From the MTR:

The DCI recommends that at least ten percent of all decks be checked over the course of the tournament.

However, there is not such information on what percentage of decklists should be counted for correctness.
Recently, there seems to be a push for judges to “watch more matches” and “actually watch the games rather than move about and only respond to calls”. There will never be enough time to watch all matches and thus we need to choose how much time do we devote to our other reponsibilites (i.e. other than watching matches).

In recent tournament reports, for example, it was said:

Decklists are important to the integrity of Comp REL events, but we shouldn't allow management of them to take up more time than watching for all the other things that are important for event integrity as well. If you've got a designated scorekeeper, maybe suggest to him that he count a few lists every round if he gets bored, and as you get closer to the playoffs, make sure all Top 8 lists are reviewed (which you'll need to do anyway, so you know they're accurate for submission to SCG). (http://apps.magicjudges.org/forum/post/114626/)

We didn't dedicate any time to count decklists. The scorekeeper was free to do so during rounds in that downtime before slips start coming in, but we dove right in with deck checks, counting the lists when we checked the deck (if they hadn't been counted by that time already, mostly just in the first round or so). (http://apps.magicjudges.org/forum/post/116412/)

Counting deck lists: This isn't something that needs to get done for
all lists. Watching more matches of magic is a greater service to the
players. (http://apps.magicjudges.org/forum/post/113833/)

Is it still the case that ALL decklists should be counted for correctness (i.e. no 59-card decklists) and Game Losses for them assessed?

Because if not, then it becomes a question of how many should be checked. (Assume that all judge calls would still be covered immediately, i.e. the only disadvantage of counting decklists would be that fewer matches would be fully watched.)

In addition, if we don't count all the decklists soon and instead count them throughout the tournament, we'll have to abandon or modify this position:

Decklist errors: You really need to issue all the ones found from
counting lists at the same time. Players will question judges
impartiatility when issued found during the one section of counting
are investigated and infractions given at different times. To help
with this you don't need to check both decks in a deck check, just the
one you think has an issue. Or wait for the round to start, or even
previous round to end, just grab the player with an issue as soon as
they finish their previous round. (http://apps.magicjudges.org/forum/post/113833/)

We've moved away from attempting to count all decklists at events of any level in the last couple of years. In general, these only catch clerical errors which are not seeking to gain an advantage and which therefore have no impact on any actual Magic match. It was thought that it would be a better use of judge time to focus on other tasks - watching matches, providing customer service in terms of helping players and responding to calls, and doing deck checks themselves.

I never prioritize counting during round 1 any more (instead having deck checks do an beginning of round 1 deck check), but I will have my deck check team spend about 5-10 minutes each per round counting. Whatever they get through, I'm fine with. It gets most counted, and it gives my judges small breaks sitting down. Those small breaks go a long way in making my team still feel sharp and refreshed at the end of the day.

For our RPTQ we deck-checked close to 30 % since the focus wasn't to count all the lists. One thing I think is important to remember though is that as DC we always want to start by verifying that we have all the lists. That kind of counting is still good to get done as early as possible :).

30% is possible for smaller events; for GPs with hundreds or thousands of players, it's going to be less. But that doesn't really answer the original question.

There isn't a set number, nor percent; it is a good practice to count 5-10 lists each round, per Deck Check judge.

It's also a good practice to ensure that all players who “make the cut” have had their list verified. That's day two at a GP, Top 8 at a GPT or PPTQ, etc.

d:^D

On Sun, May 3, 2015 at 12:48 PM, Scott Marshall
<forum-17978-9872@apps.magicjudges.org> wrote:
> It's also a good practice to ensure that all players who “make the cut” have
> had their list verified. That's day two at a GP, Top 8 at a GPT or PPTQ,
> etc.


This is only of any relevance if they are going to keep using the same
deck or the list will be posted somewhere right.

–
Gareth Pye
Level 2 MTG Judge, Melbourne, Australia
“Dear God, I would like to file a bug report”

Correct - if you cut to a Top 8 draft, the sealed deck lists from Swiss don't need to be completely verified.