Latest Articles

Quirks of the Invitational Format

We are on the cusp of our first HEX TCG $100,000 Invitational, but before we jump right into the matches it is important to understand what impact some of the rules will have on the weekend. While we have seen the HEXTCGPro Invitational Tournament as well as the FiveShards Fate Cup work as invitation-only tournaments, nothing has really been a one-to-one example of how the $100k is laid out. There are a number of interesting things going on, and many will have some subtle impacts on the weekend that you might appreciate watching out for.

1) The Day 1 cut to Top 8

Day 1 is extremely short; only 4 rounds of Constructed before we go ahead and end the tournament for one-third of the field. One reason for this likely is that the field needs to be divisible by 8 in order to properly have a draft. You might ask why the field didn’t start with 8 or 16 players. The most obvious answer to that is that there needs to be a check in place during the tournament to account for any players that can’t make it for whatever reason (passport issues and illness being some of the chief problems you see constantly in the TCG world for not making it to a large event). The number 12 most likely was not arbitrarily picked for the number of Qualifiers, as it both keeps the group small so that the event maintains an “Invitational” feel, while also providing enough safety against players missing that the tournament can go forward no matter what.

In any case, at the end of Day 1 we will end up with only 8 players advancing. Obviously 3-1 and 4-0 will lock people in, but things are a little murkier after that. First off, you need to be aware of how the Invitational works when players end up tied by total points at the end of 4 rounds. You can peruse the official rules for the details, but in essence if 9th or 10th place have as many points as 8th or 7th place respectively, those players will play an additional round to see who advances to Day 2. There are certain scenarios where even 4th place and 9th place could both have gone 2-2, but after tiebreakers are figured in the client it will only be 8th pairing off with 9th and 7th pairing off with 10th if their records are the same. If you’ve played TCGs before, you know that tiebreakers can swing wildly in the last round of a tournament (and especially so for a shorter one like Day 1 will be), so only in a few cases will a 2-1 going into Round 4 be totally safe for the day.

For those interested in the math behind what scenarios can crop up where we get tiebreakers, I’ve done a bit of the math that I will share here. For the math here, I’m counting each win as being worth 3 points.

12 Players

Point Totals after Round 1

3
3
3
3
3
3
0
0
0
0
0
0

These are fairly set in stone, since someone has to win a game of HEX and there are no draws.

Standings after Round 2

6
6
6
3
3
3
3
3
3
0
0
0

Again, a lot of symmetry as players with like standings have to be matched up in swiss. Things get a little murkier after this, however:

Possible Standings after Round 3

Scenario AScenario BScenario CScenario D
9999
9966
6666
6666
6666
6366
3363
3333
3333
3333
0303
0000

There is a reason no one makes these charts for 18-round tournaments, as they get too unwieldy after even just a few rounds due to all the potential for deviation. For a short 4-round tournament though, we can still capture things in a somewhat manageable format. At this stage in the game, I’ll go ahead and cheat and tell you that Scenario A is the only one of these where there is no chance of a playoff game happening, so if you see this set come up then you’ll know that the day will end immediately after Round 4. For the rest, let’s press on.

Note: the annotation for the next table will use {X//Y} to represent that one of those two numbers is possible for that particular player, otherwise the table starts to already expand too far.

Possible Standings after Round 4

Scenario AScenario BScenario CScenario D
121212//912//9
9999
9999
99//69//69//6
669//66
6666
6666
66//366
36//36//36//3
3333
3333
03//003//0

I’ve bolded where the potential for tiebreaker games occur. Note that while I’ve highlighted the 6//3 numbers, in actuality if they end up with only 1 win and 3 points it could in fact be ANY of the 1-win players that makes the sudden death game if their tiebreakers are good enough. That makes Scenario B especially interesting, because it is the only scenario in which a player with a 1-3 record can go on to a Sudden Death game and then make it into Day 2. Otherwise, what you’ll see is that the majority of the players who end up going 2-2 out of a 12 person field will move on to Day 2 where all slates are wiped clean. We’ll talk about the implications of that a bit later…

11 Players

Point Totals after Round 1

3
3
3
3
3
3
0
0
0
0
0

I’m sure there is little stomach to see someone getting a bye each round, but you have to work with the logistics of what you are given.

