Possible new way for Tie Breakers

[Shad0w]

Remember that the goal of MW is to kill the other mage. Why would the tie breaker of choice not favor aggressive play?

To me, it's almost like you're saying that THE way to play Mage Wars is by being aggressive and other ways are "wrong". That can also turn some people off of the game.

Not that I have a better idea for tie breakers though.

I am a control combo player by nature. Just because a build is heavy on control cards does not mean it can not be aggressive. In fact I am working on a combo control build that wins on turn 7-9

[Shad0w]

Being aggressive is the only way to play Mage Wars. If you never attack or damage your opponent in any way, you will lose every time. If you do damage your opponent, then you are by definition being aggressive :-P

Zuberi, by aggressive I was referring to "aggro" books instead of say, control.

Ah. We have different interpretations of what Shad0w meant then. I don't think he was trying to imply that any style of play is wrong or inferior to any other. Simply alluding to the nature of the game in general principal, ie killing your opponent faster than they kill you, as something that should be rewarded rather than trying to change the nature of the game on some fundamental level.

Correct
As an example look at the DoT builds. Even if it is a slow build up they are always moving towards the win

[ringkichard]

Remember that the goal of MW is to kill the other mage. Why would the tie breaker of choice not favor aggressive play?

The goal of Magic the Gathering is to kill the other guy/gal too. That doesn't mean that Psychatog + Upheaval was a bad deck. Or that Blue Stasis decks were an error. Certianty is a far greater vitrue than speed.

The agressive book that deals 10 quick damage and then stalls out should not get the victory over the control book that stabilizes and then tightens like a ratchet for 10 rounds. That control book is winning.

The current tiebreaker method favors short term plans over long term ones, and that's the problem.

[Zuberi]

The current tie breaker doesn't favor any plan. It simply measures how close one is to the already existent goal of killing their opponent when time is called. The time limit, however, may favor one plan over another, but there is little we can do about that. Should we adopt an unbalanced tie breaker to try and fix the imbalance caused by the time limit? Perhaps, but I'm not sold on the idea quite yet.

I would argue that a control book which hasn't managed to do any damage to their opponent over 10 rounds while they themselves have been damaged, is not winning. No matter how much control they've exerted. A control book that's managed less damage than their opponent, even by a single point, also is not winning currently. It is possible that they have recently been approaching the finish line at a faster rate than their opponent, but they are still further away from it and thus not winning. Any claims to the contrary attempt to predict the future which can not be done with certitude.

[ringkichard]

Unless you think that accumulated damage is always more likely to give you a win than than board position, awarding the win based on damage done means that the tiebreaker will be awarding the game to the mage who opened with attack spells instead of the mage casting creatures, even when the attack spell mage is losing (defined as having less than a 50% chance of victory).

This gives a stalled agro book incentive to avoid combat and wait for the tiebreaker to salvage an impossible game.

It, for example, rewards liberal use of Wizard's Tower. It doesn't matter if the Attack spells would be enough to win: the defender has to play as if they're near death even if they're not.

Or consider Sardonyx. A great creature you'd never use near the ent of time because suddenly life left  NOW is the only thing that matters.

I'll say more when I'm not on my phone.

[ringkichard]

Ok, at a keyboard, finally.

I think there are two definitions of "winning" in this thread and the other, and the differences between them are causing miscommunication.

It's been said that whichever mage is furthest ahead in the damage race is "closest to winning" and should therefore win a tiebreaker. But if you define "closest to winning" like that, it no longer means "most likely to win". You create a whole class of possible games where Alice is most likely to win -- but Bob is "closest to winning". This is a problem.

Imagine a game where Alice has 20 life left, and Bob has 25. We might say that Bob is closest to winning. But if Alice has 3 Dire Wolves and Bob's got nothing, then Alice is much more likely to win. Alice is winning. And any measure that says that Bob is closest is measuring the wrong type of distance.

This creates perverse incentives for the players. Bob may realize that if he continues to press the attack he's doomed, but can try to flee and avoid conflict and somehow salvage a tiebreaker win from a game he could otherwise only lose.

Or remember that article Piousflea wrote about how important it is to get your opponent down to a low level of life? His reasoning was that once they're at risk of dying next turn they have to change their strategy and must make short term plays instead of long term plans. This is because only the last life point matters, and until your opponent is in a position to take that last life point, you don't actually have to make sacrifices to save your hide.

Well, if we award the game to the player with the most life, we're suddenly making all the life points matter, not just the final one. That 1 damage you took on round 3 is now very important, not just because it made death somewhat more likely, but because it's a victory point of equal value to all other victory points.

--

So, how does the current tiebreaker system affect the current meta?

