Sunday, September 13, 2015

Tournament Points and Winning

With the Vassal Armada tournament coming to a close, I've been considering the point system used in the tournament, and it's effect on the meta that I've seen.  I've been trying to figure out the "best" strategy to build for and strategize around in a tournament - do you go for the table every time or do you play a long game?  With these thoughts in mind, I crunched a few numbers, specifically tournament points and overall Margin of Victory.

Suppose you knew that with your list you could win every game, but only with a MOV sufficient for an 8-2 win.  Do you bring this list to a tournament, or do you bring something more aggressive, that will likely see a 10-0 win, but doesn't guarantee you a win every time?  Actually, it depends on the number of players in the tournament, and how many of them you expect will be bringing a 10-0 list.  Consider - if any 10-0 players are in the game, they will be matched with one another in round 2, and both will not win.  While an 8-2 list will likely be matched up against another list with a smaller victory.

For that matter, if you do decide to go the 10-0 route, how many points can you be willing to offer up to your opponent?  For sake of argument, I will be using the 300pt tournament rules, as the MOV for 400pt games is not available at the time of this article.

10-0 vs 8-2 in 3 Round Tournaments

The short answer is, 10-0 looks better the more players are present in the tournament.  The long answer says that the magic number of players is 9, where suddenly a 8-2 list isn't good enough anymore to outright win.  Consider the competition between a list that will always go 8-2 no matter the opponent, and a list that will go 10-0 unless playing the 8-2 (then will suffer a 2-8 loss).  If there are 9 players in a 3 round tournament, the 8-2 player will have a 1/8 chance of playing the 10-0 player in the first round, and a 1/8 chance that the 10-0 player will have a bye, and be possibly playing him the second round.  Aside from those two opportunities, the 10-0 player will more likely be paired up against other 10-0 players, earning 10 points each round.  

By the end of Round 1, there will be 4 undefeated players, and 1 bye player with an 8 point record.  The 10-0 player will have 10 points, and the 8-2 player will have 8 points.  Most likely, after Round 1 they will not play one another unless no one was able to earn an 8-2 or better victory.  After Round 2, the score will be greater, with the 10-0 player having 20 points, and 8-2 player with 16.  Now there are at best 2 undefeated players, but a chance for 3 (if the lowest point total undefeated player defeats his opponent with a loss from Round 1).  If the 3rd undefeated player has greater than 16 points (say a 8-2 bye in R1, and a 9-1 win in R2), he will play the 10-0 player for the final round, and the 16 point player will be up against one of the 1-1 players.  Victory 10-0 with 30 points, to 8-2's 24.  

It is worse for 8-2 with more people, and guaranteed multiple undefeated players.  At 16 players, there is are most likely 4 undefeated players going against one another Round 3, and the odds that 2 of them won their games by better than 16 points is fairly good.  If both of them swept, all 8-2 can't even win outright with another 8 points, even if they have a 5-5 tie!

More than 16 further exacerbate the issue, with 32 players containing 8 undefeated players going into the 3rd round!  In the most recent Vassal Tournament, we had a total of 7 undefeated players going into the 3rd round, with 27 total players.  One undefeated player only began Round 3 with 12 points, and was unable to get 1st place (or 2nd, or 3rd) if he got a 10-0 win, and everyone else got a 5-5 tie.

But this isn't exactly news.  More points is better is a fairly obvious lesson.  What we can take away from it is the way to build a list.

5 Round Games

Of course the trend continues into 5 round games as well.  33 player games will also be likely to have 2 undefeated players after all 5 rounds end, with the number going up the more players are participating.  Further, a 8-2 player will have and 32 vs 40 after 4 rounds, meaning even if they meet in the last round, the 8-2 player will not win the tournament by defeating the 10-0 player by winning by a score of 8-2, finishing with 40 points to their 42 points.

What Does This Mean For List Building?