Standings after Round 2

6
6
6
3
3
3
3
3
3
0
0

Again, you can’t deviate too much in the early rounds of a tournament like this.

Possible Standings after Round 3

Scenario AScenario BScenario CScenario D
9999
9966
6666
6666
6666
6366
3363
3333
3333
3333
0303

You’ll notice that this table looks an awful lot like the 12-person table with the bottom row chopped off (Two points for Gryffindor, because that’s all I did!). While the bye might appear unfair to the individuals who don’t get it, the overall tournament structure will march forward much the same as it would otherwise.

Possible Standings after Round 4

Scenario AScenario BScenario CScenario D
121212//912//9
9999
9999
99//69//69//6
669//66
6666
6666
6666
36//36//36
3333
3333

Scenario A and C are essentially the same, but a few small deviations from our 12-person table occur. Under Scenario B, there no longer exists the possibility for a 1-3 player to make it on to Day 2. No one started 0-3 in that scenario anyway, so all games are still live for every player up til the end under those conditions. Scenario D also throws in a wrinkle that if we see the conditions for Scenario D emerge in Round 3, then we can guarantee that there will be a Sudden Death game after Round 4. Take your bathroom breaks accordingly.

10 Players

Point Totals after Round 1

3
3
3
3
3
0
0
0
0
0

No byes, easy-peasy

Possible Standings after Round 2

Scenario 1Scenario 2
66
66
63
33
33
33
33
03
00
00

10 players give us a pair down already in Round 1, so we start getting more complicated from here.

Possible Standings after Round 3

Scenario 1AScenario 1BScenario 1CScenario 1D
9999
9966
6666
6666
6366
3363
3333
3333
0303
0000

 

Scenario 2
9
6
6
6
6
3
3
3
3
0

10 players will result in a far high rate of Sudden Death matches. Already, under Scenario 1A and 1C we can guarantee that there will be a Sudden Death match, and under Scenario 1B we know there will be one if not two!  Another important thing at this stage is that under Scenarios 1A, 1B, and 1C we can go ahead and guarantee that going 2-2 will make it for sure, so every player with 2 or more wins up to this point now can breathe easily.

Possible Standings after Round 4

Scenario 1AScenario 1BScenario 1CScenario 1D
121212//912//9
9999
9999//6
9//669//69//6
6666
6666
6//36//366//3
3336//3
3333
03//003//0

 

Scenario 2
12//9
9
9//6
9//6
6
6
6//3
6//3
3
3//0

Scenario 1A and 1C are locked into having a single Sudden Death between two 1-3 players. Scenario 1B will have at least that, plus an additional match if the matchup between the 1-2 and 0-3 player this round ends up in favor of the 0-3 player (meaning even at 0-3 you can rally back!). Scenario 1D and Scenario 2 could end up with no Sudden Death matches, or it could end up having two if everything falls the right way (or wrong way, depending on your perspective). Both of those situations have a lot of moving parts to pay attention to.

2) Draft to beat mediocrity!

So now we are onto Day 2. Eight players will come into the Swiss draft, but only four players will continue on towards the Top 4 Constructed. Thankfully, we know that Swiss drafts make for very easy cuts: One person will go 3-0, three people will go 2-1, three people will go 1-2 and the last person will end up 0-3. It is important to note that there is a monetary difference between all the places from 5-8, so unlike Day 1 if you find yourself in a position where you can no longer advance there is still an incentive to keep playing your best.

However, we do see that there could be some incentive to approach this draft with a slightly different strategy than other places. In your typical 8-man queue in the client, your EV maximization comes mostly comes from trying to draft well enough to win but also to just keep doing well as an average since you still get paid for placing in the Top 4 and can raredraft if you please. Invitational Qualifier Top 8 drafts were very different, as there was a HUGE payoff for being first place, incentivizing players to go for 3-0 strategies and for identifying when archetypes were left clearly open for them to move into. Being 2-1 in an IQ draft meant you were above average, but no flight or qualification meant you couldn’t settle for just above average and had to go for the total victory. The Invitational itself will be even slightly different from that, as while going 3-0 is good it isn’t actually very different from going 2-1 since the only payoff is that you face a lower seeded player. Heck, that might not even be a good thing if you suddenly find yourself paired down to the constructed specialist who just squeaked in!

