbs, that is their choice, and I do feel that is why a lot of the pros play BL even if they won't admit it.

What the leagues are called, I have no opinion, I kinda like making fun of the pros though that enter BL since it is called beginners but that's just me.

[pikachufan2164]

I've had the opportunity to talk with a few League players in chat tonight, and we've explored some options for promoting consistent activity in the Leagues. Small activity bonuses are probably the best option if we want to go that route.

A sample proposal:

• Split the League season up into six 2-week phases.
• During each phase, players gain a bonus of x rating points for every set of y matches completed.
• Activity bonuses may only be earned z times each phase to discourage match spamming and encourage activity throughout the season.
• To simplify scoring, activity bonuses will only be added to players' ratings at the end of the League season.

A sample bonus list:

Phase 1: Earn 3 bonus rating points for completing 5 matches. Bonus may be earned up to 2 times this phase.
Phase 2: Earn 4 bonus rating points for completing 5 matches. Bonus may be earned up to 2 times this phase.
Phase 3: Earn 5 bonus rating points for completing 5 matches. Bonus may be earned up to 2 times this phase.
Phase 4: Earn 5 bonus rating points for completing 5 matches. Bonus may be earned up to 2 times this phase.
Phase 5: Earn 4 bonus rating points for completing 5 matches. Bonus may be earned up to 2 times this phase.
Phase 6: Earn 3 bonus rating points for completing 5 matches. Bonus may be earned up to 2 times this phase.

Bonuses need to be small but still worth getting, since the point differentials between the top ranks can be within quite close unless there's a commanding lead (for example, the difference between 3rd and 5th place in BL is only 6 points).

A cap on the number of times a bonus can be claimed in each phase discourages match spamming, and more points are given during the middle phases in order to encourage activity (since there's a rush to play matches near the beginning and the end of the season).

Since the minimum requirement for prizes and avoiding a ban is 30 matches in the season, the bonuses encourage players to pace themselves a little bit so that they can claim all those nice bonuses.

Of course, all of this is still an idea, so feedback is much appreciated.

[10:12:12 PM] BatCountry: You're a LO pika, right?
[10:12:47 PM] pikachufan2164: Yes, Bat
[10:13:09 PM] BatCountry: What are your views on people playing enough league games to rank and stop playing
[10:13:29 PM] BatCountry: Bucky or somebody was complaining about tttt and I was wondering what the deal with that was
[10:14:11 PM] Revoltion: He's complaining that the shortest route to victory was blocked
[10:14:29 PM] pikachufan2164: I would say that we want to promote activity, but it will be an absolute mess if we try to enforce inactivity penalties
[10:14:31 PM] Revoltion: Its a valid point, at least in my eyes.
[10:15:41 PM] pikachufan2164: Well, since TStar already explained that the score differential (K-value) isn't really all too drastic even with a huge difference in rating
[10:15:57 PM] Jen-i: I think one solution is to give activity based bonuses
[10:16:31 PM] Revoltion: A set number of games per month should be easy enough :/
[10:17:40 PM] pikachufan2164: But when we try to do inactivity penalties, we either have to be draconian and go "no excuses,"
[10:17:53 PM] pikachufan2164: Or try to come up with the answer to "what is a good excuse?"
[10:18:36 PM] pikachufan2164: Either way, we'll end up having people complain about how the penalty is unfair
[10:19:20 PM] Revoltion: If the time frame is large enough, then no one should be able to complain. Fair warning at the beginning of the leagues, and if you don't think you can do it then don't sign up/expect penalties.
[10:20:17 PM] pikachufan2164: Though Jen's idea of a small activity bonus might work
[10:21:00 PM] pikachufan2164: Since Leagues will be 3 months long, we could divide it into six 2-week phases
[10:21:37 PM] Revoltion: What defines small? A free win against one of the lower ranking opponents?
[10:21:44 PM] Revoltion: (In terms of points)
[10:21:50 PM] pikachufan2164: Each phase gives a scaling amount of activity bonus (x bonus rating points for every y wins)
[10:22:24 PM] pikachufan2164: Middle phases pay out slightly better activity bonuses than early and late phases
[10:22:28 PM] Revoltion: Wins?
[10:22:39 PM] pikachufan2164: (To compensate for the lull in activity in the middle of the season)
[10:22:55 PM] pikachufan2164: I was thinking of raw rating points as the reward
[10:23:19 PM] Jen-i: yeah - that works well
[10:24:23 PM] pikachufan2164: Or maybe x bonus rating per y games played
[10:24:28 PM] pikachufan2164: Not sure
[10:24:47 PM] Revoltion: Bonuses shouldn't be given for wins.
[10:24:51 PM] Jen-i: games played might be better
[10:25:06 PM] Revoltion: That'll have the same effect as the rich getting richer.
[10:25:07 PM] pikachufan2164: True
[10:25:48 PM] pikachufan2164: Should there be a cap on how many bonuses per phase you can pick up?
[10:26:02 PM] Revoltion: Yeah.
[10:26:14 PM] Jen-i: yes or people like me get too many
[10:26:14 PM] pikachufan2164: I think there should be, to discourage match spamming
[10:26:32 PM] Revoltion: Spamming games shouldn't be able to nullify losses.

