Season Four Possibilities

[Thomas David Baker]

I'd like to kick off an Extended constructed season (+ possibly EDH or something similar alongside it).

But we have to solve this problem:

- Bob loses his first three matches, including one to Jim
- Jim is 3 points ahead of Sue who has not played Bob
- Bob is sick of his deck, has no hope, and won't play Sue

I thought about various things:

1. Forget the "pick up games" style of previous leagues and run a knockout tournament with explicit pairings at each round.
2. Refuse entry to the next season to those that don't play at least X% of their games.
3. Introduce "divisions" and relegate those that don't play enough matches at the end of each season.
4. Introduce a "formal challenge" where you can give three Games Club nights to your opponent and if they don't play you on one of those you get the points (this is pretty much a jazzed up version of allowing concessions).
5. Allow deck changes during a season. Either X cards or change-your-sideboard or one or more complete changes of deck

But I didn't really like any of them. Maybe (5) but if you are 0-3 or 0-4 before you think "hmm I need to change my deck" the incentive to battle with your new deck is fairly low.

The attributes I want to preserve:

- It's relatively easy to get a game.
- There is a ranking of players at the end of the tournament that produces 1st, 2nd, 3rd at least.
- All players are encouraged to play matches not hide from them.

This got me looking at "ladder" competitions.

The classic ladder is one where the winner swaps place with the loser. But there is no incentive in this system to play if you are the player higher up the table.

People have thought about this before.

Leapfrog System

Two players play. If the higher-ranked player wins, they get 3 points. If the lower-ranked player wins their points total becomes the higher-ranked player's total + 3. The loser gets 2 points if they won a game, or 1 point otherwise.

You can only "leapfrog" once per day. At the end of each day your best possible leapfrog is applied plus any other points earned.

The potential benefits here are:

- Lots of matches will be played (playing and losing is better than not playing).
- But a single win can put you top of the league (its never hopeless).
- But you cannot rest on your laurels as you will surely be overtaken.
- Players can potentially be allowed to play each other more than once (although there will have to be a limit otherwise two players could just team up and play each other off into the stratosphere and then refuse to play anyone else).

A complete worked example using the last constructed league

This example is bogus because instead of games drying up at the end, those in 2nd, 3rd, 4th would be trying to play anyone they could to try and catch 1st place while 1st place waits to see if anyone will catch him and then scrabbling around for a game once it has happened. But of course if any of these near-winners loses one of these matches this puts someone else right back in it. Exciting!

But, also, possibly, flawed. Is the best strategy to do nothing then play high-ranked players on the very last day of the tournament? Is it impossible for you to win the tournament if you don't play on the last day? Its certainly impossible for you to win the tournament if you don't win your last match, which is quite swingy. Mills finishes 6th in the example with his 7-1 record because he didn't play much at the death. But perhaps that is OK? We want to encourage playing, after all. And he would have had more potential opponents if we allow playing each other player once a week or something similar.

Perhaps we can mitigate this by limiting matches per week, matches per day, or something similar. Perhaps it doesn't matter because it's exciting and if you miss out there's always next season.

I also think that there's something to be said for Steve's suggestion of a knockout between the highest 4 ranked players (1 v 4, 2 v 3, then a final between the winners) regardless of how we generate the rankings.

Perhaps the secret is to have a completely open-ended leapfrog system for every format going. "Hey Bob, want to play an Extended ladder game?", "Sure, and let's play a Vintage ladder game after", "Great". There are monthly winners (top on the 1st of the month) or a monthly Top 4 play off and possibly prizes for that, but the ladder keeps rolling on forever.

There's the germ of something very good in the leapfrog system in that it solves a lot of our problems. I'm interested in what you guys think before I decide on a final format. Can you spot fatal flaws? Are you completely turned off by the extra complexity of the leapfrog system? Have you got a great insight that will improve any of these ideas? Does any of this solve the problem of hating your deck after 2 matches? Whaddyathink?

[Simon O'Keeffe]

Just to confirm I understand the leapfrog. If I play only one game total, but it is on the last day, and against the current leader, then I could win the league from 1 match?

Maybe keep points from playing separate from leapfrog points.

e.g.

Bob has played 9 and won 8 = 24 + 9 game count points = 33
Charlie played only bob and won = 24 + 3 + 1 = 28

[Ben Titmarsh]

I think we're in danger of over complicating things. I know concessions are a contentious issue, but why don't we just say they are allowed from the beginning to avoid all of this. In reality only 2/3 people are going to be running around getting concessions, and in theory the player with the most actual wins should still take it down.

If someone hates their deck (in reality this shouldn't happen all that often, post rotation constructed doesn't come around regularly enough), just let them concede.

I understand where the animosity was coming from with regard to concessions in the last league, but if a clear statement is made from the beginning then people who feel strongly enough (anyone?) can abstain.

[Thomas David Baker]

@Simon - you are understanding the system correctly. I guess if the table was:

Bob 50
Jim 48
Sue 0

Then Jim might agree to play Sue in an attempt to pip Bob, lose 0-2 and give Sue the title by one point. In practice I imagine it wouldn't happen like this. But if it did, that would be quite an exciting game! We could display Wins and Losses in the table, but ultimately we forego a strict "who won the most" system for a wilder, possibly more exciting system that copes with some of the constraints we are dealing with.

