Strength Of Schedule As First Tiebreaker in SR2012

[Junn Khan]

At PAX a few years ago we had a tournie for qualifying for the PAX championship/invitational. I won, based on SoS though there were three of us undefeated, though since the top two went there should only have been two of us. How did this happen? So many people dropped that it the system crashed and 2) all of my opponents stayed in the tournie while second place had about half and third had one. For this reason I hate SoS as a tie breaker and believe that control points are better measure of skill and thus, should be rewarded.

[milothewise]

@ Scout and Jestor -

All your arguments and proposals are based on you having more control over CPs than SoS. You are objectively wrong about that. Matchups affect both of those numbers more than any other factor. You seem to want to be rewarded for getting paired against the scrub who is destined to go 0-2 drop, and tabling him (or winning on scenario). Way to go, I guess? I want you to be rewarded for winning against a harder field.

As Cloud Gatherer mentioned at post #52 (did you even read his post?), random round one pairings are not optional. Some people, at random, will have a more difficult game one than others. Acknowledge that.

As far as not being able to affect SoS - win. More wins = more pairings against higher-rated opponents.

[PG_Pszito]

(!Old timer alert!)

Caster Kill has always, and since creation of the system, been the primary win condition. Scenarios were birthed to balance the issues of tougher casters to catch/kill just hanging back and letting the clock run out. This was counter-productive to the desired environment, although it was an understandably natural evolution of tactics given the rules at hand.

Now that scenarios have developed into a more viable win condition (less of a 'dont abandon the center board' restriction) they have a much stronger case for being a primary win condition. No one is contending this. However scenarios are still a secondary consideration to losing your caster given it is the only non-variable win condition and which is only made exception by a variant rule. Any variant that changes the static parameters of winning hugely changes the dynamic of any game and having a safeguard/counter-measure to any win condition is both wise and inherently more balanced. Simply removing one or the other creates a single minded focus to the system that rewards only one type of play. We should keep the discussion on one track or another if we intend to find any applicable solution.

It also seems there are two different directions said solution is being pulled for here, if not to stick with SoS as a primary tiebreaker (which is what we're talking about right? just tiebreakers... as in only relevant to anyone not undefeated). I completely agree that the penultimate solution would be any scoring method that incorporated all aspects of potential victory; caster kill, scenario CPs, victory points, and finally scoring units. It is the only way to accurately/completely represent a players progress to victory. To make all those both equitable and easily calculable by tournament organizers is beyond me and, as stated, a cornucopia of experienced PGs to discern. Thus the decision of which is valued higher has to be made, landing us at the current SoS>CP>VP system. Since players are seeded by SoS (ie: victory record), not CPs or VPs or any combination there-of, it stands to reason that the most important factor is SoS. If you've played against the toughest (or luckiest) opponents the day can throw at you, why is that not a more important factor than how you beat them? The whole argument that CPs are even in the same category of victories seems very chicken and egg, one cannot matter without the other. Sure, the the randomness of who within your seed you get paired with each round (and who your past opponents do the same with) delivers an uncontrolable variable... so does each round's scenario. Tournament organizers are to random roll which are played each round, eliminating groups of scenarios that have already been played. Both have variation and no player can control what they land with, only how to proceed with it. If you alter either of these tournament structures then of course one aspect can begin to outweigh the others, again dynamically changing how players build lists/play/determine who is victor. This is fine, if it is the desired style of play. For an international standard, we have to land somewhere.

Officially, as of now, it is SoS. I have yet to see any strong case to ratify this beyond the desire to completely reorient the focus of play to being scenario driven. I'm certainly not saying this can't or won't happen but I know what it will do to competitive play; devalue community and entire ranges of models. Lets keep our options open.

[scout's honor]

@ Scout and Jestor -

All your arguments and proposals are based on you having more control over CPs than SoS. You are objectively wrong about that. Matchups affect both of those numbers more than any other factor. You seem to want to be rewarded for getting paired against the scrub who is destined to go 0-2 drop, and tabling him (or winning on scenario). Way to go, I guess? I want you to be rewarded for winning against a harder field.

As Cloud Gatherer mentioned at post #52 (did you even read his post?), random round one pairings are not optional. Some people, at random, will have a more difficult game one than others. Acknowledge that.

