Category Archives: Staff

Advice targeted at ops and stage staff.

More Goldberg, Less Rube

By all accounts, Grand Prix Los Angeles got off to a rocky start. @BenKrantzStudio asks, “Why is there a HUGE line for byes check in for #GPLA?”

Interesting question.

OK. Well, the question as posed is actually pretty mundane, so let’s tackle that first.

Why is there a HUGE line for byes?

Well, there isn’t, not really. There’s a huge line for player issues. It’s true that many of these are people who have questions about their byes. Which happens because:

  1. They’re confused about how many Byes they’re supposed to have. Maybe they had Byes at a previous event and now they don’t (because Planeswalker Point seasons rotate; the switch to the annual system can’t come soon enough). Maybe they don’t understand exactly how the Planeswalker Point system works. Sometimes they’ve been told that they should have byes by a friend or by a well-meaning staff member who is just mistaken.
  2. They think they won a trial that didn’t get reported on time by their local store, or which got invalidated (make sure, when you win a trial, to get your store to submit that tournament right away and to verify that it’s a legal and accepted tournament).
  3. They just plain don’t read the posted bye list correctly and freak out about where their byes went, when they’re already safely in the system.
  4. They’ve filled out their registration wrong (since sadly, online registration doesn’t and currently can’t check the correctness of your DCI number) and used an incorrect DCI Number, which means that they bye lookup system that is based on DCI Number will fail to find your byes.
  5. Something screwy happens with the bye processing and the printed bye list doesn’t have the right data on it. This can happen if the data import has been screwed up in DCIR, or if the TO is using custom software to important preregistration data and something has gone wrong with how they put the bye information into that data.

5 is relatively more rare, but did happen here in Portland (though we caught this early enough to post an updated bye list reasonably early in the morning). As well as scads of the rest.

In an ideal world, there’d just be a site you could go to that would canonically tell you how many byes you have for each event (almost as if the people who make the game ought to have a site that has information about tournaments, your current Planeswalker Points, and your match history … *sigh*). But sadly, that doesn’t seem to be our world.

But as I mentioned, bye issues are a subset of player issues, which are slightly more interesting.

Why is there a huge line for player issues?

To answer this, we need to know more about what types of issues we’re talking about and the sorts of things that can trip players off as we get the day started.

Enemy number one is, by far, “I’m not on the seatings.” Except it’s usually more an angry rant having to do with how they paid and aren’t on the seatings. Most often, it’s simply because they didn’t see themselves, which happens because:

  1. They simply missed it.
  2. They missed it because their name only appeared on the right side of the seatings. Unlike most of the other postings, where you can always find your name in the left column, a player seating does not duplicate entries so that your name always appears in the first column. You may only appear on the right side. Players, check both columns. Judges, cut the darn forms in half so it’s obvious that the two sides are different.
  3. They missed it because something is weird with their name. Either it was typed in wrong during their preregistration, or somebody couldn’t read their handwriting for on-site registration, or the first and last names got swapped when somebody was entering data. Players, check under your first name if you can’t find yourself under your last name, and please write legibly. Staff, make sure you’re being careful when you enter in registrations, any typos can cost a lot of tournament time.
  4. They missed it because they used the wrong DCI number, it got entered in by DCI number, and auto-completed to somebody else’s name. These are doubly pernicious because they also result in a stub player who ends up not showing up (since they never really signed up), which leads to extra byes and extra work.
  5. They’re actually missed, either because of a missing slip, a miscommunication, or these days, frequently because something went screwy with the preregistration system (payment wasn’t successfully processed, something went wrong with the tools that convert from preregistration to entry, etc…), which then needs to be validated and those players added.

Second most common are people with issues with a Sleep-In Special or VIP system. Maybe they signed up for one of these and they don’t see it in the system. Or maybe they bought one, but showed up anyway and are now panicked that they aren’t on the list (generally Sleep-In Specials are filtered out of the player meeting so we don’t have lots of empty seats to contend with). Or maybe they are a VIP with a bye and don’t realize that that often automatically signs them up for a Sleep-In Special.

Third is people who want to know where to turn in their decklist or to get their free stuff. Hint: I don’t know a TO in the world where the line for the scorekeeper is the right place for either of those. Decklists are collected at the player meeting (or however specified if you signed up for a Sleep-In Special – you did read the fine print, didn’t you?), and free stuff is given out in some way that doesn’t make the whole tournament wait for you.

Besides those and the aforementioned bye issues, other reasons that people end up in the line are because they want a Fixed Seating (Note: Please tell us you need one of these either before the rush after the seatings are posted, or after the player meeting. I know that this is the most obvious time for you to think about it, but this is the time that we have the most customer service requests and where delays slow the tournament start the most), because they want to know where their friend/son/daughter are sitting, because they see a line and want to know what’s going on, or any manner of other craziness.

But enough answering what you asked. Let’s talk about the truly interesting question, a question that is much closer to the thing that you actually want to know.

Why in the @#$(*&% is it taking so long for this tournament to start?