I think it might be worth a try one time at least. We already know a strict league system is not perfect for the situation.

[Thomas David Baker]

@Ben - Using the three seasons we've played so far if we allowed concessions at least two, and probably all three, would definitely have been decided by who could arrange the most concessions in the last week or so. I don't think there's any point running leagues where that is the case. There's definitely a problem here that needs to be solved.

These leagues differ from "normal" leagues in that they are guaranteed to end unfinished. We could require matches to be played at certain times and places but then it doesn't fulfill the round-the-edge, as-and-when, pickup-game style that was one of our core requirements.

In Season One on the final day we had Mills pressured into playing Paul Y (with an illegal deck) when he really didn't want to so that Paul could go top, then Paul refusing to play Steve so he could stay on top (he did relent and do the sportsmanlike thing in the end - and won), then Dan not being interested in playing Steve because Dan had no chance of winning the competition. In the end Dan agreed to play Steve although his heart wasn't in it. I let all the dubious results stand and declared the whole thing a draw because it was easier than having to pick sides. But it left a bad taste in the mouth.

I guess it would have been better (by some definition) for all those dubious last day games to be concessions instead. But who would have won the league? Whoever out of Steve, Dan and Paul could arrange the most concessions. That's not a great outcome.

Season Two was solved for us by Tom Reeve kicking everyone's ass conclusively although if Paul L had arranged for concessions from those he didn't manage to play he actually would have nicked it.

And we just had our kerfuffle over Season Three. Rob had stopped carrying his deck around, had no chance of winning and didn't want to play Warren. If we allow concessions Warren gets the title from Roque because Rob can't be bothered. If we don't Warren is cheated of the title because he doesn't get to play one of his matches. Neither of these outcomes are great.

With legal concessions the best tactic in at least two of the preceding three seasons would have been to (a) be popular, (b) have everyone's phone number, (c) play no games for the first couple of weeks, then (d) get everyone who was out of the running to concede to you, and then (e) play out any matches that you really have to, unless they are against rivals in which case refuse to play.

If everyone was an upstanding Ben Titmarsh of a character then the fact that optimal behavior makes the competition crap would be one thing. But I'm afraid to say there are plenty of shady characters, factional rivalries and other human nonsense going on at the Games Club. I think we need a system that incentivizes players to do the right thing rather than just relying on their honor and judgment.

One simpler possibility is to say that at the end of the season everyone below you that you haven't played automatically concedes to you. Early on this works because you want to play to get points and get higher up the table. But as soon as you get a few wins and have a decent gap above another player you will avoid playing them because you're going to get an auto-win anyway unless they find some matches elsewhere. Meanwhile those who have already won a few but also lost a few will lose interest and be harder to pin down for a game by those trying to catch up that haven't played them. This is fairer than allowing concessions (no popularity contest) but doesn't exactly encourage a fun competition.

[Thomas David Baker]

One option would be to make the leapfrogging less dramatic.

We could have a maximum leap size (equal to your current points, equal to half the leaders current points, 10 points, or derived some other way).

Or we could say that the potential leap is to (opponent's points - 1) rather than (opponent's points + 3) so that you can never give away the lead by playing a distant rival. I quite like this last one.

[Tom Reeve]

So what you're really saying is, you want me to play more so I can remove the need for shadiness by kicking everyone's ass?

>_>

[stevie_b]

Quite a pickle we have here folks.

First of I think tom, by trying to find a 100% infalliable system u r missing the point that this is a £2 casual, let's play more constructed idea. The actual league table is simply a way to formalise those casual matches.

Having said that u r dealing with trying to solve an intractable moral hazard dilemma - nobel prizes are awarded for less.

Swapping decks mid league seems unfair on the players played before or after the swap - once there is an established meta this should not be a problem - if we had standard now there would be jund, Baneslayer x, boros bush, nissa eldrazi, Gary rubbish and hence less randomness.

I am not a fan of the leapfrog system - it seems overly complex, swingy and open to abuse of sandbagging wins at the end.

Although the Paul young case and warren concessions were problems they were actually 'solved' by peer pressure and naming and shaming on this forum.

Finally intoducing a playoff would mean that u would have to actually win some games to win the season not just acculmulate consessions.

[Glenn Goldsworthy]

what about the top 4 at the end of the season go into a knockout playoff (1st plays 4th and 2nd plays 3rd with the winners playing for the title)

this would get rid of the need for shenans as everyone in the top 4 has something to play for, the only issue would be if there were 2 players on the final day who could make top 4 and 1 gets soemone to concede but thats not really "in the spirit of a £2 buy in touney with mates"

Also EDH is where it's at people need to get a league sorted for games of 4 people and a points system for achieving nothing through 15 card combos

[Thomas David Baker]

I've started an EDH league with a Top 4 play off as if anything can be played in good humor and without controversy that's probably it. There are no concessions there because the points aren't even for winning but rather for doing daft stuff. I still think we can do better for the more competitive league. When we come up with something I'll kick off an Extended season (I won't wait for the EDH season to end).

[Charlie Grover]

Play offs should be best of 5.