Organized Play Guide Update - OPG 4.1

by Matthyas Kiraly, Director of Organized Play

10th March 2014

The second update to The Continuing Committee's Organized Play Guide for 2014, version 4.1, is now available to download from the home page and takes effect immediately.

The following is a summary of all major changes from the last released version (4.0), in order of appearance in OPG 4.1. This is then followed by further detail on the major changes.


SUMMARY OF CHANGES

[UPDATED] Section 3: Registering a Tournament
Recommended minimum number of rounds has been updated in an effort to decrease the reliance on tiebreakers past Strength of Schedule to determine final rankings.

[UPDATED] 4.2.2 – Limited
Added requirement for chosen card pool for championships-level constructed tournaments to be submitted to the Department of Organized Play for review and approval.

[UPDATED] 4.2.3 Traditional and 4.2.6 Academy
Updates to the wordings of these card pool definitions.

[UPDATED] 7.3 – Time Warning and Extra Time
Added allowance for tournament directors to allocate extra time to games due to extenuating circumstances.

[UPDATED] 7.4 – Ending a Round
Updated based on Extra Time rule in 7.3.

[UPDATED] 7.7 – Determining the Winner After the Final Round
Addressing an issue with round-robin tournaments, the tie-breakers of CVP (cumulative victory points) and Differential have been swapped in their positions in the tiebreaker stack (3rd and 4th respectively).

[NEW] Section 11 – High-Level and Specialty Tournaments
New section that covers the basic definitions of each of our championship and specialty events, including information on k-values, timeframes, how the events are allocated, and the byes awarded.

[NEW] 14.3 – Interim Organized Play Rulings
The Director of Organized Play may authorize an interim ruling be made between OPG releases, which will officially override existing OPG entries until which point the OPG is updated to incorporate the interim ruling.

[UPDATED] Glossary
Updates to the wordings of the “Printable Physical Card” and “Virtual Promo Card” definitions.

====================

SWITCHING CVP AND DIFFERENTIAL
As per 7.7 – Determining the Winner After the Final Round

This is expected to be the most controversial change and was not made lightly.

The issue of the effectiveness of CVP at round-robin events has been around since the decision to implement our current set of tiebreakers back in 2011. During the planning for this OPG update, this topic was raised again for the OP team to consider, and it was no less controversial than when tiebreakers were first raised for discussion back in 2010/2011. Only now we have the benefit of experience and stats.

Now, before I go any further, let me just clarify this is not about “which is better” of CVP and differential at breaking ties. You could raise perfectly valid arguments for both and simply not be able to split them. This change is not over which is supposedly better. A problem has been identified, and this is about fixing the problem.

For those not familiar, CVP was identified early on as a poor tiebreaker at round-robin tournaments in that, while it breaks the tie, it does so on the basis of the first round pairings, prior to a single game being played. Indeed, CVP does break ties efficiently, but it fails to provide a quality outcome at these events, which is unacceptable - the tiebreaker is effectively a random event prior to the first round.

Examining tournament statistics from the 1st of January 2012 to Saturday the 8th of March 2014, with tournaments broken down by 4-6 player events, 7-12 player events, and 13+ player events, raised some interesting facts.

Firstly, 71% of all tournaments are 4-6 player tournaments. As such, naturally you would expect more ties that progress past SoS to occur, and they do. Almost 50% of all such ties occur in 4-6 player tournaments.

Specifically, at 4-6 player tournaments, there are 186 instances of the ties being broken by CVP. Of these, 65 of them (over 1/3) are round-robin tournaments.

Emphasizing that point: over a third of all 4-6 player tournaments that see CVP used to break ties, are round-robin tournaments. Which is where CVP provides the least quality outcome.

CVP is argued to perform better at our larger (13+ player) events, however the number of these events that see CVP being used only make up 5% of our total tournaments.

It is also noteworthy to mention that proportionally, the greatest number of ties that progress past SoS are seen at events with 13+ players (by a factor of two over 7-12 player events, and a staggering four-times the number, proportionally, at 4-6 player events). Keeping that in perspective, that’s still only a total of 65 events.

While analysing this data, I discovered that there is a very good correlation in the number of ties at a given event being inversely proportional to a factor of the number of rounds played given the number of participants (lots of players, play too few rounds, get heaps of ties)*. Thus if we focus on increasing the number of rounds tournament organisers use for events with a larger number of participants, we believe the use of tiebreakers past SoS would drop dramatically (our estimations are by at least 50% if just one extra round is added to such events going forward).

