In recent months there have been a fair number of discussions on how tournaments should be organized and run to best reflect the competitive environment they seek to achieve. While there is no true solution to all situations, in this article we’ll look at the tools offered to organizers around the world and the common systems they use to run their events. This is an explainer that seeks to help the community understand competitive events and arrive at a solution that best suits them. This article will largely focus on the options avaiable in Best Coast Pairings (BCP), which is the standard in the US and common in other parts of the world now too.
As a note, we do strongly recommend that Tournament Organizers (TOs) take a transparent approach to their tournament setups so that players can understand how both the pairing and placing systems will operate and make conscious decisions on how they approach the event.
There are a few different ways you can structure tournaments based on how you want to have players face off against each other. A tournament need not adhere to one of these and may in fact combine them – we’ll talk more about that later when we discuss pods.
- Single Elimination tournaments are events where a player is eliminated from competition – either from the event entirely or from contention for the top spot – after a single loss. Single elimination brackets see one winner advance from each game in a round, while losers are either eliminated or may play each other for lower seeding. Many large tournaments advance to a single elimination structure for the final rounds, once a top 8 or top 16 have been determined.
- Double Elimination tournaments acknowledge that single elimination tournaments may eliminate someone too early and sort losers into a losers bracket, giving them another chance to compete later in the event. As such, a player must often lose two games early on to be removed from contention.
- Round Robin tournaments have each participant play all the other participants an equal number of times – in a 40k context usually once, since there’s unlikely to be things like a home and away game to consider. You’re more likely to see these in a league than a tournament.
- Swiss tournaments assign players points for a win, loss, or tie, and pair players based on their current points total, seeking to match players by skill level. Most 40k tournaments use Swiss pairings of some kind or another with different tiebreakers. We’ll talk about those in a moment.
Pairing and Placing Metrics
Pairing vs. Placing
Before we start, it’s worth noting that both the pairing and placing systems utilize the same metrics and options. However, the pairing system is applied at the end of each round while the placing metric is applied continually as results are updated then finalized at the conclusion of the event. It’s possible, and common, for an event to use one series of metrics for pairings and another for placings.
The variables that may be individually utilized or stacked into a ‘tiered’ system are described below:
Individual performance as measured by wins – the aggregate number of wins where ties are awarded as .5 wins.
Player A has had the following tournament performance: Rd1 – Win, Rd2 – Loss, Rd3 – Draw, Rd4 – Win. Player A has 2.5 Wins
Individual performance as measured by the total battle points scored – aggregated across each round.
Player A had the following tournament performance: Rd1 – 97, Rd2 – 67, Rd3 – 86, Rd4 – 94. Player A has 344 Battle Points.
Win Strength of Schedule (SoS)
A measure of the quality of a player’s opponents – the total of opponent’s wins divided by the number of opponents
Player A plays three opponents (B, C, D) who at the end of three rounds have B: 2 wins, C: 2 wins, and D: 1 win – the total of Player A’s opponent’s wins is 5. Therefore 5 / 3 = 1.67. Player A has a Win SoS of 1.67
Battle Points SoS
A measure of the quality of a player’s opponents – the average of opponent’s battle points divided by the number of opponents
Player A plays three opponents (B, C, D) who at the end of three rounds have the following scores B: 214, C: 297, and D: 134 – the average of Player A’s opponent’s scores is B: 71.33, C: 99, and D: 46.67. Therefore (71.33 + 99 + 46.67) / 3 = 72.33. Player A has a Battle Point SoS of 72.33
Wins Extended SoS
An abstraction of Win SoS and a measure of the quality of an individual’s opponent’s wins – the total of all a player’s opponent’s SoS divided by the number of opponents. Should be used in conjunction with SoS.
Player A plays three opponents (B, C, D) who at the end of three rounds have each played three opponents resulting in their SoSs of B: 1.33, C: 2.0, and D: 1.0 – the total of Player A’s opponent’s wins is 4.33. Therefore 4.33 / 3 = 1.44. Player A has an Extended SoS of 1.44.
Battle Points Extended SoS
An abstraction of Battle Points SoS and a measure of the quality of an individual’s opponent’s scoring – the total of all a player’s opponent’s SoS divided by the number of opponents. Generally used in conjunction with SoS.
