Online League Format Changes on January 1

Published 5 Dec 2020 by antti (last edited 25 Dec 2020)

Starting from 1 January 2021, we will make two changes to our online league format, one a large one and the other a smaller one.

Tl;dr version

  1. Pairings will be rolled out monthly instead of bimonthly, are determined directly by league standings, and your results will affect your league standing.
  2. Games that end because of mistakes and oversights will start to get relatively less focus compared to games that are cl0se throughout.

Long version

Up until now, new pairings in our online league were done twice a month arbitrarily by our teachers. As the first, larger change, to make the games feel more competitive and the pairing system more predictable and transparent, we will switch to a stricter league format with five players in each league (except for the lowest one, which will have 4–9 players), which players getting promoted and demoted once a month.

The second, smaller change is that we will slightly adjust the level of depth in the game reviews. All games will still get reviewed, but some games in more and some in less detail. This is to reflect the fact that some games simply end up getting decided more quickly because of larger mistakes. In order to keep the system fair, we will make sure that half-option participants are guaranteed one ‘in-depth review’ and full-option participants two ‘in-depth reviews’ per subscription – however, the actual numbers will probably be at least double the above.

While the pairing and review systems will change a little bit, as you can see, our core service remains the same. So do our pricing options at €40 for four reviewed games in two months and €80 for eight reviewed games in two months plus Jeff’s coaching.

Below is a description for how the league pairing schedule works. Because a five-person league can have a varying number of half and full-option participants, the number of games played and therefore the exact schedule will vary.

The general process is as follows:

  1. At the start of a month, participants have been ranked in order by their previous results.
  2. New participants are placed in the list at positions roughly corresponding to their playing level, and leaving participants are removed from the list.
  3. The participants are divided into leagues of five players each, starting from the top. The bottom league will consist of 4–9 players, according to the teachers’ discretion.
  4. Inside the leagues, the players play two or four games in one month, depending on their subscription option, against players determined by the pairing table.
  5. After a month’s games are finished, the players are then sorted according to their results. The main criterion is the player’s winning percentage in that month’s games, and the tiebreaker is the player’s current position in the league.
  6. After the players are sorted inside the leagues according to their results, the bottom two players in each league switch places with the top two players in the lower league. After this, the leagues are deconstructed, and the process repeats.

Press below to see an example of the above process.

Example of re-sorting by a month’s results

League table with results

#Name12345Win-%
1A101050%
2B001150%
3C111075%
4D000125%
5E101050%

Re-sorted league table before promotions and demotions

#NameWin-%
(main criterion)
Former placement
(tiebreaker)
Result
1C75%3Promoted
2A50%1Promoted
3B50%2Remains
4E50%5Demoted
5D25%4Demoted

Internally, the system is therefore fair and predictable when ranking existing players in order by the results. However, there are cases when a player would ‘normally’ get promoted or demoted, but this doesn’t happen because of participants joining or leaving the online league.

Below you can see in detail how the pairing schedules look like depending on the number of half and full-option participants. The schedules have been fashioned so that, generally, higher-ranked players play higher-ranked players and lower-ranked players play lower-ranked players.

The ‘all-half’ pairing schedule
#12345
1
2
3
4
5
One full option and four half options

If one participant has the full option and the others the half option, then the all-half pairing schedule is followed, with the exception that the full-option participant plays everybody. This removes the pairing between the players that the full-option participant would not play against in the all-half schedule. For example, if player 3 has the full option, the schedule looks like this:

#12345
1
2
3
4
5
Two full options and three half options

If the two full-option participants would not normally play against each other in the all-half pairing schedule, then all-half schedule is followed and the full-option participants play two games against each other.

If the two full-option participants would play each other in the all-half schedule, then they play a second game with each other, and the rest of the pairing is adjusted while minimising rank differences in the pairings. For example, if players 3 and 5 have the full option, the schedule looks like this:

#12345
1
2
32
4
52

Compared to the all-half schedule, in this case, the 2 v. 4 pairing has been removed, and the 2 v. 5 and 3 v. 4 pairings have been added.

Three full options and two half options

This schedule follows the all-half schedule, but all full-option players play an extra game with each other.

Four full options and one half option

If four participants have the full option and one participant has the half option, then the schedule otherwise follows the all-full pairing schedule, with the exception that the half-option participant only plays pairings according to the all-half pairing schedule, and the opponents not included in his/her pairing play an extra game against each other. For example, if player 2 has the half option, the pairings are:

#12345
1
2
32
4
52
The ‘all-full’ pairing schedule
#12345
1
2
3
4
5

This post has no comments so far. Please log in to comment!