No offense, but did you even read my posts? I've pretty clearly indicated that indeed the difficulty of the track you get is both random and uneven. That's not the whole story though, so let's not gloss over everything else. I've also indicated how what control you have over your CPs is limited, but I don't see how you can argue that you have any control over SoS or no control whatsoever over your CPs. A teensy bit is still more than nothing, last I checked. Finally, I've argumented that the fact that you can take a measure of control over your CPs is actually a bad thing, not a plus. Not sure where you think I'm standing on all this, so this just for clarification.

[SportsComedian]

While I agree with what most have said that SOS is most accurate to gauge the finishes for tournaments in most cases. I think exceptions should be made, such as 3-round tournaments, where you don't get enough games to properly determine your SOS. The more rounds you go, the less the random first round matters to the overall standings and flow of the tournament.

I also think that while SOS gives the most accurate standings for tiebreakers, it can take away the fun and meaning of later matchups. Using something like CPs, you might know what you have to do to land "in the money". Being in the top 3 is fun, something to strive for, and will usually net you store credit or something. The difference between 7th and 8th place due to tiebreakers doesn't really matter to most, because by that point nothing is on the line.

I think that to make tournaments the most fun the whole way though, you should use a mechanic that you have the most control over even during your final game. If you're 1-1, and you know it's likely there'll be a couple 2-1 finishers, you know you need as many CP as possible in that final game to have a chance at winning. You bring your better control caster list, you come up with a strategy to make sure you get a decisive win, and you go to it knowing what you'll need to do to win. If its SOS, and a crapshoot of the other tables finishes, or if worse you've already figured out that you can't place in the top 3 regardless because you lost in the first round to the best player, and then got matched up with the worst player in round 2, the rest of the tournament is a lot less fun, especially the final game.

I think putting the path to victory in the hands of a player is a good thing. Sure its dependent on other people's CP scores, but it's something you have control over in your games too. And yes, winning every game puts you in the most control of your destiny. It's always your fault you're not first, because you lost. But sometimes it's not your fault you couldn't finish 2nd or 3rd, it's the SOS calculations and the random matchups. I think that's counterintuitive to the goal of tournaments, to create a fun competitive environment for players throughout.

But as I've said many times throughout this thread, SOS is a good tiebreaker for larger tourneys. But for 10 player 3-rounders, it just doesn't do what it's supposed to do.

[Jestor]

@ Scout and Jestor -

All your arguments and proposals are based on you having more control over CPs than SoS. You are objectively wrong about that. Matchups affect both of those numbers more than any other factor. You seem to want to be rewarded for getting paired against the scrub who is destined to go 0-2 drop, and tabling him (or winning on scenario). Way to go, I guess? I want you to be rewarded for winning against a harder field.

As Cloud Gatherer mentioned at post #52 (did you even read his post?), random round one pairings are not optional. Some people, at random, will have a more difficult game one than others. Acknowledge that.

As far as not being able to affect SoS - win. More wins = more pairings against higher-rated opponents.

You haven't been paying attention to me either. If I continue typing I'll probably get another suspension.

[milothewise]

What rubs (at least some) people the wrong way about SoS is that it tries to both quantify and mitigate your luck of the draw but can't do this really well...It doesn't objectively make the tournament experience better for the contestants and it doesn't necessarily result in a more correct ranking, it just lets you say "these are the X (theoretically) best players of the event, they get to decide amongst themselves who the winner is".

This is primarily what I was responding to from your posts. Felt representative.

SoS does not mitigate luck of the matchup, we agree there. You get a random matchup round one. Question is how we proceed from there. Rewarding in-game stuff like CPs or APs gives advantages to experienced players paired against newer/poorer players, and rewards playing a certain way. I think both of those results are worth avoiding.

The costs you're describing from SoS seem so minor in comparison: If you lose round one, you're in bad shape for the tournament. You will face, on average, lower-quality opponents (judged by match wins) for the rest of the day, and your SoS will suffer. If one of your opponents drops, you are similarly in a poor SoS position, this time through no fault of your own. But at the same time, the guy you lost out to on SoS won his first round and lost to (on average) a better opponent. His opponents didn't go 1-2 or 0-2 drop. Your perception is that the outcome is unfair and random, but his perspective is that his matches were tougher, and he deserves to place above you. I agree with him.

It won't always work out this way, it's a built-in problem for any metric that there will be exceptions, but on average SoS is the best, most efficient way to rank players in a tournament. That's because on average, the best players beat the best players at the table, so those percentages matter.