Online Preregistration, VIP, Sleep-In Special, constant video coverage … all of these things feel run of the mill these days and are pretty much par for the course, so it’s easy to forget that these are all relatively recent innovations in an attempt to enhance the player experience. So are giant 2000 person tournaments.

Wait a second … recent? Hasn’t it been years? Shouldn’t this be down to a science?

Well, recent relative to when the software infrastructure that runs these things was designed (and yes, the need for new software and technology is foreign to nobody, but that software isn’t ready).

Even based on just tournament size (let alone all of these add-ons), the process of getting a tournament started, including all of the infrastructural work that needs to be done to make sure that it proceeds relatively smoothly, is like a giant rickety Rube Goldberg machine. And every one of these enhancements just sticks another pair of contraptions onto the machine. So now, instead of just having to shoot the ball through the flaming hoop, it has to do a 90 degree turn in mid-air through a combination of magnetic tricks.

Here’s a reasonably representative subset of the things that need to be put into place before we get started with a tournament:

  1. We need a seating. Which means all of our players need to get into the software, which includes an import from our preregistration data, data entry from on-site data, identifying and fixing any erroneous data (this piece is an insanely complicated topic on its own, figuring out the balance between tournament integrity and expediency – more on this topic someday in the future), and dealing with any changes as they roll in from people on-site. This seems like it ought to be pretty straightforward, since it’s just data manipulation, but Preregistration is a maniacal beast; we’ll get to that in a bit.
  2. We need name ranges set up with signs printed out to distinguish them. As much as possible, these should be balanced.
  3. Both of the above, and more, are complicated because of the Sleep-In Special, and the fact that the software doesn’t know anything about the concept. Fundamentally, the challenge is that there are some tasks that want to take into account people who have Sleep-In Special and treat them differently, and there are some that want to treat them the same as everyone else. The two points above represent one of each. The biggest of the former is the seatings for the player meeting. Because you don’t want to deal with a bunch of empty seats during the player meeting (which gets in the way of determining who’s legitimately just not showed up or is a player that might represent a problem in your data set), and Sleep-In Specials will by definition not be there, you need to pull them apart. The standard tactic for doing so is to append something to their name to push them all the way to the end of the range. Chief among the latter is determining name ranges, which don’t want to be biased due to any name shenanigans from players with SIS, since those tags will go away as soon as the tournament actually starts.
  4. VIPs complicate things further, to a degree depending on which services are being offered. Luckily, fixed seatings for these are generally not done anymore, as that’s its own ball of wax. But it’s common to want to give them their own set of pairings, which either means leaving their names tagged separately for the tournament (in which case you get to deal with the parade of folks who don’t understand where to find their names), or to come up with some alternate tooling that can pull out the VIPs in a programmatic sense based on their undoctored names. This must be its own separate system, since again, the tournament software was written far before the notion of a VIP existed. Most VIP programs also offer SIS for those who have at least one bye, so processing the VIPs who have byes separately from the VIPs that don’t and making sure the tagging works properly for each to accomplish the requirements in point 3 has to be considered as well.
  5. Now that we’re past the introductory stuff, now it’s time to actually fix up the files. Remove all the weird tags like SIS. Decide whether to leave VIP tags. Deal with any problems that we’ve found from the player meeting, ideally deleting anybody who isn’t present to cut down on unnecessary byes, without dropping somebody who actually wants to play. Putting this step last is a little misleading, though, as in the ideal world, much of this can and has been done in advance so that this step is quick.

And then a list of things that don’t block the tournament from starting, but will slow things down as time goes on if they aren’t done early:

  1. Drop Lists need to be created. It’s quite common for players to decide some time after they’ve submitted their result slip that they want to drop from the tournament. The process by which this is done needs to be determined and the right materials generated. The trick here being, dropping based on name is really slow and expensive, but players tend to know their name. Dropping based on player number is fast and efficient, but this is an arbitrarily assigned number that nobody knows, and furthermore the software doesn’t have any efficient printout that makes it easy for people to find their player number. So to make sure that round turnaround isn’t adversely impacted, you ideally want to make a printout that maps player name to player number.
  2. Many head judges like to keep track of any time extensions given by judges on a master list. No big deal here, just need to print an appropriate form to track these.
  3. Depending on whether online pairings are available (which they generally are these days) and what toolset is being used for this, various amounts of setup need to be done to make sure that these are available and work. Technically, not needed to start the tournament, though people complain a lot anytime they aren’t working.
  4. As video coverage expands, the amount of support that the coverage team needs increases. Many of them, these days, want additional online data that requires some setup of tooling or spreadsheets to efficiently get the data they need over to them.
  5. A full judge list needs to be generated for the main event for the purposes of penalty entry. Ideally done well in advance, but judge lists can easily fluctuate until the last minute, and the software only accepts 100 total judges, which is often surpassed by the full staff in modern tournaments, so we need to be judicious about who is actually on the floor and will be issuing penalties, which can’t always happen until close to the last minute.
  6. Fix all of the missing DCI numbers. Many DCI numbers will be messed up due to preregistration inaccuracy, and if we left all of these people out, we’d have a riot on our hands. Instead, we tend to leave those folks in and assume we can take care of their erroneous numbers early in the tournament (technically not needed until the tournament is submitted, but not having one is a catastrophic error that prevents tournaments from being submitted, and the longer you wait, the more you risk one of these players no longer being available).

You may notice that the vast majority of this list is eased substantially by having accurate data. Unfortunately, online preregistration does not lend itself well to this goal, as there is no preregistration system that is integrated with the tournament software. So the various systems that are created by TOs to do registration can’t and don’t validate player information, which leads to challenges up and down all of these various lists. An incorrect DCI number can’t be caught by the preregistration system, and wreaks insane havoc with the system.

Whew. Pretty sure I’m forgetting something. But hopefully you get the idea.

And now, finally, we’re equipped to talk about the core of your real question.

What in the @#($& made Los Angeles take so long to get started?

It might not be obvious from the above, but there is a loose sequencing among all of these tasks. The ball rolls down the hill at just the right speed to launch into the cup. The cup swings in just the right way for the ball to be dropped into the funnel. And so forth.

And if somebody just tilts the first part wrong, the whole thing blows up. Suddenly you’re moving parts all over to catch the ball that’s gone the wrong direction, and it’s tough to get caught up and get the ball back onto the intended track.

In Los Angeles, we had two sets of issues. One was that the registration data came in quite late; it wasn’t finally ready until about 10 minutes before the tournament started, for various reasons. This got the whole machine started off on the wrong foot. The other was that various issues with preregistration payment processing caused a decent number of people not to be registered that should be. In a limited tournament, we can hide these because we can get people building a deck while we add people as needed. For constructed, though, you have to wait until everyone is in so that you can pair them against opponents. So it’s extremely visible when this happens.

Additionally, we decided not to tag the VIPs to make them easier to pull out, opting instead to write tools to pull them out separately. But the late-breaking data meant that setting up this tool was delayed.

Finally, once these initial tasks were delayed, getting everything else ready pushed into the first couple of rounds, so we spent a lot of time getting caught up.

Throw all of this together with the usual things that happen at large tournaments (players keeping slips in their pockets for 10 minutes while we try and hunt them down – this isn’t a made up story, this accounted for one of our longest rounds), and suddenly things feel slow.

But, you may wonder, other events have to do all these same things and don’t have these problems, so what gives? That is definitely true. But the GPs are all run by different independent TOs, and they all do things a little differently. They all have their own processes. They all have their own tools. So you can’t assume that anything that works in one tournament will be the same in every other tournament.

And, you might wonder, why are we still using this old software that doesn’t know anything about any of these new processes? Surely it’s obvious that we need new tools, right? Well, yes, it is. And they’re being worked on. But they aren’t done yet.

But what about these common processes, like preregistration, shouldn’t this just work the same for everyone? And shouldn’t it just be integrated with the tournament software and be able to do all the validation? Or even better, just be integrated with your Planeswalker Points account? Or Magic Online account? Well, yes, I agree. That sure would be nice, wouldn’t it?

So there’s the story. The problem were clear and ultimately preventable. And the right people are working on how to make the machine a little more reliable in the future. At least, the machine that this group of staff uses.

Anticipation

It’s Saturday at Pro Tour – Magic 2015. Rich Hagon comes walking up to the scorekeeping desk.  What do you do?

You don’t actually notice until he arrives and starts talking to you. Not surprising, since you’re focused on something else. Sometimes he needs stuff from you, but sometimes he’s just looking for players or here to see other stage staff, so there’s no reason to break your concentration until he starts the conversation. It takes a bit to figure out what he’s asking for, then you start figuring out how to get him that data. You’re probably an OK scorekeeper (you did make it to the Pro Tour level, after all). However, you’re being a reactive scorekeeper. Somebody asks you for something, and you deliver. The vast majority of not only scorekeepers, but staff of all sorts operate in this reactive mode, and the separation between the bad and the average-to-competent staff boils down to whether you know how to deliver what’s requested, how to generate it if it’s unorthodox, and how quickly you’re able to do so.

Don’t get me wrong – there is a lot of variation on this scale, and a lot of learning that people need to do solely on execution, speed, and creativity of delivery. Even the most experienced and skilled staff are constantly looking for ways to get a little faster. So it’s not that this isn’t important. It’s just that no matter how good you get at being reactive, the best you’ll ever be is OK.

You stop what you’re working on, since it isn’t urgent, and pull up the DCI Reporter window and wait until he gets into earshot before asking him what he needs, as coverage is a high priority at these events, and more likely than not if he’s come out to the main area, he probably needs some data that you’re going to be the one to provide. You’re being a proactive scorekeeper. This is where OK staff start to become good staff. You’ve done this enough times or observed this enough times to understand when you’re likely to be involved and when you aren’t, and you’ve learned to speed up the process by being ready and engaged when you’re likely to be needed.

This sort of learning tends to be the sort where you practice your job over and over again. Instead of just paying attention to how you’re doing things, what works and what doesn’t, though, you’re also paying attention to your own workflow patterns. It seems like when this person approaches, they’re usually looking for you as opposed to for the head judge. At the end of the round, a judge is going to want outstanding tables when there’s 5 minutes left to go. A player walking up holding a slip in the later rounds of the tournament is probably looking for the box to submit side event result slips. One walking up with nothing in their hands but standing with their buddies probably wants to drop.

Note, none of this causes you to perform any processes more quickly. However, you’ll notice that people start to say that you’re quicker. This is partly because you’re more immediately responsive, and partly because people who need things from you tend to think of response time as the total time between when they need it and when they get it. They don’t separate between the time it takes to understand the problem and the time it takes to actually solve the problem, so cutting out the former feels like cutting out the latter to people.

You notice that it’s near the beginning of one of the later rounds, and that he is walking toward you not looking at his tablet. As he approaches, you pick up a copy of the standings that you already pre-printed and hold it out to him, so by the time he arrives and starts to talk, he looks down and notices what you’ve offered, takes it, and continues on his way. Now we’re talking. You’re being a predictive scorekeeper, and this is where the magic happens, where people think that you’re pulling off some crazy Sherlock shenanigans and start saying your name in revered tones as a psychic wizard.

What really happened here?

Why did we have standings already printed? In the last few rounds of a Pro Tour, various coverage people are going to need access to standings to know who to watch and to do math on possible Top 8 qualifiers. It can be different ones at different times, but this happens frequently.

How did we infer that this is what he was going to ask for? Well, if Rich is walking around late in a round, he’s probably trying to corral interviews or otherwise looking for other staff, but early in a round he probably is going to be watching matches and more likely to need you. If he’s staring at his tablet, it’s generally got the name of a person on it that he’ll be looking for, and if not, then he probably hasn’t yet figured out who he’s looking at yet and needs data to figure that out.

Another example: Blake and Nate are over on the side table trying to sort out feature matches for the round, and Nate walks over to ask you to print the master list of deck archetypes. Instead, you hand him the specific pairings with archetypes listed next to each name instead, which is already printing by the time he arrives at your computer.

How is this possible? Well, they’ve been picking features every round, but this round it’s taken longer than usual, and while you’re prepping paperwork for judges, you overhear them discussing which archetypes have and haven’t been seen on camera already. As soon as archetypes become a selection criteria, it’s a good bet that they don’t have all the data they wish they had. You, on the other hand, have the right tools to create a merged pairings/archetype list.

Examples of this sort abound. When you see a player running toward you at the beginning of the round, calling over a judge to stand next to you before that player arrives so that you know one is on hand. When a judge walks toward you holding a decklist, pulling up your reference for who’s sitting where.

There are a few key skills that combine to make these and just about any other similar situations possible.

You need to be comfortable enough with your own work and workflows that you can pay attention to what other people’s workflows look like, and what their motivations are likely to be to solve those workflows. “What is this person likely trying to accomplish right now?” goes a long way toward helping to figure out what they’re probably going to come asking for.

You need to be effective at your job without having to focus so hard at it that it keeps you from observing the things going on around you, since that observation gives you the evidence that you need make better deductions and pick up on things that you might be able to help with.

You need to understand what other people think about what you know and what you can do, both to help predict the types of things they’ll ask you for (which necessarily will fall in that category), and to synthesize ways that you can help that they haven’t conceived yet, which nobody but you can do effectively since nobody else will know to ask.

While these examples are from the scorekeeper’s perspective, the same principles apply regardless of your role. Let’s take, for example, judges. Noticing the line in front of the stage at the beginning of the day, grabbing a set of pairings and starting to work through that line for people who just can’t find their own seats. Seeing the cluster of people around a table and walking over to be a pair of eyes on a potentially contentious match. Catching that there are an odd number of people at the deck check table and standing nearby to help with a deck check that is one person short.

Everyone is interacting with other people and other staff of all sorts. Learn to do your job well. It’s important. But if you’re not paying attention to what it takes for everyone else to do their job well, then you’re never going to be as good at your own job as you could be.

Be magical. Predict, don’t just react.

The Checklist

All right … now that things are a little less frantic, let’s turn that last post into something with more useful content.

First of all, some context on what happened. When DCI Reporter takes in the last result of a round, it then spends some time processing all of the results and getting the points for the wins associated with each player record. When that process completes, then you have the right records for generating pairings for the next round. Unfortunately, this process is not instantaneous, and furthermore, the program does not block you from doing other things while it’s processing. So if you happen to pair the round quickly enough, it will go ahead and do so with the player records it has, which will not include the points from the most recent round.

This results in completely unusable results and, for the record, is one of a vanishingly small number of things for which you absolutely must re-pair (which really is a dirty word – there are very few other circumstances in which you should ever do a full re-pairing. More on that in a future topic for scorekeepers). Which means it’s entirely and cripplingly bad.

There have been a few comments lately of the nature of, “it’s 2014, why not just fix the program rather than talk about how the program is as a given?” And while I can see that point, it’s not a useful one for run-time operations of an event, because we have to work with the software we currently have, as it currently works. Anything else is unhelpful pipe dreaming when you’re actually mid-tournament and trying to make sure everything goes well.

So, reality is, that behavior exists, and you have to work around it. In this case, the best fix is actually a bit retroactive and not proactive. You can’t guarantee you won’t run into this bug, but you can verify after the fact whether you did or not. When you finish generating pairings, the pairings by table always pop up. Takes only half a second, if you remember, to look at the top pairing and make sure it has the number of points you expect based on the round. If it doesn’t, just kill and remake the pairings.

This is but one of a number of things that can and should be getting checked reactively to prevent any issues from turning into real problems. Here are some more:

Players:

– Always check your points, every round. They are listed on the pairings next to your name. Make sure you look at your own row, and make sure you look at your points rather than your opponents. Many players seem to use table number as a substitute for checking points, and while it’s generally true that higher records mean lower table numbers, this is not a guarantee. Fixed seatings can force tables into unexpected ranges. Seatings are sometimes done randomly for a variety of reasons. There is no real substitute for actually checking your points, and if an error has been made, there’s only a limited window during which it’s reasonable to fix.

– Always check your result slip to make sure your name is on it and that you’re playing the right person. A few times a tournament, somebody sits at the wrong seat, plays the wrong person, fills out the result slip, and turns it in without ever seeing the issue. This leads inevitably to confusion, and generally to penalties and game losses. You don’t want to be that person, nor to receive any of those things.

– Always check your result slip to make sure it is accurate before you sign it. It’s a document. Which you are signing. With your signature. Do I really need to say more about this? What would your mother think?

Scorekeepers

– The aforementioned pairings check to make sure the points stuck.

– Make sure you check for last minute drops after you think you’re done. If a judge is helping you manage the drop list (and they probably should be, if we’re talking about a large event), one always seems to manage to sneak through.

– Make sure you check to see if you sent your coverage files to the coverage team. I actually recommend you leave your e-mail client in the Sent Items folder and title your e-mail messages with something descriptive and easy to parse so that you can easily see at a glance what you’ve sent and remind yourself to do so if you’ve forgotten. This is often the slow link in getting pairings/standings/results up on the coverage site, which makes players grumpy.

Judges

– Make sure you’ve checked that you’ve got all the pages you expected out of the printer before you run off. Being in a hurry doesn’t help if you don’t have all your materials, and printers unfortunately still run out of paper.

– Check your result slips as you put them down on the tables. It only takes one stuck piece of paper to make you redo a ton of work.

– Double check a player’s result when they hand you a slip. They should have done so before they signed it, but … well, you know.

This is by no means a comprehensive list, of course. If you’ve got your own tips, please let me know!

That Happened

Gaaaaaaaaaaaaaaaaaah the “DCIR doesn’t include the most recent round’s points in the new pairings if you pair too quickly” bug makes me want to punch myself in the kidney. That and our connection to the online pairings site went down precisely between putting up the wrong pairings and the right pairings.

Re-pair is a filthy word. I feel dirty.

Days since screwing up a GP: 0

(Just in case anyone thinks I only pick on other people here …)

What Happened?

There’s a crowd of people sitting around table 243, including a pair of judges. Things seem a bit tense, understandably so, as a few minutes ago, a little bit of sloppiness in communication lead to a huge disagreement about whether this creature attacked or not, and before you know it, an eight minute extension was issued after the judges finally sorted out what actually happened and got the match ready to go again. Which is a bit of an unfortunate coincidence, since this match also got deck checked this round, which was another seven minute extension (a fast deck check, really). We’re finally entering extra turns now, though, and whether it’s because it’s an interesting match or because it’s the last match going, either way there’s a gaggle of people watching.

It’s two minutes into the round, and a frantic and upset player runs up to the desk. He was already late to start with and trying to rush to his seat, so he’s already under duress. The inevitable and universal first statement comes spilling out.

“I’m not on the pairings.”

Here we go again.

“We have you dropping last round, did you not mean to?”

“No, that got entered wrong.”

And now the usual ritual of finding the table number and pulling out the slip.

“This says you lost and dropped, what happened?”

“Oh.”

Depending on the person, some different variant of mumbling and bumbling around why they might sign something you didn’t verify ensues. Three more minutes pass while we verify what happened with the opponent, and validate that no funny business was afoot.

Problem is, this player wasn’t near the bottom of the standings, they were actually X-1, so they’re in contention. Can’t really just give them a bye, nor can we just match them against one of our lower ranked players because getting the matchups right is actually a tournament integrity issue. So now we have to issue what’s called a cascade, in which we break a series of pairings and mix-and-match them to get valid pairings. This affects four tables this time, and it’s already now seven minutes into the round, and by the time all is said and done, now we’ve got four tables of people who are starting to play a good twelve minutes into the round, with full time extensions (since for most of them, at least, it wasn’t their fault).

“Player X and Player Y, please report to the main stage. Immediately.”

The immediately is an extra bit of urgency, since it’s the third time we’ve called them, and theirs is the last match slip that we don’t have, despite having triple-checked our stack of sorted slips for it and having sent multiple judges to verify that the table is actually empty. Eventually, perhaps we’ll tack on the public shaming clause – “the entire tournament is waiting on you right now.” For now, though, we’re still hoping they get the message and come up on their own.

Finally, one of them ambles up.

“Where’s your result slip?”

“I brought it up.”

“Are you sure? Can you do us a favor and check your pockets?”

“Oh … uh … sorry,” and the inevitable sheepish crumpled up slip from the pocket. No matter how many people we tell that they need to bring their slips up right away when finished, this will always happen.

Little-known fact about DCI Reporter – it won’t let you insert players mid-round, even if you intend to give them a bye or otherwise can perform one of the rational fixes that would be possible when a player is inadvertently left out of a tournament. In this case, it’s a simple miscommunication by which a series of players with byes who were given free entry to the tournament (in exchange for doing some promotional work) weren’t actually put into the system.

The round’s actually over, and we have all the slips. But since the software wouldn’t let us put those people into the tournament until the round was properly over, now we actually have to take the time to insert those players, and manually add their byes round-by-round (another fun software quirk) before we can get around to pairing the next round.

“Player X and Player Y, please report to the main stage. Immediately.”

The immediately is an extra bit of urgency, since it’s the third time we’ve called them, and theirs is the last match slip … well, you get the idea.

Player Y walks up, and Arthur (the quite accomplished and upstanding judge – let this be a lesson that things happen to everyone and you have to stay vigilant!) asks, “we’re missing your result slip, where is it?”

“Um, I handed it to you, remember?”

Arthur reaches into his pocket, and then prepares for public shaming.

WP_20140621_002

It’s been a few minutes now, and the judges are positive – there’s nobody left playing on the floor. Certainly not the 20 matches of missing slips that the software is waiting on.

Is there an errant judge walking around with a pocket full of slips? While one judge starts circulating the word to do a pocket check, another follows the usual first step of going to the group of judges sorting the slips and checking whether they slipped through somewhere. And lo and behold, there they are. We might never know why – sometimes there’s a stack set to the side for entering penalties that the scorekeeper accidentally forgets to enter the results for. Sometimes a judge gets confused and passes a stack of slips directly to the sorters rather than putting them in the box to process. Sometimes the sorters take from the wrong pile and end up with a stack of slips that hadn’t been entered yet. Doesn’t actually matter so much, the minutes are gone either way.

It’s fifteen minutes after the round has completed, as evidenced by the timer slowly counting up in the corner of the room. Pretty long turnaround time between these two rounds, really. So … who’s fault was it?

Perhaps you didn’t receive a Batterskull at a recent Grand Prix (or perhaps you received more than one – I certainly hope you’re ashamed of yourself …). Pretty obvious what happened … right?

Hopefully you’re not as sure as you were a few minutes ago.

As the Round Turns …

It’s Pro Tour – Journey into Nyx from Atlanta, and we’ve just finished Round 9. Which, as it turns out, is the first round of the second draft as opposed to the last round of the second draft. Awkward. Almost disastrously so.

But I suppose I’m getting ahead of myself …

A couple months back, I talked a little bit about coverage, and specifically, why it takes so long to get data onto the coverage site. The follow-up question that seem to be rampant this weekend, from viewers and on-site staff alike, is, “Rich has been filling with this segment for a long time, why is it taking so damn long to pair the round and go back to showing matches?”

The question, as it turns out, is ill-formed. The round might be paired. Or it might not … it can be hard to tell.

Too obtuse? Well … it turns out, there’s a lot going on. Let me tell you the story of the Round 9 to Round 10 turn at Pro Tour – Journey into Nyx.

Rich Hagon is at the desk, doing a segment about draft picks with Zac Hill. A bunch of prep already happened to make this segment possible, which probably involved graphics work he did with Deb and John in advance to get the packs put together (which were typed up in the first place by Trick, myself, and some other number of text coverage staff). And that doesn’t even count the two or three camera operators shooting that piece. And the countless behind-the-scenes folks working on lights, production, streaming, and things I don’t even know.

It’s an interesting bit of content regardless, at least to some viewers, but it’s a segment to fill time between matches because there isn’t anything to show. Don’t tell that to the viewers that are already agitating for matches (to put it delicately … complaining about why we aren’t showing matches all the time is probably more accurate), though. It’s not that nothing is happening – there’s exactly one match left going on the floor … where there are no cameras, and if you don’t think it’d be disruptive mid-game-3 if someone asked you to pick up your stuff and move, then you’re more unflappable than most.

There is a veritable army of judges waiting to post pairings once that last slip comes in. Which it is coming in. But they’ve got a good bit of waiting to do, as it turns out. Trick, Nate, and Tim are also hanging out, and they come first, since they’re picking the matches for us to feature this round. But even they get to wait, as we’ve just finished Round 9. Which, as it turns out, is the first round of the second draft as opposed to the last round of the second draft. Which is awkward. Almost disastrously so.

Here’s the thing. DCI Reporter, the software we use to run this thing, was written a long time ago, in an era where it was never anticipated that we would ask it to do the unholy things that we ask it to do. In the grand scheme of things, the Pro Tour scheme of bouncing between drafts and constructed isn’t even all that unholy. First circle stuff, really. However, it means that our drafts no longer line up with even multiples of three rounds at a time, which used to be a safe assumption with any draft tournaments.

Why on Earth could that possibly matter?

Well, there’s a helpful dialog box that happens when you try and pair a round in a tournament that’s in draft mode that is one after an even multiple of three. It reminds you that you probably wanted to make new pods, and defaults to, “yes, do that for me.” Scorekeepers tend to act on experience and instinct and know all the places that they can blast through the myriad of unnecessary dialogs in DCIR to save some time, and if you’re in the vast majority of even experienced scorekeepers never to have seen this particular warning, then you’re not alone. Neither had Kali.

Once upon a time, that dialog made sense. I dare say it probably saved a lot of mistaken rounds paired against old pods that shouldn’t have been. So I can’t say that the dialog was poorly thought out. What I can say, however, is that now that we’re not aligned to rounds of three, it has undesirable effects.

Kaboom, the tournament is gone. Well, more specifically, all the pods are gone. So the round can’t be paired. And DCIR expresses its displeasure in dramatic fashion.

“Where are the pairings?” ask Trick, Nate, Tim … and Greg, Rich, Deb … and a bunch of producers … and all the viewers. Well …

Luckily, like any good scorekeeper who’s probably been burned a few dozen times, Kali has a backup, and she re-enters the results that weren’t in that backup like a pro. Some file syncing shenanigans ensue (it takes 3 computers in sync to do everything we do at the Pro Tour, but that’s a story for another time), but eventually, we have pairings, and Trick and crew spring to work.

Meanwhile, Rich (who somehow produces while still talking live on camera without skipping a beat), Deb, and Greg (the overall coverage manager) are probably wondering what’s going on. Along with BDM and Randy, sitting in the booth now waiting to get going. Which we can tell because they’re asking on our internal chat. And they need to know so they can plan the length of their segments and make the transitions graceful. We usually try and help keep them up to date as best as we can, but in this case, fixing takes priority over informing – sorry about that, guys.

So, features matches are chosen, we put up the pairings, the round starts, and we cut to match coverage, right? Well … no, not so much.

See, we have graphics to be set up and a feature match area to get prepped first. Also, the folks on camera are going to need to know the matches, get their decklists and some play history, and anything else they can get their hands on to sound knowledgeable on camera. And the staff in the feature match area need to know where players are supposed to sit to match these graphics as soon as they walk over. Which means that we need to get the feature match data all put together and distributed before we tell the players to go anywhere.

Of course, DCIR doesn’t know anything about feature matches. Or Top 25 rankings, which need to show up on graphics. Or deck archetypes (which to be fair weren’t needed for this round, thank goodness, but get spliced in at this point as well). So we have to pull matches, records, and countries out of DCIR and then splice in the rankings and archetypes manually before sending them out to everyone. A minute later, Rich and Deb can start the wind-down on the segment they’re working on, Kali can post pairings online, head judge Lems can fire up the judge machinery to post pairings and get the round going, and we can update the scrolling ticker with the upcoming feature matches, all while the graphics crew gets prepped with the right overlays and graphics for the impending feature match. And then we need to prep our software that streams the results data to the news desk and the updated standings to the scoreboard graphics. But that can wait until the round starts, and until another story.

At any rate, a lot going on. This picture, as it turns out, is probably incomplete. In fact, I suspect exactly zero people, myself included, know everything going on to get a round up and going for Pro Tour coverage. But next time you wonder what’s taking so long, you might want to recalibrate your idea of what “so long” is (the preceding story happened over the span of eight measly minutes), grab yourself a snack, and appreciate the insane work that is happening to make this thing work at all. We do, grumpiness aside, appreciate your patience and thank you for watching!

The Pulse

Every tournament has a pulse, and every round has a heartbeat. That heartbeat is much like your own – it can change a little, speed up, slow down, more blood, less blood. Sometimes there’s even a palpitation to keep things interesting. But at the end of the day, it’s the same action, again and again.

Understand that heartbeat, and you understand how to help a tournament run smoothly. 90% of what you need to know to get it right lives in this chart (click for full size):

RoundHeartbeatBlank

This chart is the heartbeat. It consists of the six distinct phases to every tournament round.

1) Prep – The previous round is done, let’s get the next one started by getting everyone into the right seat. The round needs to be paired, players need to figure out where they’re playing, and they need to get to those seats.

2) Start – Shuffle and present. Players begin play. Tardiness penalties for late players apply. Deck checks happen. Any problems with pairings due to mistaken pairings or other problems need to be corrected. Hopefully this is settled by the time we reach the No Show/Match Loss threshold, after which everyone is playing happily and we can …

3) Cruise – Play Magic. Handle Rulings.

4) Monitor – We’re entering the home stretch, and a lot of playing is starting to finish and results are getting reported and thus a lot of data entry is happening. Help keep the area clear so things stay efficient. Also, though Slow Play is always an issue, this is the phase at which it starts to crop up more frequently as players consciously or unconsciously start to watch the time left in the round.

5) Finish – Time to get the round closed out. At this point it becomes practical to make sure that we have coverage of matches that are still going.

6) Shutdown – We’re into extra turns now, do what needs to be done to get this round finished. If you can’t concretely help with this, everyone benefits best if you get off your feet and out of the way.

Don’t understand? Let’s take a look at a concrete example. Let’s take your average paper team member. Note that this is a generic version, as always, make sure you are in sync with your head judge and/or team lead about the specifics for any given tournament:

 RoundHeartbeatPaper

Or perhaps, more mysterious to the judge at large, how life looks for the scorekeeper:

RoundHeartbeatScorekeeper

Ignoring one-off tournament-level activities (for example, put out the draft sets in preparation for a draft – more on this in a future post), this is all you need to have an effective day. Figure out your chart, write it down if you need, and follow it. Rinse and repeat.

The experienced judges have this committed to muscle memory at this point, and it shows. It is no accident that when we need somebody to handle something, John Alderfer’s name comes up a lot (to pick a concrete example, though he certainly isn’t the only one). He, and others of his ilk, are uncannily frequently in the right place at the right time. This is not coincidence – somewhere down the line, he internalized this timeline and he follows it, round in and round out.

The less experienced or less well-considered judges don’t have this instinct yet, and that also shows. The super-experienced judges know their own pulses as well as those of everyone else, and they weave that understanding together to be maximally helpful to the things that others need to do. After all, staffing is a team exercise.

Find your role. Figure out your chart. Follow it. Someday, it will be instinct and you won’t need to codify it. Until then, fake it until you make it. Any senior judge should be able to help you fill yours out.

The Minutiae of Minutes

As judges and staff, what is our job at a tournament?

This seems like a question that has a ton of answers, particularly if you consider how those answers change depending on what your specific role is.

It’s not, however all that complicated. I’d argue that you have exactly two:
1) Give a good customer experience – get the rulings right, provide good customer service, help answer people’s questions.
2) Get done faster.

To be fair, this really could be one task, as getting done faster has a big impact on the customer experience. However, it’s important enough that I’m going to call it out. This may not be visible to all of the floor judges, but every single head judge that I’ve ever worked with obsesses about getting done faster. How quickly do we get started. How long does it take for people to get seated. What are our turnaround times. What tables have time extensions. All of it boils down to the same thing – how do we get this thing done faster, so that we’re eating at Miku instead of McDonald’s. So people get to hang out with their friends instead of crashing straight away. So everyone gets a good night’s sleep.

Far as I can tell, the first point is constantly in the minds of the judges. They spend tons of time talking about policy, rulings, debriefing on interesting judge calls, working on coverage around the floor, watching each other, etc … And while I can’t say that the second isn’t in people’s minds at all, it’s evident that it isn’t as instinctual for people to think about how their actions affect this goal.

Don’t believe me? I get asked constantly why I bother having people sort result slips at the end of the round while we’re waiting for the last slips to come in, a task that takes quite a bit of time overall and wastes a lot of minutes. Why do this if I’m constantly asking people to add value and save time?

Here’s the rub: the time of yours I’m wasting isn’t, by itself, worth anything. It doesn’t matter. Not all minutes are created equal.

Other than making sure you get off your feet and are resting (which, you can do while you sort slips), while we’re waiting for the last three slips, unless you’re one of the judges who is watching and making sure those matches progress apace, literally none of your time can be spent in a way that makes the round end faster or slower – it’s inconsequential right now.

However, it might not be in the future, when a player comes up and complains that they didn’t get all of their points for the round (probably because they didn’t fill their slip out right, but that’s another story). Now we’re talking impact. Either the round waits until we find their slip and verify everything to repair it, or that table gets a time extension – either way, now we’re into the realm of slowing down how fast we get done. Now, having the slips sorted saves time that means something. And if we wasted twenty judge minutes during dead time to save one minute during live time, every single person on the floor should take that trade.

There are exactly three times that affect end of day timing:
1) Any time that affects round turnaround and starting the round. How fast you post the pairings. How quickly we can find result slips of people with issues. How easy you make it for people to find their pairings and sit down.
2) Any time that affects match extensions. How quickly do you get there, make the ruling, and get out. How fast do we repair pairing issues. How quickly do you handle your deck checks.
3) Any time that affects getting the last slip in. Making sure people turn in their slips right away. Not getting in the way of the last result being reported.

That’s really it. Everything else is manufactured urgency and doesn’t actually help. All that stressing out you do about how quickly the result slips get off the printer and cut. How quickly you run up the third to last slip. That third and fourth and fifth judge you send to watch the match that a competent judge watching for slow play is already on. Doesn’t do anything.

Relentlessly save time. Every minute matters – a single minute saved per round is ~10 minutes per day. A couple of these and, to steal a cliché, we’re talking about real time. But only the minutes that matter. Take the time to figure out which ones those are. And which ones aren’t. The faster you learn to answer the question, “does this actually get us done faster,” the better you will eat and the more you will sleep.