Previously, the ballot (on which judges record their marks and their vote for the winning side) was a triple carbon copy, with the top copy in white and the other two copies in different colours (typically pink, green, yellow etc.) The past practice was for Convenors to organise the printing of ballots prior to the tournament.
However, the introduction of computer-generated personalized ballots has meant that ballots can only be printed the night before each day of debates. While this might appear time-consuming, the trouble saved from decoding handwriting, tracing down judges etc. has made this endeavour worthwhile.
Design of ballot sheetEdit
After a history of discussion and experimentation, confusion has been resolved as to how Points of Information, time penalties and other penalties are recorded on the ballot. The offering of POIs is now marked in a separate column on the marksheet and must range between -2 and +2. Mark Gabriel from Singapore (CA in 2007) has designed a useful guide for the marking of POIs. That guide will appear here.
Apart from such penalties/rewards, the ballot always has separate boxes for the Style, Content and Strategy marks for each speaker. Each mark may only be within a given range.
A ballot will also have boxes for:
- the round number (0-8 or the name of the break round)
- the motion for debate
- the date
- the Proposition and Opposition teams
- the name of the judge
- the total team marks
- the winning team (which should be double-confirmed e.g. “Slovenia – Opposition”)
- the judge’s signature
In the case of personalized ballots, the judge's name, round number, date, and Proposition and Opposition teams will be pre-printed. At the bottom of each ballot, every member of both teams will be assigned an alpha-numeric code e.g. SIN-1. It is important to emphasize to judges that they should check the codes against the speaker order before filling it in. Judges should also write down the name of each speaker so that the CAP can re-confirm which speaker was speaking in each round in the event of a mistake.
The point of the tabEdit
From Rule 7: At the end of the preliminary rounds, teams shall be ranked according to the number of wins. If teams are tied on the same number of wins, they shall be separated where practicable by elimination debates and otherwise on the following priority: (a) number of adjudications in favour of the team; (b) average judges' scores for each team.
The Rules don’t say it, but should two teams be tied on average judges’ score (which equate to total team points, as all teams will have debated 8 times), then a coin toss will be used to separate the teams. There used to be a Rule about looking back to a possible preliminary round meeting between two such teams, but it was removed in 2002 (see the Council Meeting minutes).
The point of the tab (a computer-run, tabulated record of the ballots compiled by every judge in every debate) is therefore to allow the CA to determine the break according to the requirements of Rule 7.
What the tab should doEdit
Different years have run different tabs and collected different sets of information. The ideal would be that the tab collects every piece of information on every ballot. An argument against this in the past was that it takes a long time to enter all the data – but in fact it takes about an hour to enter every ballot from one round of 32 teams, and with the advent of the CAP plus larger Organising Committees, there should be plenty of ability to do this.
In reality, a tab generally collects everything except the breakdown of Style /Content/Strategy marks – instead the tabulators will enter the speaker mark for each speaker, as given by each judge. In recent years, though, the tab has captured the breakdown as well to serve as a check-and-balance mechanism against calculation errors. These should be recorded against the speaker, the team and also against the judge, so that judges’ overall marking patterns can be monitored.
Any spreadsheet can easily use formulae to
- check that the judge’s addition of speaker marks to reach a team total are correct; and
- check that the overall vote is correct, based on the speaker and team votes submitted.
It can then automatically generate:
- round-by-round results (e.g. “New Zealand 2 – Australia 1”), which can be printed off or published online;
- a round-by-round tab, if desired, of teams ranked on wins and judges;
- the final ‘break’ tab; and
- if required, an individual speaker tab (see below).
Judges make mathematical mistakes – with or without calculators, they happen. A responsibility of the CAP is to make sure that such mistakes are corrected and don’t affect the competition. A rule of thumb of recent years has been to double-check every ballot and correct them, if necessary, before results are published .
If every speaker mark is entered into the tab, the tab can double-check easily that the addition of speaker points to reach a team total is correct. If only team totals are entered, then the CAP is trusting that the maths on the part of the judge is correct. When debates are won or lost by half a point in some cases, it makes sense that the speaker points should be entered.
Section 6.3 of the Notes for Adjudicators essentially says that a judge’s instinct and marks really should coincide. As a CA, if the two part company (i.e. a mathematical mistake allows one team to win while the judge has written the other in as winner), then you would tend to trust the judge’s written decision and adjust the marks in an appropriate manner to make them tally. (Of course, you would speak to the judge first – but a judge would generally prefer to defend the team they’d voted for rather than defend their maths.)
Practice in 2001 and 2003, and perhaps others, was to check the maths on each ballot and, where a mistake occurred, correct it in another colour ink and have one of the CAP sign it. This is fair to the teams and coaches are reassured that the CAP is monitoring every ballot.
The current practice is to speak to the judge concerned before making any amendment to the ballot. As a general principle, it is best to have CAP members check the ballots at every venue during lunchtime or on the bus back as it is becomes harder to hunt down a judge once everyone has retired to their rooms in the hotel.
- Judges forget to fill out all the boxes.
They may forget to complete the date, round, their name or often the winning team. If this happens, track down the judge after the round and remind them to do so in future rounds.
