TWDA: What it is and is not useful for

First off, I want to applaud Jeff Thompson (The Lasombra) for all his hardwork collecting and archiving the Tournament Winning Deck Archive (TWDA). It is an invaluable resource for the VTES community. I think that it does the following exceptionally well:

1. Archives successful decks
2. Recognizes players who have won lots of tournaments and with a variety of decks
3. Provides a database useful for new players to copy decks and learn about design, and for experienced players to improve their design and get an idea of prevalent play environments

But on the forums, I often see people treating the TWDA as some sort of holy grail. "The TWDA says this, so that's the end of this discussion! This card appears in the TWDA but this card doesn't!" People commonly refer to the TWDA is a magical crystal ball that accurately predicts how strong/weak and competitive various clans are. I do not think it does a very good job in that function (or even moderately decent). As far as I know, Jeff Thompson has never suggested that the TWDA is an accurate weathervane for the strength of clans, but lots of players seem to act like it is a decisive and undisputable evidence for such a thing.

I think using the TWDA to gauge relative clan strength is a very, very flawed approach. This thread is obviously to highlight why it isn't a good tool for this purpose, and also to suggest ways that it could be refined if you actually *wanted* such a mechanism.


1. The TWDA only catalogs the winning deck

Obviously, every deck and clan would like to win. But competitiveness and strength can't be measured simply by the winner. A deck that is "second best" or that consistently gets into finals (but just has trouble clinching the win) is still a competitive and strong deck.

Some CCG's use some sort of "cut-off" system to evaluate the success of decks: Top 4, Top 16, etc. If you wanted a more accurate gauge of deck performance, you should really be collecting and looking at all the finals decks (and possibly even decks that tied to get into the finals on GW/VP/TP but lost due to a random coin flip). On multi-day events, you could look at the cutoffs to reach the second day.


2. Participation statistics

To figure out how well a clan is doing, you also need to know how much it is being played and how well all the clan decks are doing. Looking at the TWDA "wins" doesn't distinguish between the following scenarios:

A. Underrepresentation

The clan is fine, but people just don't like it and don't play it. If the clan isn't represented at all at the tournament, of course it can't "win." Maybe the clan is strong but it's just a boring deck. Maybe people dislike the clan (due to RPG considerations). Many people dislike the art due to personal aesthetics.

B. Overrepresentation

"Even a blind squirrel finds a nut once in awhile." If a given tournament has a lot of players playing a given clan, then there's a higher chance that that clan will "win" and be in the TWDA (assuming the decks aren't all contesting the same vampires).

On the extreme side, if I get 9 friends together and we all decide to intentionally play Samedi in a 10-player tournament, obviously a Samedi deck is going to be entered into the TWDA. But that doesn't mean anything about whether the Samedi is good or bad against other clans since everyone is playing Samedi.

C. Performance statistics

Suppose a !Brujah deck wins a 100 person tournament. Looking at just the winning clan doesn't differentiate between:

- Maybe there were 50 !Brujah decks, and 49 of them got 0 VP's (and just the 1 was successful)

- Maybe there were exactly 4 !Nosferatu decks played at the tournament and all 4 of them made the finals but lost to the !Brujah (when combined with the above point, that would indicate the !Nos are better off than the !Brujah, even though the !Brujah won the tournament)

- Maybe there were 50 !Nosferatu decks and they all got 0 VP's

In each of these scenarios, completely different conclusions would be reached based on important data that isn't captured by the TWDA.

In other CCG's (L5R for example), they do factor in the participation rate (how many people are playing that clan) to see how well the clan is really doing.

In VTES, if you wanted a more accurate measure, you could for instance calculate a clan's *average* GW's and VP's. That would let you know how well the clan is doing with decks not reaching the finals or winning (and if you use the earlier suggestion to track all finalist decks, then you have even better data).

Of course this would require a lot more work, and would likely necessitate required deck lists from all players.


