Wednesday, March 13, 2013

Battle Points in the 40k GT? Simultaneous Tournament Theory

The NOVA Open is a win-loss, win-path, bracketed GT.

There are a number of reasons for this, ranging from fairness to sportsmanship. I'll cover those briefly, and assure the reader first and foremost that none of this is changing.

Reminders on Why Win/Loss, Win-Path, Brackets
40k GT's are by their very nature relatively non-competitive events. They are often full of highly capable players, but they are NOT competitive.

Competitiveness in a tournament setting is determined largely by the average delta between the average player. If you have a basketball competition of 6 players, with - at their prime - MJ, Kobe, me (trust me, that's a lol), some random HS basketball player, a 6 year old, and a 5 year old in a wheelchair all involved ... the tournament is not going to be competitive at all. The winner will probably be MJ or Kobe, but if pairings are random those two may play each other at any point in time ... and even if they luck out and face each other in the final, that's going to be one of the only competitive games ... well, again, unless all the pairings work out by dumb luck just right.

Furthermore, the amount anyone wins any of those games by is irrelevant and completely meaningless. We'll get back to why there.

What's important is: in a 40k GT, we don't have a national ranking system that is fairly and equally applied to all attendees, *and we never will.* There are some stellar locals who attend each of the major GT's, and ONLY those major GT's, for any number of reasons. There are relatively average travelers who occasionally win a GT or place highly, but routinely lose when paired against numerous unknown but superior players. Ranking systems would only meaningfully impact players who were always gaming ... and even then, they so rarely play each other that the standings would be quite random in their import.

AS a result of this, first round pairings in any GT setting are *generally* random, or based on some arbitrary criteria. In this unavoidable fact lies the root of evil in designing a fair tournament when leveraging margin of victory.

Let's go back to our 6-player Basketball tournament, and random pair the first round. Note - the person who ultimately wins the event will probably do so if you go sufficient rounds for it regardless of how you score it. Nevertheless, this will help illustrate the point.

Round 1 of this basketball tournament, I go up against the handicapped wheelchair-bound kid in basketball. You can trust I'm pretty bad, but you can also trust a relatively athletic 31 year old in good health is going to probably crush the handicapped kid (don't hold this against me, please!).

Let's also have the random pairing in Round 1 put Kobe up against MJ.

A couple of things have now happened, both of them unavoidable. One, I probably have a MUCH bigger victory on paper than MJ or Kobe (whichever wins). In a Win/Loss WinPath tournament, the amount by which we crushed a totally randomly determined opponent is more or less irrelevant. Subsequently, we won't be paired based upon that level (For, if we did, I would certainly dodge Kobe/MJ, and face some other person who beat up on a relative scrub ... sounds good to me ... or at least better to me). If the event goes 3 rounds, this will become *somewhat* meaningless as long as my points allocations aren't totally whacky ... chances are I'll go up against MJ or Kobe in the last round at latest, and get appropriately beaten by the best player. BUT, what if the event ends after 2 rounds, and the winner becomes the person who won by the net most points over 2 games? Heck, any # of things might happen, and there's no way to tell who really is the best by this criteria.

That's ok though ... often times you end up with a pretty good eval of a great player in a margin of victory setting ANYWAY; random first round pairings can still knock out certain opponents in a win/loss event also who otherwise *should* be considered high up there. Of course, they can battle their way back up the charts regardless, but that's another story.

The other thing we do with win-pathing that's important is management of putting people together who are having similar days. Win-Path pairing does the following:
If you win round 1, you play someone else who won round 1. If you lose round 1, you play someone else who lost round 1. Since the first round is not seeded and we have NO WAY of telling whether your margin of victory was indicative of skill or just a dramatic skill difference, this plays no role in subsequent pairings - only whether you won or lost.

Now, so far you're all thinking "duh that's like any other pairing." Here's where win-path changes things up.

If you subsequently lose round 2 after winning round 1, you play someone else who lost round 2 after winning round 1. You do NOT play someone who lost round 1 and won round 2, even though your records are the same.

Why is this meaningful? Emotions play a HUGE role in how we interact with others. The substantial majority of the NOVA's attendees have responded feeling there was a much better sportsmanlike feel to the event than many other events they've attended, in survey response. Nothing's perfect, but there's a correlation here. With simple record-based pairings or with margin-total based pairings, you can often run into a situation where you have people paired off who are experiencing very different days. A person who loses their first round after months of planning is probably in a similar mood as someone facing the same. Contrary to what might seem common sense, this more often leads to empathy than to a combination of sadness. They both feel the same way, and in that have an instant bond that is visible when you walk around and witness games. This increases in complexity as the rounds go on.

Someone who LOSES game 1, and WINS game 2 is in a MUCH BETTER MOOD than someone who won game 1, and lost game 2. Even though their records are the same, their moods are dramatically different ... one just recovered from a loss, the other just dropped out of undefeated. You do NOT want these types of players to play each other unless you have some kind of fair ranking/seeding system that determines pairings. Again, we do not ... we cannot ... have a meaningful and accurate ranking system in play at national GT's (you can rank in leagues and such, and this solves other problems entirely).