*More information on this analysis can be found below when I discuss the changes to the recommended number of rounds.

There is a theory that deck choices will be affected by switching differential and CVP. We have no evidence that this will or will not actually happen. However, the argument over which is better between CVP and differential becomes moot if the tiebreakers don't see frequent use at tournaments. Currently, over 70% of events see ties resolved before CVP/differential, and we believe that a push for an increase in the number of rounds across all events will see a further drop in the overall number of ties progressing past SoS, therefore minimising any potential affect, certainly in the medium to long term.

In summary, we believe that swapping the 3rd and 4th tiebreakers is a low-impact change that positively impacts local 4-6 player events, arguably our most important events, far more than the impact of the swap at larger 13+ player tournaments, the bulk of which are side-events at Continentals, and Online Events, which are run with fewer rounds than they should have been.

We will be actively monitoring global trends over the next 12 months, and will review this decision prior to the start of the 2015 Regionals season if deemed necessary.

INCREASE IN THE RECOMMENDED MINIMUM NUMBER OF ROUNDS
As per Section 3: Registering a Tournament

While analysing the data regarding the levels of usage of ties at our larger (13+ player) tournaments, we discovered some interesting stats regarding not just CVP, but the use of tiebreakers in general and why we see such an increase at our larger events.

There is a reasonably strong correlation in the number of ties at a given event being inversely proportional to a factor of the number of rounds played given the number of participants.

Anecdotally, I don’t think this surprises anyone. Play too few rounds with too many players and you get heaps of ties. However we now have enough stats in the system that simple data analytics allows us to draw data-justified conclusions, rather than the qualitative arguments we’ve seen of past.

Delving into this further, setting aside those 13+ player events where less than 20% of the players (most commonly just two people) are tied past SoS, almost all of the events did not follow the figures previously set in the Minimum Number of Rounds table in the OPG (Section 3, Table 3.0.1), most often having too few rounds by only a single round.

A good example was a three-round NACC side event in 2012 with 24 players, which saw 75% of players (that is 18 out of the 24 players) still tied after SoS (50% of ties were broken by CVP and the rest by Differential). There are other examples from Worlds and Continentals side events, as well as several online events, but no-less compelling is the number of ties they produced that were resolved by the 3rd and/or 4th tiebreakers.

Almost all of the offending events were 3-4 round events.

In contrast, there were only four 13+ player events that had more than the recommended number of rounds for the number of players, and less than 20% of players tied past SoS.

As such, we believe the number of times ties progress past SoS is tied to the number of rounds the event is run, and can be greatly reduced by increasing the number of rounds played, improving the quality of the tiebreaking, and also with the side-effect of reducing the time required at large events by tournament directors needing to explore tiebreakers to great depth in order to determine the final ranking of players.

A rough extrapolation of results, as well as examination of trends with events that are run over more rounds, shows that the 13+ player events examined with ties that progress past SoS could drop by over 50% if they were run for just one more round.

The table in Section 3 of the OPG 4.1 has been updated to incorporate not only the minimum number of rounds required mathematically to determine a clear winner, but also tightened up to follow the historical trends set by larger events that were able to minimise the number of ties that progress past SoS to more reasonable levels (generally, a maximum of 10-15% of total number of players).


EXTRA TIME
As per 7.3 – Time Warning and Extra Time and 7.4 – Ending a Round

This allows tournament directors to allocate extra time to games at their discretion. If the tournament director is participating in the game, as with any ruling, the secondary or tertiary tournament director would be responsible for making the call on extra time for that game.

The timing of the extra time is up to the players and the ruling tournament director to sort out at the time that the extra time is awarded.

This allowance is separate from the allowance of extra time for tournament directors participating in their own event, as per section 7.1. In those instances, the tournament director should delegate any end-of-round responsibilities, where appropriate, to the secondary and tertiary tournament directors to minimise disruption to the tournament.


DEFINITIONS OF HIGH-LEVEL/CHAMPIONSHIP EVENTS AND SPECIALITY TOURNAMENTS
As per Section 11 – High-Level and Speciality Tournaments

In order to provide more clarity, this new section in the OPG will be home to definitions of The Continuing Committee’s big events.


INTERIM ORGANIZED PLAY RULINGS
As per 14.3 – Interim Organized Play Rulings

There may be rare situations where an OP-related ruling needs to be made between OPG updates. This entry ratifies the rulings as official and, where there is contradiction with any entries in the OPG, the interim ruling will override the entry. The interim ruling will persist until either revoked or integrated into the OPG document.

Discuss this article in this thread.