3. Flawed "Clan" criteria

It is my personal opinion that the clan criteria under the TWDA is flawed. According to the website:

"If a deck has 5 or 6 vampires of one clan, and no other clan has more members in the crypt of that deck, it will be listed under that clan."

Many people have made valid arguments that TWDA "clan performance" doesn't accurately reflect how well a clan is doing (especially due to atypical star vampires).

A. Bar is too low

With only requiring 5 vampires of a given clan, I personally think that's too low of a bar.

Storyline events require 75% (so at least 9, for a 12 card crypt), though that is admittedly arbitrary as well.

I find it odd that a deck can count as a "clan" deck when the *majority* of cards in the crypt are *not* members of that clan. Right now, it's common to have 5 members of Clan A, 5 members of Clan B, and 2 other vampires. It's even worse if all the copies of Clan A is of one vampire and all the vampires of Clan B is of a second vampire (unrepresentative all star decks like a Girls deck with 5 Aksinya and 5 Cybele).

If a deck has 5 Malks, 5 !Malks, and 2 Abominations, it counts as both a Malk and a !Malk deck. But if the deck had 6 Malks and 7 !Malks, it would count as a !Malk deck but NOT a Malk deck. I find it odd that the second deck has more Malks than the first (and a higher proportion), but the first counts as a Malk deck and the second doesn't!

I think a better requirement for clan decks is requiring a majority (> 50%, so at least 7 cards in a 12 card crypt) or *maybe* at least 50% (allowing a 50/50 to count for two clans).

The requirement should also be proportion/percentage, rather than a fixed number, like the 5 currently being used. Technically speaking, under those rules you could actually build a deck that counts as a "clan" deck for every single clan in the game! Just use 5 crypt cards from every clan. With 40 clans (including Caitiff and Imbued), you just need a 200 card crypt. Sure, it'll be hard to win. But all you have to do is win once and you've won with *every* clan (some players are trying to get a TWDA deck for each clan, but you can do it all at once with this 200 card crypt).

Requiring a majority of vampires in the crypt to count as a clan deck, also reduces the likelihood of all-star vampire decks contaminating the results. You'll still need to add a few nominal weenies or support vampires to make the cutoff, unless you play with 7 copies of one vampire (though it could still be an issue for vampires with an Advanced version that is still the same clan).

B. Factor in multiple copies of a single vampire

One possible way to deal with all-star vampires muddying the waters, is to introduce some sort of function that penalizes including too many copies of a given unique vampire and encourages multiple minions.

For example, say you used some sort of root function (say square root) for determining whether it counts as a clan deck.

A deck with 7 Mata Hari and no other Ravnos, isn't really a "Ravnos deck," it's a "Mata Hari deck." Similarly, Tupdog decks aren't really representative of Gargoyle decks. So if your 9 Tupdogs, counts as just 3 (square root of 9), then it won't be counted as a "Gargoyle clan deck."

Of course you would probably have to make some accommodations for Bloodlines, and especially Scarce clans (i.e. Abominations only have 2 unique vampires in a given Grouping pair). And it doesn't completely solve the 100% of your crypt is one vampire, so having a crypt of 1000 copies of that vampire kind of skirts around the root function.

C. Factor in the number of decks that are winning

A lot of people say Clan X has Y wins in the TWDA so it must be doing great! But there is a very big difference if each of the Y wins is a very different deck, or whether all of the Y wins are pretty much the same deck.

I think a clan is stronger if it has multiple, competitive decks (rather than just a single, super-strong deck).

So when evaluating the number of "clan wins" in the TWDA, you should also be using a function that penalizes too many wins in the same category (say a cube root or a fourth root). If we already know that a given deck (Girls, or Tupdog) does really well, racking up more wins with that same deck doesn't really tell us anything helpful.

Of course, this does require some subjective judgment of whether this intercept/combat deck is sufficiently the same as this other one. Or whether these two Nephandi decks are essentially the same.