It rewards mages with access to frequent and large attack spells. 4 consecutive Hurl Boulders might do maybe 20 damage to an opponent with some armor, but they won't kill him or her. And in a long game, the player who invests in some armor and a belt of Regrowth and some creatures will win against an All-Boulder-All-The-Time strategy. But for most of the game the creature mage will have less health than the Boulder mage, even if the creature mage almost immediately has a much higher chance of winning.

This effect makes Wizards Tower even better, because Wizard's Tower's weakness is that you have to spend mana on Attack spells when you'd usually be better off investing it in a creature or conjuration. But now, since whomever is at lower life dies when time is called, short term tactical play beats long term strategic.

Or consider Sardonyx. He's big and mean and can probably help you win the game, but he's costs you 2 life a turn. Which, if you've got 20 life remaining isn't a big deal. But now, suddenly when time is called, you don't have 20 life remaining, you have the difference between your life and your opponent's life. That 2 life loss a turn goes from being 10% of your life to 50% of your life in a real hurry, and now Sardonyx is a bad investment.

Frankly, if the old Tiebreaker didn't help Wizard so much, I think I'd be more ok with it. Yes, it looks like it helps Warlock and Beastmaster, but it's Wizard who's best at crunch time nova damage because of an elemental training and Wizard's Tower.  I really don't think we need to be making Wizard even stronger.

[Zuberi]

You present a very good argument, ringkichard, but it is late and so I shall postpone a full response until the morrow. I just wanted to mention that I don't think that there has been a miscommunication. I fully agree that being closest to winning is not the same as being most likely to win. Looking at who is closest to winning is measuring a fact, while looking at most likely to win is trying to predict the future. I discussed both methods and why I dislike playing fortune teller.

Predicting the future is impossible to do with any kind of accuracy. You present a lopsided situation as an example for your case, but the real test is with fine measurement of a close game. If we both have multiple creatures on the board spread around the place, who is to say that you are 51% likely to win and I am only 49%? Can you honestly concoct a method of predicting the future with that degree of accuracy? Meanwhile, it is very clear if I have 1 life more than you do.

You claim this measurement of life is the wrong type of distance, but it is the ONLY type of distance. Even if we were to take up your attempt at predicting the future, we would have to use some measurement of life remaining compared to damage potential to do so. My method is the exact same measurement. I merely restrict myself to the damage potential that has already occurred, while you wish to try and imagine what damage potential could occur in the future. This might work when there is a major difference between players, but in a close game it breaks down to biased guesswork.

Now, I've already typed more than I meant to, but I will end with saying we should be wary of changing the nature of the game and your predictions about that deserve consideration.

[silverclawgrizzly]

There have been some interesting points made on both sides of this argument. Tie breaking is certainly a topic I could see individual venues tackling differently. I'm inclined to side more towards the argument made by Zuberi and I'll give a few non-Mage Wars comparisons to explain why.

In the other game I play Hero Clix,you get points for KOing enemy pieces. Now the game is over and you have won when you take out their entire team. Of course that doesn't always happen, often times matches go to time limit with pieces from both armies still on the board. Now lets say we're playing 600 points and you've taken out 305 of mine and I've taken out 300 of yours. Your big Superman piece who costs 300 is on his last bit of life and I'm about to finish him off when time is called. Obviously I was about to win, but I did NOT win. Per the rules of clix, which has a very strong competitive scene, I've lost.

Now lets step away from gaming. I hire two men to dig two trenches for me and I tell them they have 1 hour to do it, and since I'm a nice guy I leave them a couple of shovels. The first guy gets his shovel and starts digging. The second guy drives down to the local construction rental store and rents a bulldozer then drives it back only to get there when the hour is up. I come back and ask see that my trenches aren't done. I ask the first guy what's up and he says "Well I tried to get it done but time ran out. I did however get this much done, isn't that worth something?" I may partially be inclined to pay him a little something for his time and work. I ask the second guy what's up and he says "Well I rented a bulldozer and though I didn't actually move any dirt in the time allotted I still think you owe me something cause I could easily do it now." I'm gonna tell him he's out of luck most likely.

Eh one more example just for kicks. I'm a pretty big boxing fan actually, always have been. I like to watch people slugging it out for whatever reason. Well in boxing obviously you wanna knock the guy out, the sooner the better really. Now if we start going at it and I'm just whaling away on you, knocking you every which way while you just absorb the punches hoping I wear myself out and conserving your strength for a big knock out punch....you had better deliver said knock out punch before time runs out. Fighters get scored for actively trying to knock their opponent out on the off chance they aren't able to do it. Whoever succeeded in hitting the other guy more wins. Now yeah sure the other guy may have been about to knock the winners block off, but PREPARING to hit someone is nowhere near as exciting as HITTING someone. Now Mage Warriors are in an arena trying to take the other guy out right? It's a fight, a fancy magic fight, but it's a fight lets be honest. There's a crowd in attendance a winners purse on the line and high stakes all around. Hence the same basic concepts of any other organized fight for entertainment would logically apply would they not?

Anyway this is just my take on it. Like I said I could see venues doing things differently.

[Hedge]

using your thinking, if neither can finish off the other I think the King, Emperor, Ceaser, or the defaocto leader should have them both killed, that is unless one of them  decides to fall on thier "sword," so to speak.   

[silverclawgrizzly]

using your thinking, if neither can finish off the other I think the King, Emperor, Ceaser, or the defaocto leader should have them both killed, that is unless one of them  decides to fall on thier "sword," so to speak.   

I agree completely. 

[Zuberi]

I debated not posting this, because I don't want to sound dismissive, but I promised a response. Your concerns about changing the nature of the way in which the game is played are something we should take seriously. However, I'm not convinced they're well founded. Do we have any data about this actually occurring?

If this burst damage occurred early in the game, and it was made in a poorly advised manner as you suggest, then I suspect your opponent could capitalize on the mistake before time is called and overtake you. If the burst was made towards the end of the game, however, then I don't see the problem at all. By then, you are both aware that time is running out, and you should be making a last ditch effort to pile on the damage and try to kill your opponent.

In the end, I would like to see how big of an effect we're actually looking at. I'm okay with players firing all of their guns at the end of the match, though I'm not sure you would agree, but I think we both agree that a strategy of hitting early and then avoiding combat should not be extremely viable. I really think that anyone capable of finishing the match within whatever allotted time has been given though, should be able to capitalize on any bad plays made during that time. The problem may revert back to the fact that many control strategies can not finish within the time allotted, and so they are vulnerable to these absurd strategies while Aggro Mages would not be. In which case, poor tactical decisions would not end up overwhelming the meta, but control mages would remain underrepresented.

I think I've already discussed the issues with control mages and the time limits. We don't have much control over the time limits, and unfortunately that does give control mages a raw deal. I don't believe that trying to compensate for this with a weighted decision method is the right way to go though.

[echephron]

I've always thought that highest score should win, where score is some function
s =life left+a*(mana remaining+sum(mana cost of each constant you control))

Problems:
Yes it ignores how close a creature is to dying, but that makes it too hard to calculate. I'm only saying its more precise than the current method. i define "winning" not as who did the most damage, but who is best at dealing damage without dying themselves, so life remaining, resistance to mage damage, and damage you can deal to the enemy mage are all variables. all these variables are far too much to compute for the average player, so thats not gonna happen. I fear even adding up the total mana cost of all constants in play will be too much work for some players.

This method does not consider upkeep or channelling, but i'm not sure how i would or how relevent it even is. the constants which are channelling boosters are counted for their mana cost, and the constants with an upkeep cost generally have a lower mana cost, so this helps balance things out on their own without having to directly look at upkeep and channelling.

My first guess is that a = 1/2, but i'm open to alternatives.

Example:
you have 20 mana, 10 life left(life-damage), and 30 mana of creatures, equipment, enchantments(facedown and revealed), conjurations in play.

Score s1 = 10+(1/2)(20+(30)) = 20+25 = 35

Your opponent has 0 mana, 20 life left, and 10 mana of constants.
Enemy score s2 = 20+(1/2)(0+10) =25

s1 >s2 so player one wins!

[Wildhorn]

Then people would just armor up to the extreme and avoid spending mana since a round of channeling suddenly become equivalent of 5 life.

[echephron]

wildhorn, you seem confused. 10 mana in your pool is the same score as 10 mana in your creature, but the creature can actually be useful by damaging the enemy mage or killing enemy constants. just dont make creatures that will die easily. killing a constant will be equivalent to dealing its cost/2 damage to the enemy mage.

or if an attack spell will deal more than cost/2 damage to the enemy mage, that could work too. spend mana to reduce the damage you take, increase the damage they receive, or destroy their cost in a cost effective manner.

[Sailor Vulcan]

It seems more and more to me like what well need is either a way to accurately and reliably predict outcomes of unfinished games, an increase to time limits or my current favorite: just accept the tie and move on. Mage wars tournaments are Swiss style, anyways. Thematically speaking, time limits make sense to me since spectators to the mage wars include some very important people and they don't have all day. if both mages are still alive in the end then neither one has really lost in story. However, if games are frequently and consistently going to time then I really think they should extend the time limit. Or put time limits on the planning phase.