Winpath pairing over a tournament with sufficient rounds achieves net results one way or another, but also ensures you face players up against each other who are experiencing the same emotional swings as their opponents. Extrapolate this further - someone who loses 2 straight games and finally wins 1 is not going to be in the same emotional state at all as someone who loses, wins, and then loses again.

The final component of the NOVA format is bracketing. After 4 rounds, players' records are reset, and they are bracketed down into sets of 16 based upon their record. Their margin of victory is at this point used to establish their seed w/in their bracket. So, the final 4 rounds of the GT along the Win-Loss front are effectively a series of 16 x Sweet 16's (or subpart thereof). You don't have to do this in a smaller event ... you just always break after a certain # of rounds into the appropriate # of smaller sub-brackets.

Why do we do this? Well, there are a couple of reasons. First of all, it adds a little fun and intrigue to the brackets. Second of all, and perhaps most importantly, it further reinforces sportsmanship and competitiveness. Extrapolating from the above on win-pathing ... in a standard GT over many rounds, someone may start 4-0, and then lose 2 straight against stiffening competition. Rest assured, going into Round 7 this guy is NOT in a good mood. He thought he was in it, and now he's way out of it ... dropped to a record that in HIS eyes is now 'bad' at 4-2. Without bracketing and win-path pairing, this player might very well pull randomly or by BP-matching a player who started the event 2-2 (and thus felt relatively out of it and ready to have a good time), and is VERY happy after winning rounds 5 and 6 to yield 4-2. You do NOT want these players to face each other! Not only is it not overly relevant to the results at this point, but you're unintentionally brewing powderkegs. Often TO's never think about this sort of thing.

Finally, bracketing allows us to reward people for being solid Generals RELATIVE to their combined list/skill levels. If you want to bring a more fluffy list to the NOVA, you CAN without penalty; YES you'll take your lumps early on, but you'll wind up more or less in a bracket of players who are your net peers. You'll then have full freedom to compete for a Generalship award in a 16-player bracketed tournament against lists and player types that very closely mesh with who you are, and with the list you brought. Rest assured, there's just as nice a prize and just as fun a time in the 0-4 bracket as the 4-0 bracket.

So what are the downsides? There are a variety of opinions on this, but I'll try to summarize the ones I most commonly have heard:
1) Win-Loss requires too many rounds in a Major GT, which can be exhausting. 8 Rounds is too much for some folks.
2) 40k is a match-up / rock-paper-scissors game in the eyes of many players; as a result, bad matchups and unlucky losses can happen, and you shouldn't be eliminated from Generalship competition by getting unlucky with dice or matchups.
3) 40k should have ties of some sort, because you shouldn't force a winner or loser to be declared off a close game.
4) Some people just really like margin of victory and think despite the randomness of match-ups, big wins mean more.

Well, what if we could solve all of these, without changing the format on the W/L front at all? What if we could also make scoring your rounds easier and faster in the process? What if we could add Battle Points as well, margin of victory if you will, and award a Generalship prize based upon THAT in parallel to the Tournament Champion prize for the winner of the top bracket?

Enter this year's scoring thesis in draft.

Adding Battle Points to the NOVA 40k GT
The NOVA Open typically has tiered victory conditions (win by Primary, or tie and win on Secondary, or tie and win on Tertiary). This year, some rounds also have more cumulative victory conditions (Add value you earn from Primary, Secondary, and Tertiary ... compare to value your opponent win ... highest wins). Both types of mission have their ties broken by Tiebreaker.

So, (these numbers arbitrary for purposes of explanatino) what if winning on Primary (in tiered missions) or "x" margin (in cumulative missions) earned you 10 points, winning on secondary or "y' margin earned you 7 points, winning on tertiary or "z" margin earned you 5 points, and going to Tiebreaker earned BOTH players (win or lose) 3 points.

Well, now even though winpathing and win/loss are in full effect, you're ALSO earning Margin of Victory based battle points in every round. Additionally, for the Battle Points track, Ties are now in play.

Additionally, similar but different from last year, Battle Points will stop being scored at the end of Round 6. Renaissance Man will be scored similarly to present (50% appearance, 40% win-percentage, and 10% battle points). The highest Battle Points scorer will earn a prize equal to the Bracket 1 winner for BP, effectively an alternate Best General. Additionally, any player who prefers Battle Points will not only be able to consider Tiebreaker wins/losses as ties, they actually will be for purposes of that track of the event. None of this affects the W/L lovers or the sportsmanship-enhancing win-path pairing.

So what we'll do is ... at the end of Round 6, there'll be the option to remove yourselves from the GT. Anyone who doesn't enjoy playing 8 rounds, who wants to sleep in on Sunday, who wants to tour DC, who wants to check out the seminars and paint competition, who wants to check out the Basic Black Range Day (yeah, on Sunday, they're taking those who fill out the Seminar to the NRA firing range and museum) ... well, they can do any of those things, and have it not impact them in the GT.