4. 2010 vs 2011

Some people occasionally reference just 2011 TWDA data. But since we haven't had any new cards since early 2010 (Heirs released February 3rd, the couple of storyline promos legal April 1st) most of the 2010 data should be just as valid as 2011 data since the card pool environment has not changed.

Suppose a clan or deck was really successful in 2010 (after April 1st). If it didn't do so well in 2011, that does *not* mean that the clan sucks (because the deck clearly did well in the second half of 2010, when the card environment was the exact same).

So people referring to 2011 TWDA data should really include 2010 TWDA data as well (and if you want to get technical, 2010 TWDA data starting from April 1st).


5. TWDA isn't all-inclusive

Some players might not want to enter their decks into the TWDA, so these wouldn't be counted.

Some tournaments might not be counted because they aren't VEKN sanctioned (hey, maybe they found the secret to killer Abomination decks!).

Some tournaments might not be counted because they have less than 10 players. 10 players is a bit of an arbitrary number. Why 10? Why not 20? Why not 50? Why not 5? (Since Jeff manages the TWDA, he can use whatever criteria he wants. But when other players start extending this to overall game balance and how clans stack up, why limit yourself to just a subset of the possible data?).


6. For Library cards

Some people also say "Library card X is or isn't in the TWDA." I think the argument for library cards is less convincing (*unless* you factor in the number of copies in each deck). Since there is a lot of card flow, changing 1 card in a 90 card deck is unlikely to significantly impact things.

If I take a deck that is consistently winning tournaments, swap out a single card and put in a wallpaper card (Eyes of the Dead or Up Yours!) and then win with it, the fact that Eyes of the Dead of Up Yours! is now in the TWDA does not mean that it's a good card. If I convince everyone at a tournament to sub in one copy of Eyes of the Dead, and everyone does this for several tournaments so Eyes of the Dead keeps popping up in the TWDA, that says nothing about the strength of the card.

Aside from those extreme scenarios, when it comes to looking at the prevalence of a card, you really need at least two separate measures to track things. Drawing upon my past Economics training, you want one metric that simply measures how many decks it appears in (regardless of whether there is 1 copy or 20 copies) and a different measure that rewards concentration within a deck (so you can square or cube the number in each deck, and then take the sum of that. This would help counteract the prior situation where everyone includes just 1 single copy of a card.). Another consideration which affects things is Unique and once per game cards, but I haven't spent enough time to figure out a workaround to factor that in.


*** Now some people might argue that some of these factors may "balance out" over time, with a large number of tournaments. It might work for the "people conspiring to mess things up" situations, but generally I don't buy that argument. Since the underlying criteria is flawed (only looking at the winning deck, using a flawed criteria of a "clan deck") I don't think the "balancing out" argument resolves the problem.


I'm sure there are other valid criticisms and plenty of possible fixes. But this is just a lengthy summary of why I think the TWDA is NOT useful at all for determining how "well" a clan is doing.

So whenever players continue to claim "Clan X is too strong or Clan Y is too weak because that's what the TWDA says," I'm just going to continue to roll my eyes and ignore it since it is a terrible metric for that particular purpose.

Did you really write all this up because you and me disagree that Tzims are competitive? I'm fairly sure I already stated in that post initially that the TWDA is not a particular accurate tool for determining what's good and what's not, but it's is certainly the best indicator we have.

Also: did you ever take any sort of statistics class?

The TWDA only catalogs the winning deck

Yes it does. However, it is *extremely* unlikely that decks that can ENTER the finals consistently (as you put it) can never win one. The more likely scenario is that these decks are not very good at winning games in the first place and thus don't claim a lot of finals spots and end up not winning a lot of them as a result.

The clan is fine, but people just don't like it and don't play it

Given the timespan, sample size and group of players that the TWDA represents, it is also extremely unlikely that a certain clan is underplayed for other reasons than simply not being competitive. People don't play Samedi because they are bad, not because they don't like the card art.