Not sure why this rankles so much, but it does. I'm pretty solidly in the SoS camp, and I feel like you are consistently ignoring or misrepresenting the arguments for what I see as the objectively better system.

[milothewise]

You should NEVER lose because you were randomly paired against someone who sucked.

I did read many of your posts, though not all of them. This characterizes what I saw as your position. I disagree. You should lose because your first round opponent was worse than the first round opponent of the guy who beat you on SoS. The only alternative is to reward you for beating up on that worse opponent, which is not good tournament practice and not good sportsmanship.

[scout's honor]

Not sure why this rankles so much, but it does. I'm pretty solidly in the SoS camp, and I feel like you are consistently ignoring or misrepresenting the arguments for what I see as the objectively better system.

I just have the impression that for those that really don't like SoS the objective relative merits of the possible systems, and specifically SoS, don't matter as much as the subjective ones. It's possible that by the standards of SoS, objective as they are, you get "ranked down" compared to the rank, for lack of a better term, of the last table you played (the examples of people playing at the final table, losing, and ending up in 4th or 5th place being the obvious ones). That may be objectively correct but not many players in that position will feel they've gotten what they deserved and using a system that results in players getting ranked in a way that feels right is a quality too.

I'm not really advocating CPs over SoS or vice versa, I'm suggesting it might be better to stop using tiebreakers altogether. The issue with that is that it'd pretty much throw any commonly used tournament format out the window.

[milothewise]

Fair enough. Pegged you incorrectly as pro-CPs based on devil's advocation, I guess. I do agree that the objections seem to be largely based how the final rankings have "felt."

I think tiebreakers at some point are going to be required for tournament play. As Cloud mentioned, ideally that would be an easy-to-calculate amalgamation of CPs, APs, and SoS, but that doesn't...you know...exist. What remains as useful and efficient is SoS.

[nicholas_342]

How does SoS handle bye's? It seems kind of messed up that just because a person get's stuck with a bye that means that they don't have a chance of winning no matter how good they do for the rest of the tournament. Or am I misunderstanding this aspect of SoS?

[PG_Cloud-Gatherer]

Do you want to see the equation for a "true" tournament score? It's not for the faint of heart, but I found the thread from back in December where we were thrashing this out. The quoted posts are both mine, but others agreed that the proposed equation was probably as close as possible to a perfectly accurate score, and in no way feasible for actual use.

First, a bit of preliminary math - a proposed tweak to SoS to account for drops, and a mention of why other proposed alternatives were not ideal:

"For SoS purposes ONLY, your TP total would equal (tn/g), where t= tournament points, g= games actually played, and n=number of rounds in the tournament. That way, in a 4-round event, the guy that goes 4-0 is still worth 4 and the guy that goes 0-4 is still worth 0, etc., but the guy that goes 1-1 and then drops isn't worth 1, he's worth 2 because you're figuring on a 50% average on the day. The guy that goes 0-3 and drops is still worth zero, but really, did you expect him to win his last game?
...
The problem I'm seeing with the proposed SoS alternatives so far is that they tend to privilege one kind of victory over another (be it assassination, scenario, or even attrition), and WM/H doesn't do that."

Then, the big composite equation combining CP, AP Destroyed, and Assassinations into a single score that equally weights all the categories:

"... tn/g is more about tweaking SoS to handle drops on its own than about providing an alternative to SoS, which would lead to moz's aptly-named "Puppy-Kicker Syndrome" because non-SoS alternatives don't consider how hard you had to work for your wins.

If you're okay with that, though, I'd go with S=(tn(k/g + c/r + a/gx))/g, where S=tournament score, k=assassinations, c=control points earned, r=control points required for scenario victory, t=tournament points, n=number of rounds, x=point size of the event, and g=games played. Apologies if this is just like someone else's equation; I didn't really look that closely at them because I think SoS is actually just fine once you account for drops.

This equation modifies the t from (tn/g) with a composite score made up of...
k/g = number of assassinations divided by the number of rounds played (i.e. possible assassinations)
c/r = number of control points earned divided by the number of control points required to win scenario victory in every game played (this one is actually a little more complicated, because r will vary based on the specific scenarios, but it automatically accounts for drops if you do it right)
a/gx = total army points destroyed, divided by the point size of the event times the number of games played

By reducing each of the three metrics (assassination, control points, and AP destroyed) to a ratio, you are not privileging any one victory condition or playstyle over the others. The only other adjustment this might need would be to account for Byes, in which case the player would have half an assassination (admittedly a strange concept but you need it for the math to work out), half the control points to win by scenario, and AP destroyed equal to half the size of the event. That gives a solid composite of 1.5 for that game, which is equal to winning and destroying half your opponent's army.

