Imagine following scenario:
The tournament is over. There were only swiss rounds with no cut. Final standings have been printed. TOP2 players have both the same amount of match points and all their tie-breakers are identical.
In this case, Wizards event reporter sorts them in order of the player ID. Player A has lower ID, therefore he is shown as 1st and player B is 2nd.

My questions:
1) Is player B entitled to request the same prizes as player A?
1b) Can he request at least a forced split? (What if there are prizes that cannot be split?)
1c) Can player B at least request a random determination of the final standings?

2) What if players A and B are 8th and 9th after swiss portion of tournament before cut to TOP8?

Tie breaks are tie breaks, and are usually arbitrary, but saying that, the official tiebreakers are in the MTR section 3.1 -

Match Points first
Opponents' match-win percentage
Game-win percentage
Opponents' game-win percentage

Even though WER displays standings with a 5th tiebreaker it's not an official tiebreaker. You would have to determine how to break the tie beyond that.

Previous examples on the judge mailing list have been generally along the lines of use the WER standings order; if you have time get the 2 tied players to play a head to head; if they've played each other already whoever won that match; toss a coin to see who breaks the tie. This is about the only time tossing a coin is appropriate to determine something.

Realistically it doesn't happen very often, where it's more likely to happen is in events where there are a small number of players and the number of rounds is 2 or 3 rounds more than would normally be necessary to have a single undefeated player.

In the “old days” the event software had the “unknown tie-breaker” which was the order in which a player was entered into the software. I guess today, based on your testimony, it uses your player ID. I wonder if this is the DCI number as expanded to 10 digits or a simple text comparison of the first characters of the DCI numbers?

Mark's advice is most solid. When possible, a play-off is optimal. I feel that even if they have played before, that is still best. Some judges and organizers say that in such a case, if two such players have played before, then their match result is the final tie-breaker. Otherwise, in the one other case in which this happened to me - long ago at a PTQ - we also used a random method to determine the outcome. And yes, this is the only time such an action is okay.

I've seen many strange things in my days of judging, but 5th and 6th
tie-breakers are not one of them.

On Thu, Jan 24, 2013 at 4:24 AM, Ryan Rubley <
forum-2652@apps.magicjudges.org> wrote:

> I thought the MTR used to list 5th and 6th tiebreaker possibilities as
> “Opponents' Opponents' match-win percentage” and “Opponents' Opponents'
> game-win percentage” even though DCIR/WER didn't actually calculate these.
> I don't see it in the MTR, so maybe I'm remembering wrong.
>
>
> ——————————————————————————–
> If you want to respond to this thread, simply reply to this e-email. Or
> view and respond to this message on the web at
> http://apps.magicjudges.org/forum/post/12799/
>
>
> Disable all notifications for this topic:
> http://apps.magicjudges.org/forum/noemail/2652/
> Receive on-site notifications only for this topic:
> http://apps.magicjudges.org/forum/noemail/2652/<http://apps.magicjudges.org/for%20um/noemail/2652/>
>
> You can change your email notification settings at
> http://apps.magicjudges.org/profiles/edit
>



–
Adam Cetnerowski
Gdansk, Poland

I've seen two types of oddities when it comes to tiebreakers:

- I see the final tiebreaker being invoked about once a year, at the draft FNMs which I regularly attend. (there, there's few enough matches that this can happen) In these situations, the two programs differ slightly in implementation:
- DCIR's 5th tiebreaker is “order of entry into event” - the first person entered into the event gets the higher rank.
- WER's 5th tiebreaker is “order of entry into WER” - WER keeps a unique table of players for all events, (you have the same player ID in all events on that computer) and it uses the order of entry into that table as the final tiebreaker.

- WER has had an occasional rounding bug with tiebreakers, where one person will win out by tiny percentages like .0001%. (i.e. a margin mathematically impossible at these tournaments)

I thought the MTR used to list 5th and 6th tiebreaker possibilities as “Opponents' Opponents' match-win percentage” and “Opponents' Opponents' game-win percentage” even though DCIR/WER didn't actually calculate these. I don't see it in the MTR, so maybe I'm remembering wrong.

The MTR lists all tiebreakers applicable to Magic tournaments, but OOMW% and OOGW% aren't used for individual events. I remember definitively that OOMW% is used for some team events, at least in DCIR; I _think_ OOGW% is as well, but I'd actually have to launch programs and look.

My questions:
1) Is player B entitled to request the same prizes as player A?
1b) Can he request at least a forced split? (What if there are prizes that cannot be split?)
1c) Can player B at least request a random determination of the final standings?

2) What if players A and B are 8th and 9th after Swiss portion of tournament before cut to TOP8?