Player A plays three opponents (B, C, D) who at the end of three rounds have each played three opponents resulting in their Battle Point SoSs of B: 69.33, C: 72.1, and D: 78.22 – the total of Player A’s opponent’s battle point SoS is 219.65. Therefore 219.65 / 3 = 73.22. Player A has an Extended Battle Point SoS of 73.22
Random assignment. One note about random assignment is that on the borders of a grouping, any player from group A (say undefeated) may be ‘paired down’ into any member of group B (one loss). This makes pairings fantastically unpredictable for systems which use this as a tiebreak or even the primary source of pairing.
Individual performance as measured by the total opponent army points value destroyed
Player A plays three opponents (B, C, D). At the end of three rounds Player A has destroyed the following of each opponent’s army: B: 1,157, C: 2,000, and D: 789 – the total of Player A’s opponent’s wins is 3,946 points. Player A has a Points Destroyed score of 3,946.
Similar to wins, this measures a player’s individual performance but are instead awarded a numerical value for wins (3), losses (0), and draws (1).
Player A has had the following tournament performance: Rd1 – Win, Rd2 – Loss, Rd3 – Draw, Rd4 – Win. Each win is worth 3 points while a draw is worth 1 point. So Player A has 3 + 0 + 1 + 3. Player A has 7 Swiss Points.
Path to Victory
Reflects the difficulty of a player’s position in a swiss system by focusing on the order wins and losses were achieved.
Player A had three wins and a loss, Player B also has three wins and a loss. However, Player A’s path is as follows: WWLW while Player B’s path is as follows: LWWW. By Path to Victory Player A is looked upon more favorably in this system than Player B due to the later loss.
Similar to the other SoS metrics, this measures quality of a player’s opponents – the total opponent’s Swiss Points (in a Swiss Point system) divided by the number of opponents.
Player A plays three opponents (B, C, D) who at the end of three rounds have B: 3 wins, C: 2 wins, and D: 1 win and 1 draw – the total of Player A’s opponent’s Swiss Points is 9, 6, and 4. Therefore 19 / 3 = 6.33. Player A has a Swiss SoS of 6.33
Margin of Victory
A measure of the strength of a player’s wins – the total difference battle or victory points a player achieves less the total opponents battle or victory points.
Player A plays three opponents (B, C, D) who in three games have scored the following: A-94 vs B-26, A-67 vs C-82, A-89 vs D-54. Player A has scored a total of 250 BP and their opponents have scored 162. Player A has a total Margin of Victory of 88 Battle Points.
Common Pairing Systems
Pairing and placing settings are generally used as a series of cascading metrics. Each of the above listed measures may be tiered so that pairings or placings will prioritize one before further subdividing similar groupings along additional secondary or tertiary metrics in an order pre-determined by the TO. During pairings the system will step through the assigned logic and breaks ‘ties’ by the order of these follow-on metrics. Some common setups are listed below with a brief description of their logic and the pros/cons of their use. Please note that depending on the order of the metrics (such as Wins -> Win Path vs Win Path -> Wins), they may wildly swing how both pairings or placings are handled. As such the order that these metrics are applied may be just as important as the actual metrics themselves.
Wins, Battle Points, Strength of Schedule
Probably the most common tournament pairing type, and the default in BCP. The system first identifies and groups players by their number of total wins first, then moves to total battle points to break any ties, then finally a player’s SoS if there are ties in BP.
The strength of this system is that it’s simple and intuitive. Players almost universally understand the system and find it a ‘fair’ method of pairing. However, there are a couple of significant downsides that aren’t immediately apparent.
First, the SoS component can be challenging in early rounds of an event but even in the late rounds, player drops will negatively impact player’s SoS due to the loss of potential wins associated with those players who will have dropped. As a result, through no fault of a player’s own, SoS may underrate their performance. As a result, SoS should only ever be used in events where player drops are not likely.