Overrepresentation

This can certainly be a factor. A deck that shows in the TWDA is more likely to be a good deck than one that doesn't. Of course if people see good decks they give them a shot and might end up winning with them. See last year's Girls! invasion and this year's endless stream of animalism based decks.

Performance statistics

Again, it is extremely unlikely everyone suddenly plays !Brujah and the meta gets warped beyond recognition because of it. The more likely scenario is that tournaments tend to be a mixture of "netdecked" TWDA entries, mixed with toolboxes and some oddball decks. And even if you run into a warped meta for a day just one win will not really influence any sort of statistics you might want to draw out of the TWDA.

Flawed "Clan" criteria

This is certainly the case, yes.

2010 vs 2011

Wah? This seems just a random inclusion to strengten your own arguments in "that other" thread. Of course the most recent data is just that - the most recent data. That accounts for the current metagame, which mostly revolves around dealing with animalism and poolgain. The cardpool being identical has nothing to do with it.

TWDA isn't all-inclusive

It's inclusive enough. And even if it wasn't, the decks not being recorded (of which there certainly are enough) are very likely to have similar spread to those that are being recorded. Cos ya know, that's how statistics work. There isn't some evil force that makes people not report when they won with their Osebo decks.

The funniest part is that you write a 4 page essay about how the TWDA is a bad metric and then leave out the only convincing argument: the minimum tournament size is set to 10.

People win 10-men tournaments with stupid decks all the time by virtue of getting into the finals with 0 GW 2 VP and then getting a random VP before timing.

*** Now some people might argue that some of these factors may "balance out" over time, with a large number of tournaments. It might work for the "people conspiring to mess things up" situations, but generally I don't buy that argument. Since the underlying criteria is flawed (only looking at the winning deck, using a flawed criteria of a "clan deck") I don't think the "balancing out" argument resolves the problem.

So what you're saying is "the laws of statistics don't apply to the TWDA because if they do, half my argument is gone". Alright, that makes sense. I guess I have an answer to my question here...

But this is just a lengthy summary of why I think the TWDA is NOT useful at all for determining how "well" a clan is doing.

Well, ya know, you can think whatever you want. The reality of it all is that on average, clans with good cards, crypts and synergy will be played more, win more and thus be in more finals and thus win more tournaments.

Similarly, clans with bad cards, bad disciplines and bad crypts will, on average, get played less, not score an awful lot of VPs and just end up randomly winning 2R+F 10-men tournaments.

If you want to argue that !Nos, Blood Brothers and Baali are as competitive as Toreador, Ventrue and !Malks, then feel free but I'm just going to continue to roll my eyes.

Izaak wrote:

Did you really write all this up because you and me disagree that Tzims are competitive? I'm fairly sure I already stated in that post initially that the TWDA is not a particular accurate tool for determining what's good and what's not, but it's is certainly the best indicator we have.

Don't flatter yourself. This has little, if anything, to do with you. And nothing to do with the Tzimisce. Lots of players have tried referencing the TWDA as a definitive source. I have had lots of discussions (both on and off the forum) with many different people.

Also: did you ever take any sort of statistics class?

Yes, I've taken several. Taking a quick look at my transcripts, I see at least 4, though I might have missed a few (especially those that are cross-listed).

Have you?

The TWDA only catalogs the winning deck

Yes it does. However, it is *extremely* unlikely that decks that can ENTER the finals consistently (as you put it) can never win one. The more likely scenario is that these decks are not very good at winning games in the first place and thus don't claim a lot of finals spots and end up not winning a lot of them as a result.

It's unlikely, but you're still drawing assumptions.

In addition, player behavior is actually quite relevant here. Many players like to play a deck for the purpose of getting it into the TWDA, and will keep playing it until they win with it. I think it would be relevant to know how many times a player tried to get his TWDA in there (roughly how many tournaments he played it, and how many finals he got to) before it finally got there.