Admittedly, this seems pretty complicated to do by hand for a major event, but I'm confident that something like Dohyou or even a simple spreadsheet could handle this with no problems. If you have a properly-formatted (and filled-out) scorecard for each player and you know how many control points were needed to win all the scenarios you used, it's just simply what one of my college professors called "plug and chug" - just plug in the numbers and chug through the arithmetic.

ETA: If you wanted to, you could even use this in conjunction with SoS. You'd just have to substitute S for t in the standard SoS formula and not worry about accounting for drops because S has that built in."

So there you have it: the "perfect" tournament score equation. S=(tn(k/g + c/r + a/gx))/g. How usable do you think that really is?

[Rightfin]

Lots of interesting points here, but it is quite evident that some opinions are passionate on this subject.

I'll stick to facts, and some previously group accepted perceptions:

One (and by no means the only) goal of a tournament is to fairly, and equitably, determine a winner. Fairness is also desired to not skew results based on any one particular style of play, but instead to represent a general mastery of the game in all forms. Is anyone not in agreement here?

• SR2012 pairs down winners through a series of elimination rounds. At the conclusion of each round, the number of undefeated players diminishes until the final round where the final two undefeated players match off. Ultimately there will be only one clear undefeated winner. (There are exceptions to this format (Insufficient time, unforeseen victor dropout, field of players too large, DQ, or a particular event where rounds were not paired off).
• The winners determined through this manner will not be affected by SoS, CP, VP, or through any other means – how they achieved these multiple victories could be through scenario/ assassination, death clock time-out, DQ, etc. In any case they are irrelevant and completely unaffected by any form of determination in this thread discussion.
• Unless there is a future ranking systems, handicap, professional league, or multiple event elimination formats developed it will impossible not to randomize the initial pairings. It is probable that the players may represent a wide cross-section of player skill, experience and/or luck.
• There will be a finite amount of time available for the event. The time allotted is based on expected attendance, fatigue level of players, and interest. Generally 3-5 rounds are the acceptable norm.

Where SoS comes into play is due to points 3 and 4 above. 3-5 rounds based of random pairings is not a large enough mathematical sample to create the “perfect” system. It has been stated the SoS is simply the best system available to determine standing due to the potential amount of players with duplicate win/loss standings

In theory, SoS will skew the standings based on the effort required to reach your final win/loss record.
Essentially, two players with 3-1 records will favour the player who may have defeated three top level players against a player who may have defeated three novice players.

Am I missing anything here? Is this, in essence, the Strength of schedule argument? No emotion, no opinion, is this the most basic (and brief) purpose of the system?

I'd like to counter the statement said by a couple of posts that this really is “the best way”, but I want to get my facts 100% straight first.

[quindraco]

S=(tn(k/g + c/r + a/gx))/g.

r is undefined; typically the number of control points necessary for victory is infinite, since you need both a minimum number AND more than your opponent.

[PG_Cloud-Gatherer]

r is undefined; typically the number of control points necessary for victory is infinite, since you need both a minimum number AND more than your opponent.

That depends how you define "necessary for victory"; it's possible to get a number >1 for that particular part of the equation. You have to disregard "and more than your opponent" when you're doing that part. This actually works as intended; if c>r, that means you played against a good opponent (or colluded with your opponent to farm CPs, in which case you'd both be ejected from the event for cheating).

[Terraneaux]

Fundamentally using CP's means that the best play is, sometimes, not to 'go for the throat' and win the damn game. When I'm playing a game I don't want to worry that my early assassination will hurt me in the long run over tournament standings. Fundamentally, SoS means that players will be rewarded as much as possible for winning games as quickly and decisively as possible. CP's means that players will be encouraged to **** around and draw out games in the hopes of scoring CP's.

[PG_ThatRickGuy]

Here's an idea:

Use CP for 1st tie breaker.
If you win a game by assasination, you get the 'victory' CPs. (ie: scenario is first to 3 CPs wins, you win by assasination, you get 3 CPs, regardless of how many CPs you gained in the game)

This rewards people who win the game while it works against people who go to dice down.

If you're playing an assasination list and go to dice down, you're probably losing anyway.
If your playing attrition/control and go to dice down, you probably picked up a CP or two.