That said, unlike last year, anyone who wants to play a full 8 rounds, win or lose, will continue to do just that. And, Bracket 1's winner - Tournament Champion - will still be fully and normally determined. Plus, so will all the Bracket Winners from 2-16.

Basically, what this change will do is add a third track to the NOVA Open within one simultaneous tournament. Before there was the Renaissance Track, like the oldest school GT's, where appearance score was critical to determining who the best overall hobbyists really was; and there was the Championship Track, the modern style GT, where winning and losing alone determine the best tactician / general for the weekend. Now, those tracks are unaffected, but there's ALSO a Margin of Victory / Battle Points track, like the more traditional BP based tournament.

Additionally, instead of trying to bastardize all of these things into a single track of awards and scoring, they're left independent of each other ... for pundits and attendees alike to value each style the way they personally see fit. If you want to call a Tiebreaker-determination a tie, and view the highest BP as the true best general ... well, you can. If you want to see w/l as the way to go, you can! If you want to see painting and modeling as critical components of what makes 40k the game that it is ... you can! Heck, as per normal, you can even compete for real generalship prizes without having to take a hardcore, WAAC army or playstyle ... find your way to your appropriate zone, and compete on even footing against your peers for just the prize you find most interesting.

/deep breath

Input and questions welcome!


  1. Love it. Might be the best idea ever. Might I make a suggestion, though? If you are going to allow the same player to win the W/L portion and the BP portion (or not allow it), announce that before the GT. Wouldn't want someone upset because they won both but couldn't win both and not know it before the event.

  2. I have been running tournaments at my FLGS like this for the past couple years now. It works out very well for all involved. It tends to create very interesting games at all rounds past the first. There is a pretty defined winner, and the lower places are easily established. It also makes for easier math and the incorporation of "soft" scores.

    The first round does become either a total ROLFstomp or it becomes a very good first game for both players, but I agree with you that ranking systems simply cannot be created to use for first round pairings. Even at a local store with ~20 guys, you still have the occasional new guy or traveler type.

    I'm pretty glad to see NOVA incorporating some new elements into the format. I'm interested to see its evolution.

  3. Amazing idea! I am glad you shared with us. Truly ground breaking stuff.

    1. Actually the scripts and scoring I was running at the Indy Open did 90% of this stuff. My method of generating pairings just happened to take win path into account. And my break down for overall was more like 50% sports, 30% wins, and 20% margin.

  4. *and we never will.*

    Only a Sith deals in absolutes.

  5. This is sort of like things coming full circle. The idea of basic simultaneous tracks existed in the old GT/RTT system for years. People then got all bent out of shape over which award was more important: Best Overall (Ren Man) or Best General (Battle Points) - often pushing to do away with the former. This was a definitive factor in birthing the W/L format that is prevalent in today's Indy GT circuit. So now you bascially have an amalgamation of both the old and new formats in one. AdeptiCon is going the "bastardized" route for sake of simplicity and the fact that we have multiple events that address the different styles of play, but ultimately our Championships event is built on this same foundation: Warmaster = W/L/D path, Master Tactician = Battle Points/Honor path and Best All-around Showing is the Ren Man (an old idea refined by the first iteration of the NOVA format, in that is truly equalized contributing elements of the award) - we just have a different cutoff/qualification process that doesn't require brackets. I wrote a bit about it here:

    6th edition screams for something like this if you wish to stick closely to rulebook missions. The inequality of Primary and Secondary objectives in and of itself makes past formats non-functional. I also am glad to see larger events landing more in the middle in terms of appealing to different types of players and differing playstyles - it can only be good for all events and the community as a whole.

    Like I mentioned to you earlier Mike, I am somewhat wary of the complexity of overlapping formats and its appeal to the average gamer. I almost had to post tl;dr to the above! Our missions, in particular the 40K Team Tournament ones, got overly complex over the years and it took us a little time to see/recognize it and begin to make the necessary corrections to scale it back.

    That said, of course I like this direction and it will be interesting to see how it all settles once you have the experience of running it under your belt.

  6. great idea mike, i see it catering to all of the tournament styles wrapped up in one . Ill be watching closely to see how it works out. Very bold idea given the fact that you have multiple events going on and there is still a great amount of thought into the tournament scoring. Will this pertain strictly to the Open GT or will it also be implemented for the Invite as well since in the past both tournaments tracked each other regarding best General and renman.

  7. "Finally, bracketing allows us to reward people for being solid Generals RELATIVE to their combined list/skill levels. If you want to bring a more fluffy list to the NOVA, you CAN without penalty; YES you'll take your lumps early on, but you'll wind up more or less in a bracket of players who are your net peers."

    Hells yes! Beer + Blood Angels = WIN!


    - said no one intelligent ever