It means that you need to keep your eyes on the margin of victory, and how many points you are willing to sacrifice to get that 10-0 win.  Or more accurately, what route you will be taking to get the 10-0 win.  And that has everything to do with your build - the ships you take and the objectives you select as Player 1 or as Player 2.

91 points is the threshold for 8-2 (3/10ths of your list, plus 1).  151 is the threshold to 9-1 (half your list, plus 1 point), and 221 is the threshold for 10-0 (Giving you 79 points to give to your opponent, if there is a table.)  But these numbers do not assume any objective shenanigans relating to points, which can work to your advantage, or detriment.

Consider the GenCon Special

The now famous, or infamous list that won the US Nationals, the list was based around a single VSD with Screed on board, plus 3 GSDs with ACM, one with Demolisher.  How many points do each of these come to?  The VSD - 101 pts - enough to drop to 9-1 if lost.  Mark that one "do not lose" in big neon letters.  The rest?  Demolisher is 73 points - safe to lose.  The other two are 63 points (again, safe to lose - as long as only 1 is lost).  Combine any two of the ships (aside from the VSD) and it is at most 136.  Again, safe to lose and still get a 9-1 if you table your opponent.

It isn't just that the list itself is very much bent at getting the "table", but it is about the brutal efficiency of it.  It can lose anything (but the VSD, the tankiest ship it has) and still get 10-0.  It can lose any 2 ships, as long as the VSD survives, it gets 9-1.  Those are some very good margins.

Consider the Dual Guppies Type Lists

Right now, the big name for rebel lists is two AFIIs, with various upgrades, and other things thrown in for good measure.  Here is the catch - the AFIIs, each one is going to be kitted out at significantly more than 79 points.  That means losing one AFII brings you into 9-1 territory.  Lose the "other things" that go along with the AFII, and you are likely looking at over half your point total, giving you at best the 8-2 outcome.

The Unfortunate Fate of Squadrons

Squadrons suffer the worst in all of this, in that they are relatively (compared to ships) easy to destroy.  Plus, when all ships are destroyed, they don't prevent a "table" of 300 points from being given to the opponent.  Two strikes already against them.  But losing fighters and a ship compounds the points problem.  For the Imperials, losing a GSD with ACM (63 points) means that anything more than 2 TIE Fighters (16 points) is a drop to 9-1.  A couple A-Wings and anti-squadron firepower, and 2 TIE Bombers drop, and you are looking at a less thrilling win.

Conversely, if you are working with dual AFIIs, and counting on both to survive, losing a few A-Wings (or Y-Wings / X-Wings / B-Wings) is less critical in the grand scheme of things.  Lose a CR90 with them?  50 points at best - that is 2 of any Rebel ship, or 3 Y-Wings if the CR90 was cheaper.  Not the end of the world.

What About Objectives?

So far all we've been talking about are the Armada equivalents of Spherical Cows in a Vacuum.  What objectives are being played definitely change the equation.  And also change which objectives you'd be willing to give to an opposing player for the chance to table them.

Consider for example Intel Sweep - definitely the easiest to consider because it is just straight 75 points to one side or the other.  Not enough points to sway the final score for your opponent if you can table your opponent and lose nothing or your own but the objective.  But, also enough to bump anything in the 9-1 range directly into the 10-0 range if you pick it up.

Minefields, Hyperspace Assault, Fleet Ambush, and Dangerous Territory should be easy enough to look at as well.  The first three gives no points, and the final one should result in a 45-45 split of the possible tokens.

For any of the 15 points per token objectives (Fire Lanes, Superior Positions, Precision Strike), 5 tokens gives 75 (same as Intel Sweep) and 6 gives 90, enough to upset a perfect table sweep.  Enough not to give up if the opponent is build for the objective and you are emphatically not.

For Contested Outpost, controlling the outpost for 3 turns nets 60 points, and 4 turns nets 80, and 5 nets 100.