-Rick

[Terraneaux]

Here's an idea:

Use CP for 1st tie breaker.
If you win a game by assasination, you get the 'victory' CPs. (ie: scenario is first to 3 CPs wins, you win by assasination, you get 3 CPs, regardless of how many CPs you gained in the game)

This rewards people who win the game while it works against people who go to dice down.

If you're playing an assasination list and go to dice down, you're probably losing anyway.
If your playing attrition/control and go to dice down, you probably picked up a CP or two.

-Rick

Have you done the math and decided that that's *exactly* how you want to value assassination wins vs. CP wins? How many CP wins is an assassination win worth to you?

[PG_ikildkenny]

@ThatRickGuy and Terraneaus - that is VERY similar to my Idea (Post#62)...

@ Cloud-Gatherer: indeed, the "overachieved" CP in scenarios where you could "hog" CP could be a problem ... I see 2 solutions to that:
1) calculate the CP down with a factor (so if by CP the scenario ends 5-4, the system/TO would count that as a 3-2 or even 3-1 (the factor would be 3/5, the "looser's" CP could than either be rounded up or down...)
2) (a) subtract a sufficient amount of excess points ... this amount would than be (winner's CP - max. needed CP) ... so in above excample that would be 2 resulting in a 3-2. Basically this would mean, that (b) the "CP-counter" would be resettet to 1 (or 2, depending on the scenario) below max. needed everytime the two players would score a draw of max. points (so as soon as the CP get to 3-3 for both player simultanously, the counter would reset itsself to 2-2) ...

2b) is probably the easiest way to do things, since players could do that during their game and noone stands to gain anything from hogging CP (this could and should imho be a generally used system in SR to avoid CP hogging)

@ Topic

I did some additional thinking on the subject - scout's honor hat his fare share with those.

I think there is one thing we can agree on: Final standings should be determinded by Winnings first, than tiebreaker(s) kick in.
The Problem here is, that EVERY Tiebreaker has it's drawbacks - some more than other:

SoS:
SoS is completely out of your hands (Byes and Drops make this even harder - allthough CloudGatherer's Idea (tn/g) seems to be the perfect solotiun - mainly b/c it only has to be used IF someone drops out) - imho it's more a measure of "how lucky were you to get opponents that by accident or skill played a good/bad tournament" than your opponents actual strengh (in the tournament) and thus subsequently your own strengh (since ideally you won against equally strong or seemingly stronger opponents).
This ramdomness is probably SoS's main flaw ... and not really fixable

CP:
This has been discussed to death - CP as a second tiebreaker would penalize assasination-lists for no other reason than to penalize them ... on the other hand it IS a better measure of your superiority in a game (since you were able to controll the gameflow in a matter that you could collect them - which requires more skill than to "see the killing angle").
So using CP as a first tiebreaker, player "skill" might be better reflected but it allso makes assasination less viable and thus "risky gaming" more favorable - which could leed to rediculous games

VP:
Yes, I'm bringing these in too - why?
VP is the measure for an "Attrition war winner" ... usually this is allso somehow covered by CP but in the end the player, that scored the most VP usually has attritioned there opponen more, than the one who pushes him out of the szenario and wins ... VPs should not be discarded but with two tiebreakers beforehand become far too insignificant to really matter ...

So a unifying formula (as postet by CloudGatherer - btw. a=VP in the tournamen? it's not mentioned but seems to be logical since a/gx than gives you a number between ~1.1 [jack/beastpoints] and 0 that accounts for VP earned relative to VP possible) seems to be the thing we need ... a single tiebreaker, that, in a way) combines all the tiebreakers and thus hopefully letting there drawbacks negate eachother

unfortunatly, allthough it seems to give a farely accurate result, this above mentioned formula is far to complicated. We would need a simpler one, that can be done within seconds by the TO and ideally without a calculator (or a mathgenius nearby).