Second, Battle Points as a secondary tiebreaker incentivizes a gaming of the system depending on the structure of the tournament itself (number of players, rounds, pods, etc). As evidenced by some recent events, sandbagging or submarining becomes a factor of tournament play under the theory that the path to a championship will be easier by artificially scoring low in early wins, and therefore playing a ‘weaker’ field who have not placed as high throughout the event. However, this incentive misalignment is mitigated when there will be multiple undefeated players at the end of the event (which creates its own challenges).
Wins, Battle Points, Random
A common alternative to the first format, the random tiebreaker alleviates the problems of SoS by leaving the final tiebreak to chance. A preferable system for any event that is likely to have some player drops. Otherwise pros and cons of the system are similar.
Wins, Win Path, Battle Points
A format that is probably most famously pioneered by Mike Brandt in his NOVA Open events and now used in the Games Workshop US Open events. Like the other systems it first group by wins, but then determines the tiebreak pairings by when a player took their first loss, and finally by Battle Points.
The system is a little more complex than the previous systems and has a downside in that many players intuitively feel ‘locked out’ of top placings for suffering an early loss and therefore disincentivized to play to the finish. The counter-argument to this is that those same players, by losing an early round, will have had an easier road to a top placing than those who went undefeated deep into the tournament (see submarining/sandbagging above). In practice, this system still prioritizes wins prior to the win path and so one-loss players are never locked out of a high placing (note that this would be the case if Win Path was the first decider – which should never be the case).
Wins, Win Path, Random
Similar to the above system but forgoes Battle Points entirely (more on this later). The system was used in the recent US Goonhammer Open and was found to perform well. Very similar to the above except that Battle Points are made irrelevant in the pairing decision.
Pure chaos. The only thing this system cares about is the total battle points – which are irrelevant outside of an individual game. It’s frankly an awful system and should only be used if you enjoy pain, hearing your players complain about unfair outcomes, and you want to make the world a worse place. Pros: none. Cons: drives violence at the tables as multiple defeated players nonetheless play the spoiler and deny their opponents some key battle points. It’s a bad system and you should feel bad if you use it.
Common Placing Systems
When it comes to final placings the systems are a little more restrained and tend to follow one of the above pairing systems. Part of this is that randomization is eliminated as a source of final standings – no one wants to find out that their placing was left to complete chance without first taking one of multiple alternative ways of measuring into account. This tends to be true even when the alternative measures are flawed in some way – people just prefer an explanation vs a black box solution.
Wins, Battle Points, Strength of Schedule – The same form of pairing may be utilized all the way through placing and remains a simple, intuitive method of finalizing event results. However, it may suffer from the same issues that plague all SoS metrics as described above and, while sandbagging/submarining is not an issue for final placings, the effect of an undefeated player losing in the final round on the top table suddenly dropping multiple places in the final rankings is a very common downside of this system.
Wins, Win Path, Battle Points – Similar to the pairing system, this system may also be utilized for the final placings. Unlike with the Wins -> Battle Points -> SoS system, top table finishers will remain in the top rankings due to the win path component. However, as with the pairing system some players who have lost early in the event may feel ‘locked out’.
Battle Points – Just don’t. You monster.
Goonhammer Pairing & Placing System
There are no one-size fits all settings. However, in most cases, we find that the following system provides the best standard competitive experience.
Pairing: Wins, Win Path, Random
Placing: Wins, Win Path, Battle Points
As we noted above, the Pairing System was used at the recent US Goonhammer Open to good success (and strongly positive review and feedback). The big decision here in using this system is to forgo Battle Points as a pairing metric in favor of a randomization. This neatly closes off the option of submarining; randomisation is favourable over the other options due to the limitation of SoS metrics (player drops), the duplication of already used determinants (Swiss Score – Wins), and the skew of alternatives (Points Destroyed – prioritizes killing vs mission play), which leaves the random tiebreaker as the most unbiased and therefore fair way of assigning a final pairing based on like wins and win path.
However, random is not a valid measure of a final placing and as a result Battle Points are brought back as a final tiebreaker for event placings in this system. This encourages players to always aim to play to the best of their ability and try to maximise their score.
Alternative Tournament Structures
In addition to the pairing and placing systems, some tournaments may alter the fundamental structure of the tournament itself with creative methods of scoring or bracketing. A couple of common structures are described below.
For commentary on the WTC format, we’ve borrowed the opinions of Team Scotland captain Innes Wilson.
The WTC Format takes its cue from the Minor/Major Victory systems of old, and is adapted from its use in the World Team Championship event where it gets its name. It uses a modified version of points differential to score games; this has been tweaked over the years to account for different mission scoring, and the current 9th edition version uses 5 point Increments, where each 5VP you have over your opponent results in a different share of the 20 Tournament Points available for the game. For example, a 0-5 point VP difference results in a draw, which is then a 10-10 in the Tournament Score, while a 32 point win is a 16-4 win for the player with the higher score. The highest score difference required is 51VP+ for a 20-0 win. These scores are then collated over the event into your total score. Only this score is used to determine placings, with any number of secondary tiebreakers being possible.
As an example, let’s say Player A plays three opponents (B, C, D) and across the three games the scores are as follows:
Game 1: A-94 vs B-26 (20-0)
Game 2: A-67 vs C-82 (8-12)
Game 3: A89 vs D-54 (16-4)
Player A therefore has a total Tournament Points score of 44 out of a possible 60. Only the tournament points score is used when determining placing. Individual win counts are not tracked when using WTC scoring. 5 11-9 wins results in a score of 55, while 3 20-0 wins and 2 0-20 losses results in a score of 60, which would place higher. Players are paired based only on their Tournament points. This can be seen as a positive or negative of the format; it devalues pure wins in favour of quality of wins, though it’s worth remembering that due to players being paired vs opponents on similar total points to their own going 5-0 and scoring 55 TP is not necessarily directly comparable to going 5-0 in a WLD format. (Corrode: you can think of this similarly to pod structures below, where you might have a better win record than people above you, but earned from theroetically weaker competition).
This format aims to incentivise interaction by encouraging denial of your opponent’s points as well as maximisation of your own, while also providing an incentive to play out losing games as you can still take points from any game. It also allows alternative “win conditions” such as pairing into a hard matchup and being able to play for a small loss rather than a blowout. It also disincentivises talking out games, or collusion, as both players’ scores are directly impacted by the difference in scores making it harder to collude allowing your opponent to score higher without impacting your own placements.
This system may take a couple of forms, but the most common within the 40k community has been pioneered by Mike Brandt in his NOVA Open events and now with the Games Workshop US Open events. Rather than a single tournament bracket where all players are paired in the same big group, the pod structured event will establish a defined number of rounds as a “play-in” before breaking out players based on their performance up to the round cut-point. From then on they will be assigned into pods where they will be locked into the pod’s range of final placings and will only play opponents within their pods to establish the final rankings for each pod and the tournament overall. This is an effective system for very large events in which an adequate number of play-in rounds have been established to sort the field. It also provides an opportunity for players to “win” in what effectively becomes mini-playoff structures, and allows TOs to distribute prizes throughout the field – as an example, our own Rob Jones recently went 4-0 in the 2-2 bracket at the New Orleans US Open, winning his pod and a snazzy certificate to adorn his wall. It is therefore a more inclusive structure. It’s a good system for big events with a large number of rounds, but at smaller events you run the risk of having too few play-in rounds or not enough peopel to make the brackets meaningful, which might have a dual effect of ‘locking in’ or ‘locking out’ players of mismatched skill sets.
An alternative to a full pod system is to simply create a “top cut” in which only the top x-players of an event, after a defined number of rounds, play on for the championship. The Las Vegas Open is the classic example here, though as that event has become truly enormous they’ve had to create a “shadow round” for the cut to accommodate their goal of having the top cut consist of all the undefeated players after six rounds. Other examples of this system include Adepticon, and recently, the London GT, with a very drastic cut to a top 4 after just 5 rounds.
Wrapping Things Up
As I’ve stated previously there is no one-size fits all, but the better you understand the systems and how they pair and rank players, the better you can either understand what you’re signing up for, if you’re a player, or what would best suit the needs of your event, if you’re a TO – which will hopefully help you to get the best competitive experience possible.
Have any questions or Comments? Drop us a note in the comments below or email us at firstname.lastname@example.org.