For Opening Salvo, it does nothing at all good for someone that is going for a table.  Avoid this objective like the plague that it is.  On the other hand, if you are running a counter-meta designed to knock out 10-0 built lists - throw this into the mix.  It's a trap!

Most Wanted and Advanced Gunnery are wildcards into the mix.  On one hand, they can make an inexpensive ship into a "do not lose" ship, or a bigger ship into a disaster if it is lost.  On the other hand, they give you extra points for tabling your opponent.  The important thing with both is to ensure that you are getting maximum value for the points over what your opponent picks up.  It may still be worth it to sacrifice the 88pt (39 bonus) CR90B w/ Tantive IV and Raymus if doing so you can drop 200+ (85 bonus) points of VSD.

So What Is Worth Sacrificing?

The easy answer is anything is expendable if it is worth less than 80 points, and 2 can be if together they are less than 150.  So the sweet spot is right around the 74 point range.  GSD1s w/ ACM are 63, and with Engine Techs are still under budget.  CR90s are rarely more than 50 points each, though it is basically impossible to get 2 under 80 points (naked CR90Bs would work).  But 3 under 150 is definitely doable.  Nebulon Bs are also inexpensive for this  role, with Salvation and a Turbolaser upgrade coming in below 70.  A naked AFII-B also can work, coming in at 72 points, as does a naked VSD at 73 points.

Consider the Following Rebel Squadronless List at 298 points
1 • Assault Frigate Mark II A - Mon Mothma - Raymus Antilles - Advanced Projectors - H9 Turbolasers - Paragon (137)
2 • Nebulon-B Support Refit - Veteran Captain - Nav Team - Salvation (65)
3 • CR90B Corellian Corvette - Overload Pulse (47)
4 • CR90A Corellian Corvette - Leia Organa - Jaina's Light (49)
5 • Objectives - Advanced Gunnery - Fleet Ambush - Dangerous Territory (0)

Again, you have a ship purpose built that you cannot afford to lose, but with a strong defensive build around it.  All other ships are expendable in the end.  No two ships when lost will knock you out of a chance for a 9-1 win.  The objectives work with the list - Advanced Gunnery is a trap for anyone dumb enough to go toe to toe with your Paragon.  Fleet Ambush lets you bring your forces to bear on half their ships at once, and negates any deployment advantage the opponent would get for bringing fighters.  Dangerous Territory gives you a point advantage unless the opponent wants to crash their ships through obstacles, making taking out their ships that much easier.

So what do you think?  Something worth playing around with?  Do you build a list with high variance on points with your objectives to compensate for sacrificial ships?  Do you build a list to table your opponents?  Do you have 1 ship as a sacrificial lamb, and cannot afford to lose the rest?


  1. Great Article. Thanks for taking the time to write this up.

  2. Great observation regarding keeping your expendable ships within a specific points range. I'm not convinced by the 10-0 vs 8-2 approach - it seems predicated on the fact that you will be climbing over a pile of mediocre players to make it to the pointy end of the competition. I'd probably be more concerned about how to win against the really good players and structure my gameplan that way. Going 10-0 all the way to the top table is somewhat meaningless if you lose the final 0-10 :)

    I'd actually like to see FFG revise their tournament scoring system - I'm not a great fan of it as it stands. As a long time player of FoW I like their approach that measures the victory then looks at the losses suffered by the winner. It means tabling an opponent is not necessary and you can have a near bloodless game that still earns the victor full points. I feel it would encourage more diversity in fleet builds but it would require explicit victory conditions (which would probably be a rewrite or alternative set of objectives).

  3. So does this mean that squadrons are to be avoided if you want to win tournaments? This sounds like a pretty bad outcome.

    1. Definitely not the case, though it is an easier route. Squadrons can definitely make their way onto a list, but they need to be run with the scoring in mind. You can't make carrier ships huge point sinks and not give them better than average defenses.

      I will have another article up this coming week integrating squadrons into the list.