- Judges don’t stick to the mark scheme.
For example, they give speaker marks outside 60-80 or, more commonly, Style/Content marks outside 24-32. In this case, it’s a matter of importance that the CA talks to the judge and makes sure they understand the mark scheme before they are allowed to judge again.
- Speaker names.
Usually it’s OK for judges to use first names of speakers on ballots. There may be teams however with two speakers with the same first name – this has happened recently, where both speakers had the same first initial of their surname as well. Make sure that judges know to differentiate if they need to.
- The maths don’t add up.
Running the tab: The technical bitEdit
Various tab programs or spreadsheets have been used over the years (including index cards in early years with fewer teams). The spreadsheet used in 2001, 2003, 2006 and 2007 will be made available online. It’s by no means perfect, but gives an idea of what a tab might do.
The key facilities of a tab, ideally, should be:
- accept speaker marks, speaker names and judge’s name from each ballot;
- add up the speaker marks to reach a team total;
- compare the team totals and generate the winning team name;
- produce a round-by-round list of results (teams and judges’ votes);
- tabulate a ranking of teams based on wins, judges and speaker points
- tabulate an individual speaker ranking based on main speeches (see below);
- alert when panels of judges contain a judge from one of the team’s own country;
- alert when a panel of judges contains two judges from the same country;
- alert when a judge has judged the same team more than X number of times;
- alert when any two judges have judged together more than X number of times;
- alert when any individual mark entered by a judge is outside the mark range;
- allow judges to be ‘graded’ (with subjective input from the CAP);
- display the ‘quality’ of any judging panel according to the grades of its judges (e.g. AAC);
- record the venues for each day’s debates;
- alert when teams are assigned to different venues in the same day;
- alert when judges are assigned to different venues in the same day;
- evaluate judges’ average marking ranges;
- alert when a judge’s marks significantly varies from his or her panel members
- producing results in a format that could be published online easily, during the tournament.
The Discussion page has a couple of extra comments on this issue.
With the expansion of the tournament, it has become necessary to ensure a high level of accuracy with the tab. In recent years, a shadow tab has been run in conjunction with the main tab. Typically, the shadow tab takes the form of a more primitive spreadsheet, which can capture a team's win/loss record, number of ballots, and total scores - the bare minimum required to determine the break. The shadow tab should be run by a completely different set of volunteers from the main tab, to prevent mistakes from being repeated.
At the end of Days 2, 3, and 4, the shadow tab should be reconciled with the main tab to identify any discrepencies. Where inevitable deviations appear, the relevant round(s) for each team should be isolated and checked against the paper ballots. The rule of thumb has been that the break should not be announced until the shadow tab completely matches the main tab.
Up to and including 1998, there was a knock-out “individual competition”, run in parallel with the main competition, which sought to entertain as well as mix competitors in a forum outside the main debates. Unfortunately, drunk teenagers being what they are, a number of poor debates or speeches ensued including embarrassing showpieces in front of sponsors and dignitaries.
England chose in 1999 to abolish the individual competition and instead release an individual ranking based on main speech scores in preliminary debating rounds for those speakers who had taken part in at least 4 prelims. Awards were made to the top 3 speakers at that year’s Grand Final.
In 2000 the top 7 speakers were announced at the Final. In 2001 the top 10 were announced and congratulated at the dinner after the final (limited to participants only), and an award made to the top speaker. In 2002 and 2003 similar awards were made.
At the 2004 Council Meeting the Council expressed concern about the individual awards. An indicative vote was taken as to whether they should continue: 8 countries voted that they should, 10 against, with 4 abstentions.
In 2004 the Convenor and CA decided not to make individual awards and indeed speaker marks were not recorded in the tab. However, after a number of complaints were received about this, individual rankings returned in 2005 and since that time, with prizes awarded for the top 3 speakers.
There are many arguments for and against of which the discussion minuted above includes many. They include:
- For: recognition of individual achievement, CV points for college applications, etc.
- Against: divisiveness, arbitrariness, etc.
Without guidance from the Rules, it’s entirely up to you as CA if individual awards should be made, and if they should, how many.
When individual awards are made, however, it is vital that they are cross-checked for accuracy as well. Typically, a member of the CAP will discreetly approach coaches of the top 5 or top 10 speakers to ensure that their charges have spoken in the number of rounds recorder on the tap.
See Trevor Sather's additional comments on the Discussion page.
Viewing of ballotsEdit
In the past, the expectation was that copies of ballots would be distributed during the tournament to coaches. There was no Rule about this – so they could be handed out at the end or not at all – but generally coaches were eager to acquire them sooner.
Currently, the practice is to keep all ballots in a thick ring file at the end of each day, and to allow coaches to approach the CAP to view the ballots. The ballots should be sorted by day and by round, then alphabetically according to the Proposition team name. It makes sense for ballots for each debate to be kept in a transparent sheet protector, to facilitate easy retrieval.
Crucially, ballots should only be made available for viewing after they have been checked, tabulated, and corrected for mistakes.
Future improvements to the tab or online publication (perhaps with password protection for coaches) may eliminate this administrative task from the CAP’s duties.