I don't know whether this is practicable but I'd suggest this "simple" way:

all used values SHOULD be modified by above suggestions ... so SoS has the "drop correction factor" and CP should be corrected "below max. possible"

"Tiebreaker Value" (TbV) = (SoS+CP)*VP/(Games Played*maxVP)

it's a slight variation af the above formula. basically SoS and CP are equally important. a TO can influence there relative importance by scenario selection - so in a tournament, where scenarios need a lot of CP to be won, assasination wins would be less desirable, whereas scenarios, that require less CP would make them relatively more appealing. To loosen this up, one could simply reward controllpoints for assasination - my suggestion would be 2 (since most SR szenarios need 3 CP. Those that require 2 are usually won by assasination anyway and those won with 4 see less gametime imho).

so TbV could look like this:

(SoS+CP+2*Assasinations)*VP/(Games Played*maxVP)

there is one remnant of the orinal formula, that some might not like ... the "VP Factor" ... in most cases, this should be a number between .4 and .7. Depending on EventSize this could screw more or less with the initial value and thus create some strange results. the simplest solution here would be to take VP OUT of the equation and simply let it sit as second tiebreaker.

So TbV1 (first Tiebreaker Value) would be:

TbV1=SoS+CP+2*Assasinations

and TbV2 (second Tiebreaker)

TbV2=VP (there is no logical reason to calculate on that - maybe an average value for drops ... something like TV2=VP/(No of Games in this tournament)

So placing should than simply work out as:

Wins - TbV1 - TbV2

[PG_ThatRickGuy]

Have you done the math and decided that that's *exactly* how you want to value assassination wins vs. CP wins? How many CP wins is an assassination win worth to you?

Why make it so hard? If the Scenario is first to 3 CP wins, you get the CP's you score during the game, or if you win by assasination, you get 3 CPs.

If the Scenario is first to 2 CPs wins, you get the CP's you score during the the game, or if you win by assasination, you get 2 CPs.

You could add a little bit of logic to deal with first to x with more than their opponent, or games that end in assasination after the player has earned x-1 CPs...

For example, in a scenario of Close Quarters. (first to 3 with more than opponent)
Player 1 has 1 CP.
Player 2 has 0 CP.
Player 2 wins by assasination.
Player 1 reports a loss with 1 CP.
Player 2 reports a win with 3 CP.

Or in a scenario of Gauntlet. (first to 2 with more than opponent)
Player 1 has 1 CP
Player 2 has 1 CP
Player 2 wins by assasination.
Player 1 reports a loss with 1 CP
Player 2 reports a win with 2 CP

Or a game of Guidons. (first to 3 with more than opponent)
Player 1 has 4 CP
Player 2 has 4 CP
Player 2 wins by assasination.
Player 1 reports a loss with 4 CP
Player 2 reports a win with 4 CP

There you go, now whether you win by Scenario or by Assasination you are put higher in the rankings than someone who wins a game by tiebreaker.

-Rick

[Jestor]

I did read many of your posts, though not all of them.

Please read all of them.

[DemonCalibre]

I think that this is a big problem. I just cannot understand the attitude that someone decides to just give up because they can no longer finish in first place. What they are also doing is to screw up the SoS of the person that defeated them.

Out of all the bad ways SoS seems to be the least bad. The only adjustment I would want to see is something that does not punish you for playing opponents who dropped.

This statement puzzles me. You do understand this is a for pleasure activity, and more importantly it's that persons free time, and they can choose to spend it however they wish, and are no obligation to stay to an all day event. I run Tournaments frequently, and I don't mind if guys drop. Why? because it is not my place to tell them what to do with their time. They came, they dropped their money down, decided it wasn't worth staying and left, last time I checked that was something totally within the realm of OK to do. The fact they might hose someone else's chances to win doesn't really have to be a concern for them because why are they concerned about helping someone else win?

APs, and CPS are worse tie breakers, why? because they promote degenerate behavior. I played in an event, where they used APs as a tie breaker, played against a new player, after previously being burned I did the mean thing and farmed him for APs/CPS. After making him sit through a game where I could have won any time, I tabled him for all but 3 points of his list. At the end of the game I apologized to him for farming him, and told him I wouldn't have except for the format. Then later it mattered because another guy at the tournament just assassinated his round one opponent also a new player, and farming the APs made the difference on him placing second, or third.

Frankly, promoting NPEs seems like a bad idea. Right now farming APs is not really worth that much as far as tie breaks go(I have only gone to CPS a couple times after SOS), so there is little need for me to knuckle down and farm off every last model on my opponents table when I hit a guy who is just not that great at the game, so I don't have to drag it out to farm him with no hope of winning, to benefit myself that much. I might grab CPS, but that isn't the same as intentionally and systematically exterminating a dudes army.

