Page 1 of 3

ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Thu Aug 16, 2018 12:23 pm
by Auroni
Hey everyone. This is the announcement thread for ACF Nationals 2019, which will take place on the weekend of April 13th at the University of Pennsylvania.

This tournament will be edited by Alex Damisch, Andrew Wang, Athena Kern, Andrew Hart, Ike Jose, Jason Cheng, JinAh Kim, Jordan Brownstein, Matt Bollinger, and myself as the head editor. I am very excited to work with this ensemble cast of both established and up-and-coming editors, who will breathe fresh life into the game's highest echelons while ensuring consistency, difficulty control, and polish.

This tournament will combine aspects of the previous two ACF Nationals, which will be reflected in the sections below.

Eligibility

Information on who is eligible to play ACF tournaments and details regarding packet submission can be found at the ACF Packet Submission Guidelines. (Note: because our distribution is different, do not follow the guidelines for the distribution that appear in this link)

Qualification

This tournament will use the A-value qualification system, an explanation of which can be found here. Hosts of ACF Regionals 2019 should send information about the teams that receive bids to our set email, [email protected].

Packet Submission

To ensure a fair format and adequate tiebreakers, the editors are committing to writing 11 packets this year. We will produce an additional 9 packets based on submissions that will be used in the preliminary phase of the tournament.

Like the editors of last year's ACF Nationals, I'm hoping that I can coordinate with teams on their packets so that we get somewhere around 15 quality submissions. Please fill out this form if you would like to submit a packet.

Under this arrangement, packet submission will not be mandatory, but we're counting on teams capable of writing quality packets to work with us to ensure that we have enough submissions to edit the tournament. If we can't be sure that we'll have enough packets to make this work, we'll consider other options.

Fee Schedule

Base fee: $320
-$100 discount: Monday, January 21
-$50 discount: Sunday, February 17
-$25 discount: Sunday, March 10
Buzzer discount: -$10 per working system
Staffer discount: -$100 for each staffer capable of reading a round in 35 minutes or less
Quality discount: -$50

I will work with teams to coordinate their submissions. We will continue to offer an additional $50 dollar discount to any team that writes at least 20/20 that doesn’t need any major editing (packets requiring relatively few changes to clues, answer lines, or grammar will qualify).

In the unlikely event that a team submits a packet but does not qualify for ACF Nationals, we will be in touch with you make sure that their school will receive whatever discount they earn as a discount at a future ACF tournament.

Field Size

The initial field cap will be 42 teams. Expansion up to 48 teams is very possible once we confirm staffers. Expansion beyond 48 teams may not be possible because of format, but if demand is high, we will explore our options.

If every team that played ACF Regionals, and thus has an A-value, is offered a spot in the Nationals field and there are still spots remaining, they will be filled first come, first served.

Please stay in touch with me or other ACF members if we are trying to verify your attendance at this event. We may be forced to drop teams from the field if they do not confirm their attendance in a timely manner.

Lodging

We have secured a hotel block for the tournament. More details are available here.

Staffing

If you would like to staff ACF Nationals, please send an email both to [email protected] and [email protected] with the subject line ""ACF Nationals staffing"

Distribution

We've made minor changes to the distribution for this tournament; please see below for specifics.

5/5 Literature: This includes 1/1 American, 1/1 British, 1/1 European, 1/1 world, and 1/1 other (which can include any of the preceding categories). Vary your questions by time, place, and genre, and please make sure that drama, poetry, and short fiction are represented.

5/5 History: This includes 1/1 American, 2/2 European, 1/1 world, and 1/1 other (which can include any of the preceding categories or questions spanning multiple places). Please make sure that your questions aren’t all from the same time period or place and aren’t all one type of history (for instance, try not to submit 5/5 political history or 5/5 military history).

5/5 Science: This includes 1/1 biology, 1/1 chemistry, 1/1 physics, 1 math or computer science, 1 astronomy, earth science, or other science, and 1/1 other science (which can include any of the preceding categories).

3/3 Arts: 1/1 painting and sculpture (this encompasses drawing, painting, printmaking, illustration, ceramics, sculpture, and other small "handicrafts"). 1/1 classical music. 1 question on other visual arts (architecture, film, photography, performance art, installation art, non-Western artistic traditions that do not fall under "painting and sculpture," etc) and 1 question on other auditory arts (jazz, ballet, opera, musical theater, folk music, non-Western musical traditions, popular music from the early 20th century, etc) Vary by genre, place, and time period. If you are writing a packet and have any questions about whether a given topic is suitable for this distribution, please email me.

1/1 Religion (please try to write interesting questions exploring the culture, practices, texts, and study of the world's major belief systems or of well-known historical or ethnic groups)

1/1 Philosophy

1/1 Social Science (this encompasses both hard, ie economics, linguistics, psychology, anthropology, political science, or other social science that is rooted mainly in empirical, scientific methods, ideally topics that are still relevant in their fields today, and "soft thought," such as cultural criticism, theory, and texts or people of interest primarily for historical reasons rather than current relevance)

1.5/1.5 Mythology/Current Events/Geography: 1 mythology question drawing from myths from the world's major mythological systems, or relating to "legends" that are more cultural than classical (such as American lore, European folktales, or similar stories that are less likely to show up in the canonical "mythology" category), or about mythography. 1 current events question describing interesting and relevant contemporary issues without devolving into a list of politicians and political parties or focusing on flavor-of-the-day stories. 1 geography question incorporating cultural and environmental clues instead of purely discussing physical features.

1.5/1.5 Other Academic: at least 1/1 of this category must be dedicated to questions that don't fit other categories or blend clues from varying categories. Disciplines that could be represented in this category include, but are by no means limited to, environmental art, library science, urban studies, and culinary arts. Clues that place academic topics in a popular-culture context are welcome in this category, if done sparingly and well (but please keep in mind that ACF Nationals does not have question on purely popular culture). Please email me if you have any questions or ideas for how to satisfy this requirement. The remaining question may be an "extra" of prior categories.

As overriding principles, please vary your questions by time, place, and subject matter, so that there aren't strange clusters of related questions in your packets; please think twice before submitting tossups on very difficult answer lines; and please try to write your questions on simple answers, using interesting clues, rather than writing on complex answer lines that may confuse readers or players and lead to protests.

Difficulty

In emulation of the difficulty of ACF Nationals 2017, we aim to ensure that scaled down versions of at least 60% of the tossup answers (including ones that are "common link") in each packet could appear as tossups at regular difficulty college tournaments or below. A simple way to ensure that more of your submission gets used is to keep this metric in mind when writing.

How to submit your packet

All packets should be sent to [email protected]. If you have any questions about packet writing, you can send them to that email address as well. Here is a list of the currently submitted packets:

-$100 deadline:

McGill
Minnesota
Chicago A
Penn State
Illinois
Virginia
Johns Hopkins
UCLA
Florida

-$50 deadline:

OSU
MSU
Wright State
Berkeley A
Yale

-$25 deadline:

Harvard
Rutgers
Chicago B
Texas
Michigan
Maryland B
Oxford

Re: ACF Nationals 2019 - 4/20-21/19

Posted: Thu Aug 16, 2018 12:48 pm
by vinteuil
Have those organizing nationals considered that placing it on both Easter and the first two days of Passover will have adverse effects on attendance and competitiveness?

Re: ACF Nationals 2019 - 4/27-28/2019

Posted: Fri Aug 17, 2018 3:56 pm
by cwasims
I'll also note that the revised later date will make it much easier for Canadians to attend given that it is pretty much after our April exam season finishes. The earlier date in past years has been a barrier for several people who've wanted to attend Nationals.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sat Aug 18, 2018 5:52 pm
by Auroni
After internal discussions, we have moved the date of Nationals one final time to the weekend of April 13th and 14th to make it easier for us to recruit the highest quality staff. We apologize for any inconvenience.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sat Aug 18, 2018 5:56 pm
by Cheynem
TAKE THAT, CANADA!

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sat Aug 18, 2018 8:13 pm
by otsasonr
Auroni wrote:After internal discussions, we have moved the date of Nationals one final time to the weekend of April 13th and 14th to make it easier for us to recruit the highest quality staff. We apologize for any inconvenience.
Could we get a more substantial explanation of the logic for the shift than a citation of "internal discussions" and staff availability? Obviously, the staffing requirements are a patently reasonable justification for setting the date for a tournament of the scale of ACF Nationals, but considering the also reasonable justification for the later date raised by Chris earlier, I think it would be fair to expand on the decision making process a bit more. In particular, what facts about the late April date make it significantly more difficult to staff the tournament? How did you counterbalance this against the possible impact on teams? What attempts, if any, were made to poll the community (of both staffers and teams) about their preferences and constraints?

I know these decisions are difficult to make, and I do not fault anyone for not having a rigorous analytic hierarchy model or anything, but greater transparency is, I believe, always worthwhile in these situations.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sun Aug 19, 2018 8:03 pm
by vinteuil
Auroni wrote:After internal discussions, we have moved the date of Nationals one final time to the weekend of April 13th and 14th to make it easier for us to recruit the highest quality staff. We apologize for any inconvenience.
Do you mind sharing more about this decision process? I'm having a hard time believing that it was possible to contact even the cream of the ACF staffing pool that quickly, and there are obviously reasons why this date is problematic for some schools (Canada) and people (those observing Palm Sunday).

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sun Aug 19, 2018 8:19 pm
by Cody
I did not make the final decision for the date of ACF Nationals, but April 27-28 is the date of NHBB Nationals. NHBB Nationals monopolizes a lot of staff. Unless interest in ACF Nationals inexplicably collapses, it's reasonable to plan to accommodate even more than the 47 teams from last year. That would be very difficult while competing with NHBB.

The cream of the ACF Nationals staffing pool is a very small portion of the staff needed to run ACF Nationals smoothly. (I also think it's unreasonable to expect a straw poll of potential staffers to return useful, or any, results at this juncture.)

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sun Aug 19, 2018 8:24 pm
by vinteuil
Cody wrote:I did not make the final decision for the date of ACF Nationals, but April 27-28 is the date of NHBB Nationals. NHBB Nationals monopolizes a lot of staff.
I supposed I'd be interested in seeing exactly how much overlap there is between the two staffer bases; I'd heard talk that it wasn't as big as one might expect (partly because a large number of Nationals-staffer types have disavowed NHBB).

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sun Aug 19, 2018 8:33 pm
by CPiGuy
vinteuil wrote:
Auroni wrote:After internal discussions, we have moved the date of Nationals one final time to the weekend of April 13th and 14th to make it easier for us to recruit the highest quality staff. We apologize for any inconvenience.
Do you mind sharing more about this decision process? I'm having a hard time believing that it was possible to contact even the cream of the ACF staffing pool that quickly, and there are obviously reasons why this date is problematic for some schools (Canada) and people (those observing Palm Sunday).
The set of people observing Palm Sunday is likely smaller than the set of people observing Easter itself or the set of people observing Passover.

also, fwiw, Michigan is actually out of session before the 27th/28th, and I imagine a number of schools would be in the middle of their finals week, which might also depress interest.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sun Aug 19, 2018 9:07 pm
by High Dependency Unit
vinteuil wrote:
Cody wrote:I did not make the final decision for the date of ACF Nationals, but April 27-28 is the date of NHBB Nationals. NHBB Nationals monopolizes a lot of staff.
I supposed I'd be interested in seeing exactly how much overlap there is between the two staffer bases; I'd heard talk that it wasn't as big as one might expect (partly because a large number of Nationals-staffer types have disavowed NHBB).
I haven't been to ACF Nationals so I can't say who staffs that, but Jacob's almost certainly correct on this one. That being said, schools having finals at the end of April is probably a bigger issue than Palm Sunday or NHBB nationals staff overlap.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Mon Aug 20, 2018 11:09 am
by vinteuil
vinteuil wrote:
Auroni wrote:After internal discussions, we have moved the date of Nationals one final time to the weekend of April 13th and 14th to make it easier for us to recruit the highest quality staff. We apologize for any inconvenience.
Do you mind sharing more about this decision process? I'm having a hard time believing that it was possible to contact even the cream of the ACF staffing pool that quickly, and there are obviously reasons why this date is problematic for...those observing Palm Sunday.
It's been pointed out to me that this post probably isn't explicit enough—Palm Sunday isn't a "bad" date for an organist to take off from church, it's a "you'd get fired for asking to take it off" date.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Mon Aug 20, 2018 11:17 am
by Cheynem
Is there a large number of quizbowl players/staffers who work as professional organists at churches? (At my church, the organist is a volunteer who is just a church member, so you wouldn't "get fired" for asking to take Palm Sunday off--I'm aware this is different in other churches)

To be clear, I'm sympathetic to people who can't take this day off. Palm Sunday is important to me, too, and I would be reluctant to take this day off. However, I feel like especially at this time of year, you are always going to run into dates that certain people can't take off or would have a hard time getting the time off.

I also note that I haven't been at ACF Nationals in years and am unlikely to be there this year, so my opinion isn't important.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Mon Aug 20, 2018 11:26 am
by vinteuil
Sure, I can be more explicit in that I'm a church organist.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Mon Aug 20, 2018 12:23 pm
by 1.82
I'll repost in this space what I said privately, since I think it's relevant here. I understand and respect the desire of ACF to finalize a date for Nationals as soon as possible in order to facilitate the placement of bids and the like. That said, it's still August and most universities aren't even in session yet, so I don't think that we're running up against a hard deadline to announce a date for Nationals.

With that in mind, I think that it'd be in the best interests of the community to have an open discussion of the date for ACF Nationals, so that everyone can lay out the advantages and disadvantages of particular dates, as opposed to the current situation where two date changes have already been announced by changing the title of this thread. ACF members aren't the only people with an interest in finding the optimal date for ACF Nationals, and as far as I know there are no trade secrets that require the discussion to be held behind closed doors.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Mon Aug 20, 2018 12:49 pm
by merv1618
vinteuil wrote: It's been pointed out to me that this post probably isn't explicit enough—Palm Sunday isn't a "bad" date for an organist to take off from church, it's a "you'd get fired for asking to take it off" date.
Having spent nonzero time in multiple church choirs, I can second this and any other urgencies about organist jobs and schedules.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Mon Aug 20, 2018 1:48 pm
by Cody
vinteuil wrote:
Cody wrote:I did not make the final decision for the date of ACF Nationals, but April 27-28 is the date of NHBB Nationals. NHBB Nationals monopolizes a lot of staff.
I supposed I'd be interested in seeing exactly how much overlap there is between the two staffer bases; I'd heard talk that it wasn't as big as one might expect (partly because a large number of Nationals-staffer types have disavowed NHBB).
I would judge that the majority of staffers required to run a 47-team (or more) national tournament have no vested interest in ACF over any other organization, including NHBB. To a lesser or greater extent (depending on where ACF Nationals is held), I would personally expect a conflict with NHBB to harm the staff recruitment process, and know it would make at least one desirable host site unfeasible.

The unfortunate fact is that ACF Nationals is generally constrained by a combination of two major religious holidays, ICT, and two huge high school nationals (one, this year). This makes it difficult to definitively point to a date as the best for everyone.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Sat Sep 01, 2018 10:09 am
by Auroni
I have disseminated a survey to potential teams and staffers about the date of this year's ACF Nationals. If you or a member of your club didn't receive the link to it, but you'd like to weigh in, please PM or email me.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Thu Sep 06, 2018 8:51 pm
by Auroni
30 schools and 35 staffers have filled out the survey. Both groups overwhelmingly prefer the currently announced weekend of April 13-14, 2019 (14 out of 20 schools, and 11 out of 16 staffers who indicated a particular date preference). Based on this data, the tournament will still be run at the indicated date.

Re: ACF Nationals 2019 - 4/13-14/2019

Posted: Tue Oct 23, 2018 9:05 pm
by Auroni
Of the six bids that we have received for ACF Nationals host site, we have selected the University of Pennsylvania.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Thu Oct 25, 2018 9:09 pm
by 1.82
Will ACF disclose the other five sites that submitted bids? If ACF feels the need to keep this information secret from the public, what is the reasoning?

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Thu Nov 08, 2018 3:03 pm
by ryanrosenberg
Hi everyone! I'm happy to announce that I'll be serving as tournament director for ACF Nationals this year. I hope to continue ACF's tradition of putting on a rigorous, fun, and rewarding national championship.

Due to an unfortunate timing error, Penn will not be able to confirm their room reservations for Nationals until mid-January. It is still exceedingly likely that we will be able to host Nationals at Penn; however, in the interest of caution, we have secured the University of Minnesota as a backup host site. I'd like to thank the Penn and Minnesota teams for working with us on this, and ask that teams not book travel or lodging in Philadelphia or the Twin Cities until we have confirmed rooms.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sat Nov 17, 2018 2:58 am
by Auroni
As an update, I have created a form to coordinate the optional packet submission. Additionally, all inquiries about ACF Nationals 2019 can be sent to our email address, [email protected].

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sat Nov 17, 2018 4:52 pm
by UlyssesInvictus
If we're considering staffing, how should we make ourselves available for this?

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sat Nov 17, 2018 7:51 pm
by Auks Ran Ova
Send an email to both me and the ACF Nationals email address Auroni just gave with the subject line "ACF Nationals staffing".

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Tue Jan 15, 2019 5:14 pm
by ryanrosenberg
The rooms for this tournament have been confirmed at Penn. Teams should feel free to book travel, and we'll be sending out a staffing call shortly. We look forward to seeing you in April in Philadelphia!

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sun Jan 20, 2019 2:33 pm
by Auroni
If I've gotten in touch with you about submitting packets for the $100 deadline, I'm announcing a one-day extension due to MLK day.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sun Jan 20, 2019 4:10 pm
by Aaron's Rod
ACF is pleased to announce that we have secured a hotel block at the Study at University City. Teams can book rooms on Friday 4/12/19, Saturday 4/13/19, and Sunday 4/14/19 for $194 + citywide 16.25% tax rate.

You can book by clicking here. Or, to make reservations over the phone, call 215-387-1400 and ask for the ACF Nationals Attendee Block. The window to book hotel rooms has now closed.

Parking at the hotel is valet-only Parking $42/night with in and out privileges (plus tax),although teams might find easier parking at a nearby Drexel ramp in Parking Lot G, or elsewhere in the city.

We think teams will find the hotel's location very convenient for the tournament. The hotel is within walking distance of Penn and several light rail and subway stations. To reach the hotel by public transportation, take the Airport Line to the University City stop and walk for 12 minutes, or take the Airport Line further afield and take train/tram back to 33rd station, which is a 2 minute walk from the hotel.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sun Jan 20, 2019 7:17 pm
by alexdz
Philadelphia is pretty reasonably well-connected with public transit, so in addition to the Airport Line train for those teams flying in, there are also dozens of trolley lines and bus routes that will pass within reasonable walking distance of Penn's campus.

Teams who drive to Philly and are hoping to avoid paying excessive parking may find it not all that inconvenient to park further out and take a bus/trolley/combination back to their hotel. I'm not familiar with each and every neighborhood's parking situation, but my neighborhood (Brewerytown) is far enough away from Center City that street parking is free and unpermitted. Depending on when teams arrive, if you wanted to park up in my neighborhood I may even be able to give you a ride to your hotel. Contact me if you want.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Tue Jan 22, 2019 1:53 pm
by VSCOelasticity
Aaron's Rod wrote: Sun Jan 20, 2019 4:10 pm You can book by clicking here. Or, to make reservations over the phone, call 215-387-1400 and ask for the ACF Nationals Attendee Block.
The trailing slash in the url takes you to the general reservations page instead of the event-specific one. Here is the link without it.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Tue Jan 22, 2019 3:42 pm
by Victor Prieto
We will continue to offer an additional $50 dollar discount to any team that writes at least 20/20 that doesn’t need any major editing (packets requiring relatively few changes to clues, answer lines, or grammar will qualify).
I have a few questions about this discount. It's been offered at least since 2013, but a few things have never been clear to me. Who decides what packets receive the additional discount? Are questions that are cut due to being repeats or subdistributional balancing taken into account? What if the packet contains 20/20 questions that are perfectly fine in a vacuum but is inconsistent with the vision of the editor(s), for whatever reason? By when do teams learn if they've received this discount or not? I actually have never heard of anybody who has received this discount, although it's not like I've asked around about it before. It's $50, which isn't negligible, so that's why I'm asking so many questions.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Tue Jan 22, 2019 3:53 pm
by Good Goblin Housekeeping
In a similar vein to Aaron Rosenberg's request from last year for people to not write tossups on scientists, I respectfully ask that anyone writing chemistry please not submit tossups on elements. Functional groups also frowned upon.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Tue Jan 22, 2019 3:57 pm
by Auks Ran Ova
settlej wrote: Tue Jan 22, 2019 1:53 pm
Aaron's Rod wrote: Sun Jan 20, 2019 4:10 pm You can book by clicking here. Or, to make reservations over the phone, call 215-387-1400 and ask for the ACF Nationals Attendee Block.
The trailing slash in the url takes you to the general reservations page instead of the event-specific one. Here is the link without it.
Thanks for catching this; I've fixed the original link.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Tue Jan 22, 2019 4:39 pm
by Auroni
Victor Prieto wrote: Tue Jan 22, 2019 3:42 pm
We will continue to offer an additional $50 dollar discount to any team that writes at least 20/20 that doesn’t need any major editing (packets requiring relatively few changes to clues, answer lines, or grammar will qualify).
I have a few questions about this discount. It's been offered at least since 2013, but a few things have never been clear to me. Who decides what packets receive the additional discount? Are questions that are cut due to being repeats or subdistributional balancing taken into account? What if the packet contains 20/20 questions that are perfectly fine in a vacuum but is inconsistent with the vision of the editor(s), for whatever reason? By when do teams learn if they've received this discount or not? I actually have never heard of anybody who has received this discount, although it's not like I've asked around about it before. It's $50, which isn't negligible, so that's why I'm asking so many questions.
(Speaking only as myself as an editor who has offered quality discounts in the past, and not on behalf of ACF Nationals as an institution) After collaborative discussion, we plan to recognize the packets that most demonstrate a clear effort to write interesting, creative, and fresh questions while meeting all of the difficulty and subdistributional guidelines specified in this thread. Of course, as questions from all packets are frequently cut for repeating with other submissions or editors' packets, this will not be "held against" any submission. Teams will be told about the discount in the weeks leading up to the tournament, though we will not publicly announce a list of all awardees.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sun Jan 27, 2019 9:09 am
by Bosa of York
(This is not directly related to this year’s nats but I’m not sure where else to say it.) In the first post of this thread’s linked post explaining ACF Nationals qualification, an explanation of how each team’s A-value is calculated is deferred to NAQT’s explanation of the D-value calculations through this link, which no longer goes anywhere due to revisions in NAQT’s website. It would probably be good if Stephen Eltinge or a mod could fix the link in that post to the correct one, or, even better, if ACF could either put the information in the first post of this thread or on the ACF website, rather than a three year old forum post.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sun Jan 27, 2019 9:55 pm
by Blackboard Monitor Vimes
Milhouse wrote: Sun Jan 27, 2019 9:09 am (This is not directly related to this year’s nats but I’m not sure where else to say it.) In the first post of this thread’s linked post explaining ACF Nationals qualification, an explanation of how each team’s A-value is calculated is deferred to NAQT’s explanation of the D-value calculations through this link, which no longer goes anywhere due to revisions in NAQT’s website. It would probably be good if Stephen Eltinge or a mod could fix the link in that post to the correct one, or, even better, if ACF could either put the information in the first post of this thread or on the ACF website, rather than a three year old forum post.
I'm going to check with Stephen, as well as this year's Nationals team, to make sure all of the information is fully updated first, but once I've gotten that confirmed I can add it our website. My apologies that I hadn't realized it wasn't already there. We're also planning to feature this year's A-value results on our website (as well as here on the forums) once they're ready.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Tue Jan 29, 2019 2:27 pm
by ryanrosenberg
We are still waiting on finalized stats and autobid applications from hosts in order to post A-values for this year's Nationals; I anticipate that they will be posted sometime this week, likely by tomorrow, along with a full explanation of how A-values work. After A-values are posted, teams that have qualified for ACF Nationals will have two weeks to accept or decline their bid, after which any declined or non-responsive bids will be released to teams on the waitlist.

In a change from previous years, the waitlist for ACF Nationals this year will be in priority order by A-value, rather than first-come, first-served. Once A-values are released, teams that have not qualified for ACF Nationals but wish to play should email [email protected] to register for the waitlist. As bids become available (either by teams declining or by field expansion), they will be offered to teams on the waitlist starting with the team ranked highest by A-value.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Thu Jan 31, 2019 12:12 am
by ryanrosenberg
A complete list of A-values for ACF Nationals 2019 may be found here. Teams that have qualified for ACF Nationals have been sent an email informing them of such and asking them to accept or decline their bids. A few important reminders:
  • Teams that have received a bid to ACF Nationals have until Wednesday, February 13th at 11:59 PM Central to accept their bid. After Wednesday, bids that have not been accepted will be released to the waitlist.
  • To get onto the waitlist, please email [email protected] to place your team on the waitlist. New this year, the waitlist will be ordered in priority order by A-value, so that the first team offered a bid as it becomes available is the team on the waitlist with the highest A-value. Teams that did not play Regionals will be placed on the waitlist below teams that did play Regionals.
  • Likewise, to register as a standby team, please email [email protected] to register your team as a standby team. Standby teams must show up to the tournament site the day of the tournament and will fill in for any last-minute dropped teams for free.
  • The field size is currently set at 42 teams; as we confirm team demand and staff availability, we will consider expansion to 48 teams, then 56 teams.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Thu Jan 31, 2019 12:15 am
by ryanrosenberg
An explanation of how the A-value system works, from someone who calculated it this year.

There are four key components to A-value:
  • Tossup points per tossup heard (TUPPTUH): How many points each team scored, only counting tossups and negs, divided by the number of tossups they heard.
  • Bonus conversion (PPB): Points per bonus.
  • Bonuses heard per tossup heard (BHPTUH): The percentage of tossups a team answered correctly out of all the tossups they heard.
  • Strength of schedule (SOS): The combined points per tossups heard of each team's opponents in other games, weighted by how many times they played each opponent, divided by the average points per tossup heard over the entire field. If a team has played a slate of opponents of perfectly average strength, this will be 1, with higher values indicating a tougher schedule.
The raw A-value calculation is

Code: Select all

20 * (TUPPTH * SOS + BHPTUH * SOS * PPB)
This is designed to be reflective of a team's performance in an average game, corrected for strength of competition. In addition, ACF also uses a correction for order of finish, designed to preserve the order of finish at Regionals sites in the A-value ranking. If a team finishes above another team with a higher A-value, the A-values of those teams and all teams between them in the standings are averaged and the resulting teams are sorted by order of finish at Regionals.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Thu Jan 31, 2019 11:24 pm
by Silverman
ryanrosenberg wrote: Thu Jan 31, 2019 12:15 am If a team finishes above another team with a higher A-value, the A-values of those teams and all teams between them in the standings are averaged and the resulting teams are sorted by order of finish at Regionals.
Victor, in the CMU site thread, wrote: Because of that, our A-value was averaged with both Gettysburg and Michigan A
I have a proposed change to the order-of-finish adjustment, inspired by this scenario. Before the adjustment, these three teams' A-values and placements were as follows:

Gettysburg: 257.272, 4
Penn State A: 319.240, 5
Michigan A: 258.911, 6

With the current system, the three A-values were averaged, since Michigan finished below Gettysburg but had a higher A-value, giving each team an adjusted A-value of 278.474. However, if instead the values are iteratively calculated from the top down (and reweighted as necessary), we would have the following:

Gettysburg: 288.256, 4
Penn State A: 288.256, 5
Michigan A: 258.911, 6

and no more adjustments necessary.

It seems to me that Penn State (or whichever team is in the middle in this scenario) shouldn't be penalized by the chaos around them. If Michigan's A-value had been 2 points lower, with the current system, Penn State (and Gettysburg) would have seen their A-value increase by 10. It wouldn't move them up any places in the overall A-value list this year, but there's certainly a scenario where this could be the difference between qualifying for nationals and not. (For 2019, this would also move Penn A and Hopkins A to 393.404, and Maryland A and Delaware back to their unadjusted values.)

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Fri Feb 01, 2019 12:24 am
by CPiGuy
I actually thought Steven's suggested formula was the actual procedure and was surprised to learn that it was not. I support the implementation of this procedure in future years.

also, Michigan contacted ACF asking to be added to the waitlist -- it's obviously not super urgent, but can this be reflected on the spreadsheet?

Edit: Thanks for doing this quickly!

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Fri Feb 01, 2019 12:31 am
by t-bar
Steven's proposal is essentially dual to what Dwight and I have done in the past--we start at the bottom and iteratively work up, rather than starting at the top and working down. It's not immediately obvious to me that one method or the other will lead to intuitively "better" or less disruptive results, but it's certainly something worth discussing.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sat Feb 02, 2019 11:19 pm
by Mnemosyne
Are people generally satisfied with the results of the strength of schedule formula? I was surprised that the Swarthmore site teams had a lower SoS than other top teams, considering the depth at the top of the field. I haven't paid attention to the years of discourse about these things, so maybe I don't understand something, but it doesn't look like it is properly ranking the top of the field. I looked only at PPB of opponents as a field-independent measure of strength and made this little table for comparison.

Just using the eye test, Delaware and Penn played an insane schedule compared to MSU's (comparing average, median, or even game-by-game), so why doesn't the Strength of Schedule reflect this? Are SoS calculations something negligible that no one really cares that much about?

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Sun Feb 03, 2019 9:42 pm
by tabstop
Mnemosyne wrote: Sat Feb 02, 2019 11:19 pm Are people generally satisfied with the results of the strength of schedule formula? I was surprised that the Swarthmore site teams had a lower SoS than other top teams, considering the depth at the top of the field. I haven't paid attention to the years of discourse about these things, so maybe I don't understand something, but it doesn't look like it is properly ranking the top of the field. I looked only at PPB of opponents as a field-independent measure of strength and made this little table for comparison.

Just using the eye test, Delaware and Penn played an insane schedule compared to MSU's (comparing average, median, or even game-by-game), so why doesn't the Strength of Schedule reflect this? Are SoS calculations something negligible that no one really cares that much about?
The thing your eyeball isn't taking into account is actual buzzes -- Penn and Delaware had a couple opponents who had low-ish PPB, but were also only getting two or three buzzes a game, and since SOS is based on points scored, that just compounds the effects.

Having said that, I don't see how the numbers in the spreadsheet match the description -- whether the zero TUH in the posted stats affected that or not I don't know, but in both of the ways I could think of to calculate SOS (either finding each opponent's PPTH, then averaging; or totalling up opponent's points and TU and then dividing to get PPTH) Delaware would have a higher value than either Penn or MSU. (I've shown my work in a Google sheet so you can critique my process. I didn't bother to figure what the nationwide PPTH figure was, so I didn't normalize my numbers.)

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Mon Feb 04, 2019 12:52 am
by Carlos Be
tabstop wrote: Sun Feb 03, 2019 9:42 pm The thing your eyeball isn't taking into account is actual buzzes
As has been brought up in other threads, tossup points are a bad way to measure team strength because they are incredibly field dependent, and they do not distinguish between teams past a certain skill level.

Imagine if to distinguish between two teams, either team went to a separate room and played a packet against empty chairs. There would be some gradation between weaker teams, but every team in the top 50 or so will be converting around 18-20 tossups. (This is assuming no negs, which would further mix up the stats.) Clearly, this does not distinguish teams.

Now suppose that we want to calculate the strength of the opponents of some team, say Berkeley A. We can calculate opponent tossup points by going through each opponent individually and then going through each round. However, since Berkeley played every team at the site, we can calculate the total tossup points by counting how many tossups were converted each round. In essence, this is the same as combining each pair of opponents into one team and then calculating how well they would play against empty chairs. Tossup points against empty chairs cannot distinguish the combined strength of the super-teams (UCLA + UCSD) and (Chicago + Yale) any more than they can distinguish between actual teams. Particularly in strong fields, variation in tossup points have more to do with negs and bad communication than they have to do with actual differences in skill.

The logic in not using PPB to estimate opponent strength is that some teams are better at converting tossups than at converting bonuses. However, under the current system for estimating opponent strength, such a team would actually make the strength of schedule statistic lower than if the low PPB team were worse at getting tossups. Imagine the following scenario: say team A and team B have exactly the same (low) ppb, except that team A buzzes one line earlier than team B on every question. In a round robin, replacing team B with team A would not affect opponent tossup points, since while A might be converting more than B, each other team would convert slightly less. In contrast, this would affect opponent PPB. Since A would hear more bonuses than B, the low PPB of A/B has more weight on the opponent PPB than it would otherwise. Thus, the opponent PPB would be lower and so would the strength of schedule, even though by definition A is strictly a better team than B.

Ironically, the entire model for A-value / D-value understands that TUPTUH is not a good measure of team strength— this is why there are strength of schedule adjustments in the first place. If TUPTUH is not accurate on its own to calculate A-value, why should we expect it to be accurate in calculating the strength of schedule adjustment?

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Mon Feb 04, 2019 2:36 am
by tabstop
justinfrench1728 wrote: Mon Feb 04, 2019 12:52 am
tabstop wrote: Sun Feb 03, 2019 9:42 pm The thing your eyeball isn't taking into account is actual buzzes
As has been brought up in other threads, tossup points are a bad way to measure team strength because they are incredibly field dependent, and they do not distinguish between teams past a certain skill level.

Imagine if to distinguish between two teams, either team went to a separate room and played a packet against empty chairs. There would be some gradation between weaker teams, but every team in the top 50 or so will be converting around 18-20 tossups. (This is assuming no negs, which would further mix up the stats.) Clearly, this does not distinguish teams.

Now suppose that we want to calculate the strength of the opponents of some team, say Berkeley A. We can calculate opponent tossup points by going through each opponent individually and then going through each round. However, since Berkeley played every team at the site, we can calculate the total tossup points by counting how many tossups were converted each round. In essence, this is the same as combining each pair of opponents into one team and then calculating how well they would play against empty chairs. Tossup points against empty chairs cannot distinguish the combined strength of the super-teams (UCLA + UCSD) and (Chicago + Yale) any more than they can distinguish between actual teams. Particularly in strong fields, variation in tossup points have more to do with negs and bad communication than they have to do with actual differences in skill.

The logic in not using PPB to estimate opponent strength is that some teams are better at converting tossups than at converting bonuses. However, under the current system for estimating opponent strength, such a team would actually make the strength of schedule statistic lower than if the low PPB team were worse at getting tossups. Imagine the following scenario: say team A and team B have exactly the same (low) ppb, except that team A buzzes one line earlier than team B on every question. In a round robin, replacing team B with team A would not affect opponent tossup points, since while A might be converting more than B, each other team would convert slightly less. In contrast, this would affect opponent PPB. Since A would hear more bonuses than B, the low PPB of A/B has more weight on the opponent PPB than it would otherwise. Thus, the opponent PPB would be lower and so would the strength of schedule, even though by definition A is strictly a better team than B.

Ironically, the entire model for A-value / D-value understands that TUPTUH is not a good measure of team strength— this is why there are strength of schedule adjustments in the first place. If TUPTUH is not accurate on its own to calculate A-value, why should we expect it to be accurate in calculating the strength of schedule adjustment?
I had interpreted the question as "why isn't PPB predicting the strength-of-schedule" rather than "shouldn't strength-of-schedule be based on PPB" in which case yes, my answer doesn't make a lot of sense.

I suppose someone could make an argument that a team that knows more things but can't score more points on the bonuses is not necessarily a better team, but I would agree that they are a harder team to beat and that's good enough for me. And your point about PPB being magnified is true, but maybe not as much as you think. WARNING: Numbers ahead.

Let's suppose a 10-team round robin, and our Berkeley A team has eight opponents that are averaging 18 PPB and one opponent that is getting 9 PPB, and let's just put a number in of 250 PPG for those opponents (averaged over the whole field) (this number is actually irrelevant but it gives us a baseline to start from). That means that Berkeley's SOS is based on their opponents putting up 18000 points in the games that count for SOS. If we secretly replace that bottom team with a team that is getting one more tossup each game (but also 9 PPB), then each of the first eight opponents is going to lose 28 points once, and the bottom team is going to gain 19 points eight times, so Berkeley's SOS will go from 18000 to 17928, or on a PPTUH basis from 12.5 to 12.45. This is a lot smaller of a gap than you see just from the difference between different prelim pools (for instance, just since I had already calculated them earlier, the difference between Penn A and Delaware -- same site, different prelim pools -- was 0.70 PPTUH). It may also perhaps be worth noting here that, if we are using PPB instead, the same thing would happen (since the PPB for this field would be lower than it was before since the bottom team is absorbing more of the chances), and actually by a larger relative amount (the field PPB would go from 17.1-ish to 16.7-ish). EDIT: Forgot to account for the other games not involving the replaced team! So it goes from 17.7-ish to 17.6-ish; still a larger relative drop but a lot closer.

This also means this team doesn't have to increase it's PPB to the average to get things back to even or better, since 8 bonus parts over the course of the whole tournament would do it. If they're only pulling down three tossups a game, that's 24 bonuses to work with so they need +3.33 PPB (i.e. up to 12).

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Mon Feb 04, 2019 3:00 am
by Borrowing 100,000 Arrows
tabstop wrote: Mon Feb 04, 2019 2:36 am I suppose someone could make an argument that a team that knows more things but can't score more points on the bonuses is not necessarily a better team, but I would agree that they are a harder team to beat and that's good enough for me.
This whole line of reasoning seems to presuppose that players can do well on tossups and poorly on bonuses. However, this seems like a highly suspect assumption. Tossups are testing the exact same knowledge as the bonuses just in a different form. Hence, a team that's good at converting bonuses should be commensurately good at converting tossups.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Mon Feb 04, 2019 12:04 pm
by Cheynem
I think in general you're far more likely to convert tossups (especially stuff like common link tossups) then have a high bonus conversion. I've frequently seen teams, for instance, that may not have a top fine arts or science player but convert a fair share of, say, music or bio buzzes on late clues or giveaways, but still only 10 most music/bio bonuses.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Mon Feb 04, 2019 12:19 pm
by ryanrosenberg
tabstop wrote: Sun Feb 03, 2019 9:42 pm Having said that, I don't see how the numbers in the spreadsheet match the description -- whether the zero TUH in the posted stats affected that or not I don't know, but in both of the ways I could think of to calculate SOS (either finding each opponent's PPTH, then averaging; or totalling up opponent's points and TU and then dividing to get PPTH) Delaware would have a higher value than either Penn or MSU. (I've shown my work in a Google sheet so you can critique my process. I didn't bother to figure what the nationwide PPTH figure was, so I didn't normalize my numbers.)
The SOS calculation uses tossup points per tossups heard, not points per tossups heard, and also excludes games against the team in question.

Re: ACF Nationals 2019 - 4/13-14/2019 - University of Pennsylvania

Posted: Tue Feb 12, 2019 2:49 pm
by ryanrosenberg
This is a reminder that in order to receive a bid to ACF Nationals (either by field expansion or teams declining bids), you must sign up for the waitlist. Any team, even one that did not play ACF Regionals, is eligible to sign up for the waitlist; teams will be taken off the waitlist in order of descending A-value, with teams that did not play Regionals behind those who did. To sign up for the waitlist, please email [email protected].