For the Invitational, the goal is thus to 2-1 the draft to move on towards the final portion of the event. While there isn’t a monumental difference between “I want to draft well” and “I want to draft a 2-1 deck”, there are incentives to stick to better archetypes even if they are overdrafted than in a typical draft, especially since the change to 17-card packs that happened in the middle of Armies of Myth limited means that drafts have never been deeper with playables to fill out your deck. We don’t have a ton of data on exactly which archetypes might be those better-performing ones, so while some players might have teams or groups to pull data from and make good judgement calls, all we have are the 5 IQ Limited Top 8s that were posted to HexMeta:

Draft wins per IQ competitor

RWBDSWBSDRTri-ShardRSBW
33132100
320320
12021
10011
10010
0001
000
00
0
0

Now, this data isn’t pure for a number of reasons. First, the Top 8 of each IQ was single-elimination instead of Swiss, meaning what normally could be a 2-1 deck might have been knocked out in the first round. So while an “average” point total in Swiss would be 1.5 wins, in Single Elimination it gets knocked down to .875. Second, is that some of these results came before we had the balance changes which both changed the number of cards in a pack (Hello synergy archetypes like Spiders!) and changed the health totals of various Champions (Goodbye Cressida!). Those things being understood, this is still the best public data set we have to work with, so we can take a stab at trying to produce some numbers and analyze our findings.

Average wins per archetype

RWBDSWBSDRTri-ShardRSBW
Average wins0.901.000.171.381.200.5000
% of decks “Above Average”50.0%42.9%16.7%75.0%80.0%50.0%0.0%0.0%

% of decks “Above Average” is the interesting stat to look at here. This accounts for what percentage of the archetype’s decks finished in the Top 4 of their respective IQs. Small sample size means that I would not throw around the 80% number for DR or the 16.7% for SW with extreme conviction. However when I combine the data here with my own personal tilt and I can certainly believe that many of these numbers are at least ranked in the right places. It will be interesting to watch and see if players try and “force” BS Spiders or DR Aggro in an attempt to simply meet the 2-1 deck requirement, or if any player will go for SW Coyotle or any of the non-traditional shard pairings and try to prove the data wrong.

This creates are other small effects as well. Hate drafting isn’t nearly as strong here as it might typically be. (You always hate draft the Brood Baron in an IQ Top 8 since all it takes is one loss to the Brood Baron player and you don’t qualify. You can afford to lose to the Brood Baron player in the Invitational, however, as long as you are making your deck strong enough to beat the other players you will face). You can figure your opponents are using this data also, so if you are counting on no one drafting SW but don’t want to yourself, you can at least use that knowledge to maybe choose the Ruby card over the Wild card if you are in RW since there is a strong chance that Ruby won’t come back but a solid possibility that the Wild one will. Hopefully we can get some interviews with players afterwards, as the theorycrafting that can go into an event like this will surely inspire some interesting strategies!

3) Someone’s Gotta Win This Thing!

Finally, we come to the Top 4. We now have no-holds barred, single elimination Constructed matches between the remaining competitors to see who will be taking home the big money and the title. We now have an interesting dilemma on our hands though, as while the goal in the Constructed portion of Day 1 was “3-1 is good enough, and 2-2 is probably fine”, here in the Top 4 you clearly want to be the deck that can win it all as the prize is fairly heavily slanted towards the overall winner. Some decks are geared very well towards doing “good enough” and give a lower variance for making it through Day 1, but those same decks might not be quite as good at actually being able to win tournaments outright once they get there. As an example, I’ve pulled the data from IQ #12 which was won by SaDOlution with DW Banks. Some people will drop early so we don’t perfect information like we might want, but we can still calculate the average wins in the tournament and get an idea of what baseline can be used.