Lastly, at why CPs are at the very least no better a metric then SOS. Is the scenarios aren't all equal about giving out CPS to both players. In some scenarios like Supply and Demand, and Incoming. It's very hard for the second player to gain CPs, particularly if your opponent is playing a Pin/Turn Denial list that excels at those scenarios. So having a bunch of land mind scenarios where if you don't win first turn, your ability to gain tie breakers that round immediately drops to nearly zero is also not particularly superior to SOS.

[2LiveIs2Die]

SOS should always be first.

[PG_ThatRickGuy]

There are basically two options:

Your performance.
Your opponent's skill level.

If you want to tie break by your performance, you need to have the scoring system designed such that it does not reward people for abusing the hell out of new players.

If you want to tie break by your opponent's skill level, looking at their wins on the day is a really poor way of doing it. ELO would provide a significantly better solution. If someone drops, you still get the points for their ELO when you played them. The ELO exists outside the scope of the event, so you wind up with similar skill level matchups on round 1, where as SOS doesn't become even remotely accurate until after 3 rounds.

etc...

So that's my 2 cents: either improve how scoring works to avoid newb-farming, or switch to ELO. SOS and CP tie breakers are both inaccurate and motivate poor sportsmanship.

-Rick

[Coil]

This statement puzzles me. You do understand this is a for pleasure activity, and more importantly it's that persons free time, and they can choose to spend it however they wish, and are no obligation to stay to an all day event. I run Tournaments frequently, and I don't mind if guys drop. Why? because it is not my place to tell them what to do with their time. They came, they dropped their money down, decided it wasn't worth staying and left, last time I checked that was something totally within the realm of OK to do. The fact they might hose someone else's chances to win doesn't really have to be a concern for them because why are they concerned about helping someone else win?

I am all for free will and acting according to ones preferences. It is just that to me it seems strange to have so little interest in the game that losing the chance to win (which you do by losing your first game, unless winning players decide to drop), makes you want to quit. Is winning really that important compared to playing games?

One should certainly try to win. But if you go in with the mindset that nothing but winning the tournament is acceptable the odds are that you will have a bad time. To me at least playing in a tournament is fun in itself. If I should happen to win or place well is just a nice bonus. Sometimes SoS helps me place well and sometimes it puts me further down the list. At L&L I was among the bottom in B,S&T and in the top at Speedmachine. Both tournaments were a ton of fun though.

I totally agree with your points about CPs and APs.

[PG_Cloud-Gatherer]

So a unifying formula (as postet by CloudGatherer - btw. a=VP in the tournamen? it's not mentioned but seems to be logical since a/gx than gives you a number between ~1.1 [jack/beastpoints] and 0 that accounts for VP earned relative to VP possible) seems to be the thing we need ... a single tiebreaker, that, in a way) combines all the tiebreakers and thus hopefully letting there drawbacks negate eachother

Yeah, you're right, for some reason I didn't specifically define "a" in the list of variables, though it is explained later in the text. In my grand unified tiebreaker formula, a = army points destroyed.

[Rightfin]

One question: For Con events where Masters tournaments are by invitation or qualifier, where the potential field of (usually) 16 players is culled from preliminary rounds or events. Are the first round matchups also random?

I ask as that would appear to be the opportunity to remove cloud-gatherer's suggestion that the more "unfair" perception is due to the random nature of pairings. Would this not be the perfet time to seed pairings based on results of the preliminary rounds?

[PG_Cloud-Gatherer]

One question: For Con events where Masters tournaments are by invitation or qualifier, where the potential field of (usually) 16 players is culled from preliminary rounds or events. Are the first round matchups also random?

I ask as that would appear to be the opportunity to remove cloud-gatherer's suggestion that the more "unfair" perception is due to the random nature of pairings. Would this not be the perfet time to seed pairings based on results of the preliminary rounds?

I'm not sure if they ARE random, but you're right that this would be a perfect opportunity to remove that randomness.

[computertrucker]

I like clouds formula. It rewards all styles of play. CPs, assassinations, and attrition army killers, and overall victory.It also has the ability to award for number of total games won. A hybridized mathematical system that takes into account every aspect of gameplay vs each opponent and adds up throughout the course of the entire tournament would definately give the most accurate sense of who came in first, 2nd, or 3rd..it also would not penalize someone if someone they played in an earlier round dropped from the event.

Having people drop and mess up someones else chances is not fair in any sense of the word. Rewarding one style of play over another can handicap entire army builds. A system that takes everything into account is the way to go and numbers don't lie

Me I'm glad I go to tournaments to be able to play and have fun. If I win that's cool to but thats secondary.