Here's the thing: there is no rule on any of these. This is one of those situations where the head judge has the discretion. I agree with what Mark says on approaches to this, but I'd like to add one more:

Anyone who has used various incarnations of DCI Reporter in the past several years know the increasing restrictiveness of the software, for better or worse; my personal example is if a store used to run 21-player drafts as 8-7-6 pods in the past, well, now they're stuck running 10-11 drafts. It's harder today with WER than it was with DCIR to make a decision with the confidence that the software will work with the change. The decisions what WotC has made with the software - making it much more guiding and restrictive, rather than flexible - has changed my personal belief on a situation like this to be “the software is above the MTR; dealing with discrepancies is a customer service issue”.

On Fri, Jan 25, 2013 at 1:49 AM, Jordan Baker <
forum-2652@apps.magicjudges.org> wrote:

> - WER has had an occasional rounding bug with tiebreakers, where one
> person will win out by tiny percentages like .0001%. (i.e. a margin
> mathematically impossible at these tournaments)


The last time I tried to dig into tiny percentage differences it turned out
to be WER following the MTR to the letter. The MTR says that no player
counts in another players OMW (and other) calcs for less than 33.33% but
someone who has won 1/3rd of their matches/games counts as 33.33333333…%
which is around the scale of what appears to be rounding errors. You'll
probably find that the nights you get those small differences you had 3 or
6 rounds or someone dropped at 1-2 or 2-4.

1. It's 33.0%, or “0.33” in Appendix C. (in WER and the MTR - in DCIR, it's still 33.33 repeating, as the change to 33.0% happened after the last DCIR update)

2. Assuming:
- WER orders standings based on the displayed numbers, and not their calculated numbers. (which would go to another ~10 digits when expressed in base-10)
- The smallest difference between any two PMW% (that you'd find in a tournament, that is) is 33.33(repeating) and 33.0%.

With those two assumptions, you would need to make the difference of .333…% turn into a difference of less than 0.00005%, which would require 13 rounds. (if I'm doing the math right…)

3.
WER doesn't follow MTR Appendix C to the letter. OGW% is not calculated in WER with the 33.0% floor, which the MTR states shall happen.

I think we should stop the analysis of how tiebreakers are calculated and how WER handles them. It isn't on topic. I'm not going to close the topic though as there may well be useful further discussion on the topic of how to handle two players with identical tiebreakers especially when it's 8th and 9th and needing to cut to a top 8.

The precedent for coin flipping was set by the 2HG PTQs a some years ago -
these kind of absolute ties for 4th place in a single game format were much
more common - and 2HG PTQs had to cut to Top 4 because of the draft.

I don't think there's much more to say than that. Play off if you have
time. You might well have time in constructed - you can in theory have the
playoff at the same time as 3 of the other quarter final matches. In
limited it's like an extra round before the draft, and horrible for the 7
Q'd players who have to wait an indeterminate amount of time.

Determine randomly if you don't have time. It sucks to be eliminated on a
random draw like this, but it doesn't take long for the players to agree
that there really isn't any other way.

Paul Smith

paul@pollyandpaul.co.uk

On Fri, Jan 25, 2013 at 10:37 AM, Paul Smith <
forum-2652@apps.magicjudges.org> wrote:

> In
> limited it's like an extra round before the draft, and horrible for the 7
> Q'd players who have to wait an indeterminate amount of time.
>

Time is always going to be tight. The players may prefer a 15 minute single
elim match for those two players, with only one game to be played. It's not
awesome, but it is less random than a coin flip.

Originally posted by Gareth Pye:

Time is always going to be tight. The players may prefer a 15 minute single
elim match for those two players, with only one game to be played. It's not
awesome, but it is less random than a coin flip.

I agree with this sentiment–this sort of situation really calls for an ad hoc ruling from the Head Judge. I would first check to see if the players had already played each other. If so, then the winner of that match wins the tie break. If not, or if that match was a draw, then one 15 minute game with sudden death rules in effect at the end of 5 extra turns seems the best way forward.

I think that an event that uses no playoff should give prizes based on points, not on standings. The question becomes moot then, since both players will recieve the same prize. If there is an unsplittable prize (trophy, invite, bragging rights) you can use as tiebreaker the result between the 2 players (most likely a tie) or let them play a decider game.

This doesn't solve anything if there's a decision to be made between place 8th and 9th, as stated before 2HG PTQs used coin flips. I would use result between the 2 players here as tiebreaker, before resorting to coin flip. You can also use the relative standing in round N-1 as tiebreaker after that (especially if the players decided to ID in the last round, they are more likely to accept this result as fair)

Other than that, this situation is significant and exceptional, so the HJ has the authority to make a decision he or she sees fit.