Best performing decks (among those with 5+ entries)

DeckAverage Wins# of copies
Mono Ruby Aggro5.005
BD Banks4.605
Benvolio Burn4.2915
DW Ramp4.158
Mono Blood3.7649
Cressida Ramp3.6221
Winter Moon Control3.4346
BD Non-Banks3.3611
Azurecannon2.7818

Field Average – 3.54 wins

Some notes on each archetype:

  • Urgnock has one person with only a single win but everyone else did quite well with it. Only 5 total copies though, so small sample size in effect.
  • BD Banks had 3 people do really well and 2 people do really poorly. Small sample size again
  • Benvolio Burn had a whole lot of people just outside the top 16.
  • DW Ramp was raised considerably by SaDOlutions win. Remove him and the deck is below average.
  • I did not break out Mono B by champion as it was overwhelmingly Kranok. It was the most popular deck and still was slightly above average.
  • Cressida Ramp was as close to average as any deck got.
  • Winter Moon was surprisingly below average. Could be that the deck is too tough for an average player to pilot, but also possible that people were starting to figure out how to beat it in the metagame.
  • BD without Banks also was below average. Pretty obvious that you don’t want to do this anymore, especially after Revolution’s Ruby Cup win.
  • Azurecannon really stunk up the place. Woof.

This doesn’t tell the whole story, however. Standard deviation will let us see that some decks might have had more erratic swings in their placement. If the goal is to maintain a 2-2 or better, we might want to look for low-devitation decks that also performed well, while if we want to win a tournament outright instead we might be crossing our fingers with a deck that had wilder swings in placement.

 

DeckStandard Deviation
Azurecannon1.93
Benvolio Burn1.98
Winter Moon1.99
BD Non-Banks2.20
Mono Blood2.24
Cressida Ramp2.38
Mono Ruby Aggro2.55
DW Ramp2.80
BD Banks3.34

Field Average – 2.21 STDEV

For those not familiar, the basic gist is that the lower your standard deviation then the less you had outlier performances. Again, some thoughts:

  • Azurecannon was consistent…ly bad. So many 1 and 2 wins entries, and only one person with 6+ wins in the entire set of 18 decks.
  • Benvolio Burn was the biggest surprise out of all this data to me personally. Decent number of pilots, Top 3 performing deck with a much higher than average win rate, and incredibly consistent to boot in terms of where people placed.
  • Winter Moon was full of 3 and 4 win players, so lots of people who broke even and only a handful of people who scrubbed out hard or finished really well.
  • BD Non-Banks and Mono B were pretty much on the same consistency level as the field. For those that have played these Blood decks for a long time, this likely comes at no surprise.
  • Cressida was slightly wild in terms of placements, had a good number of 6+ win showings and then a bunch of 3 win players.
  • Mono Ruby Aggro was let down by the single 1 win performer. Take it away and the deck spikes to a 1.41 STDEV. Again showing the power of small sample sizes and why you can’t use data like this to blindly make decisions.
  • DW Ramp was the opposite, where SaDOlution was really what floated the deck to the top, as removing his entry brings the STDEV down to 2.11.
  • BD Banks was nothing but players who crushed it and players who scrubbed it. Some might consider it a surprise that the change in champion power can make it vary so much compared to the other Blood decks, but when Rutherford Banks is good he is VERY good, and when he is bad it is like you don’t even have a power. This is in stark contrast to someone like Kranok who gives you pretty much the same performance across most matchups.

Naturally you can’t just extrapolate data like this forward and expect it to work perfectly. The metagame in HEX moves in an alarming fashion sometimes, so if you bank on playing a well-performing deck from the previous tournament you might find that the fundamentals of the format have changed on you when you weren’t looking. Still, this is all data the the Invitational competitors will be taking into account when they look to make their deck selections. Each one will have to weigh many questions in their mind, as the decks that make the most sense for getting into Day 2 may not also be the decks that make the most sense for winning the whole tournament once you get there. At the end of Sunday though someone will have won the tournament. Whether that person dares to defy conventional wisdom or tightly walks the safer path, you can be sure that it will be a weekend that will live in HEX history forever.

1 Comment on Quirks of the Invitational Format

  1. Thanks for the breakdown!

    This (and the hextcg.com metagame primer) will make the invitational more interesting to watch for people like me that haven’t really payed close attention to the recent constructed meta game

Leave a Reply

%d bloggers like this: