OOC:Tiebreaker procedures

Tiebreakers can be controversial at times, partly because of the different methods that can be used, and partly because of the confusion over how the tiebreakers are applied.

Some tiebreakers are seen to unfairly advantage certain teams due to how scorinators work. Other times, the method in which tiebreakers are applied can alter outcomes. It is best to be as clear as possible about tiebreak procedure, while avoiding criteria that others may consider to be biased in one way or another.

Criteria

Multiplicative style modifiers increase the effect of positive style modifiers compared to negative ones, giving teams with positive style modifiers an advantage in goal difference, as well as goals scored. As a result, they should probably be avoided outside of domestic competition, in the interests of fairness.

Additive style modifiers should not change goal difference, so two teams with all else being equal could finish with 8 goals scored and 2 conceded with a defensive style modifier, or 16 goals scored and 10 conceded with an attacking style modifier. Goal difference is therefore okay to use. However, other "goals" related tiebreakers can be problematic.

Using certain tiebreakers could lead to situations where a team's choice of style modifier may impact their chances of topping a group, or qualifying for a later stage in a competition depending on the tiebreakers chosen – which is a factor beyond the "three Rs" (roleplay, rank, and random) and should probably be avoided.

As teams with higher (offensive) style modifiers score more goals/points/etc, they will inherently have an advantage over teams with lower (defensive) style modifiers when goals scored is an early tiebreaker. Similarly, defensive teams concede fewer goals, and are more likely to have a smaller number of goals conceded, giving them an advantage if goals conceded is used as an early tiebreaker; similarly, because goal average divides goals scored by goals conceded, this can also often lead to a better result for defensive teams (for example, 8/2 is larger than 16/10).

In summary, it is often advised to avoid using the following tiebreakers outside of domestic leagues if style modifiers are used (for sports where they apply), as these can be impacted by style modifier choice:

Could favour higher style modifiers	Could favour lower style modifiers
Head-to-Head goals scored Goals scored Away goals scored Goal difference (if multiplicative style modifiers are used)	Goals conceded Goal average

Please note that this is not an exhaustive list.

Examples of different methods

Broadly speaking there are two ways to apply tiebreakers; for ease of reference, this article will refer to them as "in-line" and "recalculated".

For both our examples to follow, let us examine the following set of results from a hypothetical group stage:

Home	Score		Away
Alpha Arrows	4	1	Delta Demons
Bravo Bullets	1	0	Charlie Cougars
Delta Demons	3	4	Bravo Bullets
Charlie Cougars	1	0	Alpha Arrows
Alpha Arrows	2	0	Bravo Bullets
Delta Demons	1	5	Charlie Cougars
Bravo Bullets	7	6	Delta Demons
Alpha Arrows	2	2	Charlie Cougars
Delta Demons	2	3	Alpha Arrows
Charlie Cougars	1	1	Bravo Bullets
Bravo Bullets	1	1	Alpha Arrows
Charlie Cougars	1	0	Delta Demons

Three points are awarded for a win, and one point for a draw.

The tiebreakers in effect for this group stage were as follows:

Overall points
Head-to-Head points
Head-to-Head goal difference
Overall goal difference
Overall goals scored

In-line

When applying tiebreakers "in-line", the list of tiebreakers should be worked through in order until all ties have been broken, always moving to the next criterion until no tie remains. Multiple ties can be broken on a single criteria.

The logic of "in-line" tiebreaking on the example data will be as follows:

Three teams are tied on points, and the tie must be broken
- All three teams have the same number of points from head-to-head matches involving all teams in the tie (4), move to the next criterion
- All three teams have the same goal difference from head-to-head matches involving all teams in the tie (+1), move to the next criterion
- One team (Bravo Bullets; +1) has a lower overall goal difference compared to the other teams (+5 each), place them third and move to the next criterion
- The remaining teams can be divided on overall goals scored; Alpha Arrows have scored more (12) than Charlie Cougars (10), rank them accordingly

This results in the following table:

#	Team	P	W	D	L	GF	GA	GD	Pts
1	Alpha Arrows	6	3	2	1	12	7	5	11
2	Charlie Cougars	6	3	2	1	10	5	5	11
3	Bravo Bullets	6	3	2	1	14	13	1	11
4	Delta Demons	6	0	0	6	13	24	-11	0

This process is outlined by Osarius on the forums, here. It should be noted that any spreadsheet based tool created by Osarius will use this method of applying tiebreakers unless otherwise specified. Most similar tools will likely operate in the same way due to the calculation restrictions inherent to spreadsheets, but you should check with the creator wherever possible in order to be certain.

Recalculated

When using "recalculated" tiebreakers, once a team can be removed from the group of tied teams, the process restarts from the first tiebreaker, excluding that team. In cases where multiple teams may be removed on a single tiebreaker, they become a separate group which is recalculated aside from the other remaining teams.

Using the same example fixtures as seen above, the logic of "recalculated" tiebreaking will be as follows:

Three teams are tied on points, and the tie must be broken
- All three teams have the same number of head-to-head points (4), move to the next criterion
- All three teams have the same head-to-head goal difference (+1), move to the next criterion
- One team (Bravo Bullets; +1) has a lower overall goal difference compared to the other teams (+5 each), place them third and restart the process with the remaining teams
  - Charlie Cougars have more points from head-to-head games between the two teams (4, compared to 1 for Alpha Arrows), and therefore they should be ranked first