I think there's a difference between someone who has been trying to get a deck into the TWDA for 1 or 2 years. Versus someone who plays the deck once, wins with it, and then doesn't play it again. I do know several players who retire their decks as soon as it enters the TWDA.

Given the timespan, sample size and group of players that the TWDA represents, it is also extremely unlikely that a certain clan is underplayed for other reasons than simply not being competitive. People don't play Samedi because they are bad, not because they don't like the card art.

I disagree. Yeah, for some players clan and disciplines don't have any sentimental meaning. But with so much backstory and source material, you have lots of players who feel strong identifications with particular clans. That was the whole intention behind the Clan Newsletters in the old days (see the thread on VEKN history). You have some clans which are championed, even when they are bad or have historically been bad (there was a stretch when the Assamites were really terrible yet there were players who loved the Assamites and played them even though they were bad at the time).

If you have cases where people love the clan and play it even if they are bad, I can see other scenarios where people hate the clan and don't play them even though they are good.

This can certainly be a factor. A deck that shows in the TWDA is more likely to be a good deck than one that doesn't. Of course if people see good decks they give them a shot and might end up winning with them. See last year's Girls! invasion and this year's endless stream of animalism based decks.

You bring up a good point that tournaments are not independent of each other (the results of one event may influence the deck choices of a later event), and that trends and bandwagoning off of recent tournament wins, may lead to clumped results.

Performance statistics

Again, it is extremely unlikely everyone suddenly plays !Brujah and the meta gets warped beyond recognition because of it. The more likely scenario is that tournaments tend to be a mixture of "netdecked" TWDA entries, mixed with toolboxes and some oddball decks. And even if you run into a warped meta for a day just one win will not really influence any sort of statistics you might want to draw out of the TWDA.

I highly doubt that the clan wins are evenly distributed.

Some players tend to like certain clans, and are often known in their playgroups for playing a lot of a clan (it especially makes sense due to card inventory. If you traded for a lot of Baali cards, you're probably going to make a bunch of Baali decks and not as many Trujah decks).

Some playgroups and regions are often known for having several players who like the same clan (my current playgroups sees a lot of Baali and a lot of Salubri decks, from multiple players). Any tournaments in the region are going to be understandably skewed in favor of the popular clans.

2010 vs 2011

Wah? This seems just a random inclusion to strengten your own arguments in "that other" thread. Of course the most recent data is just that - the most recent data.

Wow, you're certainly self-centered if you think this is all about you.

I already made the exact same point in a different thread over a week ago. You probably wouldn't have been able to see that post, so that's no fault of your own. But it is presumptuous to assume that this is all just for you.

That accounts for the current metagame, which mostly revolves around dealing with animalism and poolgain. The cardpool being identical has nothing to do with it.

Card pool is relevant.

Trends and the metagame shift (often like a pendulum). You're making the same error now as you did in a different thread, by focusing on just a small cross-section of NOW.

If I have a good deck, but then all the players in my playgroup build their own decks to counteract mine, so mine no longer works well in my local play environment, that doesn't mean that my deck is suddenly bad. It's still a good deck. It might not be competitive right now in this playgroup, but it's still a competitive deck overall.

Even though the metagame might not be conducive for a particular deck at this time, doesn't mean the deck is bad. If the metagame shifts, the deck can suddenly be competitive.

There is a difference between "competitive NOW" and "generally competitive." You see to focus on just the first. So it seems a deck could be good one month. And then horrible the next. And then great the following month. And wallpaper the next. And then horribly broken the one after. I would argue that the available cardpool provides a longer-term perspective on the competitiveness of a deck rather than following day-by-day, short-term trends.

The funniest part is that you write a 4 page essay about how the TWDA is a bad metric and then leave out the only convincing argument: the minimum tournament size is set to 10.

People win 10-men tournaments with stupid decks all the time by virtue of getting into the finals with 0 GW 2 VP and then getting a random VP before timing.

Maybe you missed the point about choosing 10 as an arbitrary limit:

echiang wrote:

Some tournaments might not be counted because they have less than 10 players. 10 players is a bit of an arbitrary number. Why 10? Why not 20? Why not 50? Why not 5? (Since Jeff manages the TWDA, he can use whatever criteria he wants. But when other players start extending this to overall game balance and how clans stack up, why limit yourself to just a subset of the possible data?).

If you're worried about those small events, set a higher bar of 20 players or 50 or 100.

And speaking of sneaking into the finals, there are also hypothetical cases where a deck could get into finals (or win) because other people get DQ'ed, or people just decide to concede (maybe they have to leave). Does it happen often? Probably not. But you would probably want to specifically exclude abnormal situations like this that result in a TWD.

So what you're saying is "the laws of statistics don't apply to the TWDA because if they do, half my argument is gone". Alright, that makes sense. I guess I have an answer to my question here...

No, what I'm saying is that the TWDA is an artificial measure dictated by arbitrary parameters. The data isn't necessarily representative, and it requires making lots of assumptions (and though many of those involve corner-case assumptions, it still results in a margin of error).

Your argument is the one that seems to be relying on an awful lot of assumptions, and you know what they say about what happens when you make assumptions. Though in this case, you're mainly doing it just to yourself.

Do I need to remind you how easy it is to lie with statistics? Or how flawed methodology can lead to flawed results (like the classic "Dewey Defeats Truman" conclusion)?

Well, ya know, you can think whatever you want.

Yes I can. And it seems clear that you are going to do the same regardless as well.

The reality of it all is that on average, clans with good cards, crypts and synergy will be played more, win more and thus be in more finals and thus win more tournaments.

Similarly, clans with bad cards, bad disciplines and bad crypts will, on average, get played less, not score an awful lot of VPs and just end up randomly winning 2R+F 10-men tournaments.

Yeah, if a clan gets played more, it's obviously going to win more, be in more finals, and win more tournaments. If another 1000 Abomination decks get played, they'll win more, be in more finals, and win more tournaments, but that doesn't mean Abominations will be good.

Although good/bad cards, crypts, and synergy are factors, they're not the only ones that determine what clans people play. As previously noted, clan loyalty (or antipathy) may result in clans being represented irrespective of those factors. Availability of cards (out-of-print status, cost of certain cards) is another. You can't assume that every player will have ready access to every card. There are inherent barriers to trying to make a competitive True Brujah or Temporis deck.

If you want to argue that !Nos, Blood Brothers and Baali are as competitive as Toreador, Ventrue and !Malks, then feel free but I'm just going to continue to roll my eyes.

Have I ever argued that? (Here's a big hint for you, since apparently you need it: the answer is NO).

I'm not saying those clans are the best clans, or tier 1 clans. I am saying that those clans have some interesting things you can do with them. That you can make fun decks with them. That there are some good things about them and you shouldn't just completely write them off.

Izaak wrote:

The TWDA only catalogs the winning deck

Yes it does. However, it is *extremely* unlikely that decks that can ENTER the finals consistently (as you put it) can never win one. The more likely scenario is that these decks are not very good at winning games in the first place and thus don't claim a lot of finals spots and end up not winning a lot of them as a result.

incorrect.
I've seen it happen a lot.
anything can happen in VTES, such as 4 players all agreeing to gang up on 1 other.

Eric and I have had this discussion many times.

The TWDA is the *only* record we have of what decks win tournaments. The fact that it has been meticulously updated for almost 15 years is a feat in and of itself. There is definitely a lot of truth in the data...the facts contained therein.

The largest part of Eric's arguments are opinion based. There certainly are some "reasonable doubts" which can be raised by the nature of the data we are presented. But raising doubts is not the same as contradicting facts.

It is my opinion that *most tournament players* are trying to win and therefore playing the best possible decks they can muster with their cards. It is also clear that the larger the tournament, the more competitive it is so the likelihood of a subpar deck winning dramatically decreases.

This is really the crux of the matter. Perhaps every final table was loaded with Brujah antitribu, Gangrel antitribu and Nosferatu antitribu. We'll never know. But, if those three clans were *just as likely to win* as any other clan, then we would see a roughly even distribution of clans in the TWDA.

It is also my contention that the reason several clans are under-represented is not because they are less liked for subjective reasons, or that they are just good enough to make the finals but fail upon getting there (they'd have to have enough ousting power to make it through the prelims, so why would the final be any different?). They are under-represented precisely *BECAUSE THEY ARE BAD CLANS*. Tournament players play to win, and no-one wants to bring a knife to a gunfight.

That's not to say it's easy to identify *why* a clan is under-represented. Eric surmises quite a bit about this, and some of those are likely to be factors. But the numbers speak for themselves. Some clans are just not winning like they "ought" to be if all else were the same. That's not to say that every clan should have an exact parity with the others. But I think the TWDA is the best tool, if the only tool, to *objectively* measure the larger scope of balance in the game. Anecdotal evidence is also important, but even given all its imperfections, the proof really is in the TWDA.

Jeff wrote:

The TWDA is the *only* record we have of what decks win tournaments. The fact that it has been meticulously updated for almost 15 years is a feat in and of itself. There is definitely a lot of truth in the data...the facts contained therein.

Yes, there is some truth, but it is largely incomplete because there is a lot of important and relevant data that is missing.

The largest part of Eric's arguments are opinion based. There certainly are some "reasonable doubts" which can be raised by the nature of the data we are presented. But raising doubts is not the same as contradicting facts.

And what facts are those? (That's a rhetorical question by the way). The data does lead to some limited facts, but my problem is that people are extrapolating that data and claiming that it applies to a larger set of situations and is more accurate than it is.

It is my opinion that *most tournament players* are trying to win and therefore playing the best possible decks they can muster with their cards.

Most tournament players are trying to win yes, but I disagree that most players will necessarily play the "best possible decks." Instead, I'd say that most players want to win...their way.

There are players who scorn stealth bleed and refuse to play it, even if it is still very effective. There are players who would like to win, but prioritize having fun over winning (and that might include playing fun decks in tournaments). Some players find common (competitive) deck types to be boring and have no interest in playing them even if they are strong. There are LOTS of players who will absolutely REFUSE to play Imbued, even though Imbued is a consistently strong deck.

Paying homage to the MtG "types" of gamers, there are plenty of "Johnny" players who enjoy the challenge of playing unconventional decks (and clans) and winning with cards that other people normally scoff at. Ian Lee is a great example of a good player who tries really hard to build and play something unexpected.

In fact, the existence and renown of the TWDA, has spawned several attempts to win with every clan by certain top players. That alone means that some of these top players are going to be trying to play "crappy clans" in tournaments in order to try and rack up a win. Obviously they're trying to win, but the fact that they're playing these "crappy clans" contradicts your assertion that people are going to be playing the "best possible decks."

In all the tournaments I have played in, I have seen A LOT of players bring really weird, wacky decks (of questionable quality). That makes me think the "fun" quotient is bigger than you estimate and the motivation to "play the best possible deck" is ignored more often than you think. These are the "Timmy" players who are more interested in having fun.

Yes, there are plenty of "Spike" players who only play to win and only play whatever the best deck is. But there are also "Johnny," and "Timmy" players, along with others who don't fit neatly under any of those labels.

FYI: You might want to take a look at this old design article by Mark Rosewater:

www.wizards.com/magic/magazine/article.aspx?x=mtgcom/daily/mr11b

It is also clear that the larger the tournament, the more competitive it is so the likelihood of a subpar deck winning dramatically decreases.

Well, maybe you should be advocating for a larger "tournament minimum size" then for the TWDA, which is something Izaak also seems to want.

I do think it would be useful to separate out Qualifiers and Championships from more standard events. The "play the best possible deck" argument is more accurate when it comes to Qualifiers & Championships as opposed to more casual tournaments.

This is really the crux of the matter. Perhaps every final table was loaded with Brujah antitribu, Gangrel antitribu and Nosferatu antitribu. We'll never know. But, if those three clans were *just as likely to win* as any other clan, then we would see a roughly even distribution of clans in the TWDA.

Haze brought up a really good point, about the multi-player nature of the game. There are arguably some decks which are powerful enough that the rest of the table decides to gang up against it. So in this case, particular strong decks (which are labeled as "table threats") could be underrepresented in the TWDA. Of course this goes into the issue of how you want to define "strong" or "competitive" decks, and if it's just defined by whatever deck won (which is potentially circular logic) or whether you want to factor in multi-player dynamics (which would quickly get messy).

Also, your contention (and that of Izaak) seems to indicate that in the finals, all five decks should be "just as likely to win." That however, is blatantly untrue just from the nature of the seeding process. The first seed player has a slight advantage over all the other players since he actually gets to chose his seating, *plus* he wins ties. (Speaking of which, it would be interesting to see statistics regarding how often the various seeded players actually win).

It is also my contention that the reason several clans are under-represented is not because they are less liked for subjective reasons

::shrugs::

I don't play Imbued or Camarilla Nosferatu or Ahrimanes or Guruhi or Kiasyd or Daughters. That doesn't mean they're bad. I just don't like them or find them uninteresting. I don't really play True Brujah, because I don't have all the rares I need. I don't play a lot of Baali/Maleficia/Striga, because I traded a lot of my rares away. Even though I have a sizable collection, I am often constrained in my choice of clans due to the cards I have available (so I imagine that the players with smaller collections would face even greater constraints). Sure, I could almost certainly borrow the cards I needed for a tournament if I really wanted to, but it's usually not worth the hassle.

VTES has a good portion of players who used to play the RPG or LARP. Clan loyalties and sympathies run deep. Jeff, I know you're not really into the source materials, but I figure that with your passion for the Assamites, you'd understand. The fact that you play lots of Assamite decks (including in tournaments) doesn't mean that the Assamites are a great clan or that they are the best possible decks, you just like Assamites and want to "represent."

, or that they are just good enough to make the finals but fail upon getting there (they'd have to have enough ousting power to make it through the prelims, so why would the final be any different?).

Maybe because the finals IS different because of seeding. Maybe because a deck good at table management can win even during a tie or a timeout? Maybe because there's an advantage to first seed, so "win big or bust" decks are much more likely to be first seed (and thereby win) than a deck that can consistently get to finals but has a hard time sweeping tables?

They are under-represented precisely *BECAUSE THEY ARE BAD CLANS*.

That could be a factor, yes. I at least admit that.

But there are additional factors which you seem to ignore or refuse to recognize.

Tournament players play to win, and no-one wants to bring a knife to a gunfight.

Really? I see it all the time.

In fact I've done it a couple of times myself (when I was bored of my other decks, when a crappy deck was the only one I had available or I forgot my other ones at home, or I really wanted to experiment and finetune my crappy deck so I thought a tournament environment would be useful for feedback and experimentation).

But the numbers speak for themselves.

No they don't. We need context in order to properly interpret those numbers.

Some clans are just not winning like they "ought" to be if all else were the same.

Is all else really the same? I don't think it is.

(Maybe that's why we're at a stalemate. We're relying on different fundamental assumptions?).

But I think the TWDA is the best tool, if the only tool, to *objectively* measure the larger scope of balance in the game.

But there could be better tools. Much, much better tools.

Yes, the TWDA is one of the only tools we currently have. But I think it's better to admit that it is significantly flawed and possibly look for better tools, rather than to simply assume that the TWDA must be correct (because it's all we have!).

but even given all its imperfections, the proof really is in the TWDA.

That's where we disagree. Due to its imperfections, the TWDA is *not* providing as much proof as you seem to think it is.