[bucky1andonly]

Still doesn't do anything to those that can win a whole lot in their first 30 games and get to the top and quit playing until the leagues over.  If they don't play, they have no chance that their score will ever decrease.  If people complain about an inactivity penalty, then maybe they should be more active.

[TStar]

I'm definitely not a fan of giving bonus points just for playing.  Considering how close some of the leagues have been in the past I'd strongly prefer that the winner be determined through actual results and not some arbitrary bonus point system.  As it is, the easiest solution is to simply stop complaining and win.  If you win your matches it won't matter what anyone else does.  Trying to force activity is just a way of saying you are tying to rely on someone else losing to gain rank, instead of just winning yourself.

[bucky1andonly]

Take Jen-i for example, almost 3x the games played as tttt, shes won more than double the games he has and is still down 16 points, and a couple of losses will only make it that much harder to catch up with so little time left.  And the point is, Jen-i has to play to catch up to tttt, so this means she risks losing points, while tttt has no chance of losing any points at all and hasn't for almost a month.  By all things considered, Jen-i should be in 2nd and tttt should not even place.  And this is coming from a guy that has no chance of getting 1st, 2nd or 3rd, so it is pretty unbiased.  People need to play regularly, otherwise, wins and losses mean nothing. 

[Jen-i]

Here's my complaint TStar

Here are the "rules" to best gain score in the ELO system as we are using it:

1) Challenge all players where the difference between your score is inflated
2) Refuse all players where the difference between your score is deflated

Basically this means if we assume your actual skill is 1500 and your opponents actual skill is 1600 - you should challenge them anytime the difference between your scores is equal to or greater than 100 - and you should avoid playing them (either by not challenging or by refusing them anytime it isn't). The same thing can be said if your opponents actual score is 1400 - any time the gap is equal to or less than 100 you should be challenging them.

Following these "rules" is the way to get the best score gain out of the ELO system - however there are 2 problems with this.

1) Following these rules means no challenge will ever be accepted (when the gap between our scores is enough that I should challenge - it is then at that point where you should refuse) - Challenges will only be accepted as a mistake - because someone overestimates or underestimates their own or their opponent's actual skill. It can result in some meta-game play to artificially deflate your score - such as entering the league late. It also means that if you believe your own score to be inflated (that is your current score is greater than your actual score) you should stop playing games almost entirely - the only people you play are people who's score is inflated more than your own.

2) The league rules technically make following this system illegal - but it is impossible to enforce the rule. Refusing matches is not illegal - only refusing them with the intention of protecting your score - the only way that becomes an enforceable rule is the honor system - which really means it is not an enforceable rule.

I'm not sure what the answer is - but having a system that promotes one sort of game play and an unenforceable rule that attempts to stop it seems dumb.