Fear of Timing Out - An Examination of Online Sealed KeyForge Game Lengths and Player Attitudes

As part of my survey on the KFPL Season 2 Sealed Qualifiers on Saturday 23rd January 2021, I conducted a data-gathering exercise on games going to time. This had two parts. The first was keeping a record of the games which exceeded the 45 minute time limit; the second was a post-event survey on player attitudes. Here I will present my findings.

The Stats

Over the entire weekend, players took part in a total of 499 games as part of the qualifiers. This encompassed both the Swiss and single elimination rounds. Out of 499 games, 2 exceeded the 45 minute time limit. This was a grand total of 0.4% of all games in the event. Interestingly these both occurred in Qualifier 4, which began at 12pm EST. No game in Qualifier 3 (commencing at 4am EST) ended due to the clock running out.

There was no overlap between the 4 players whose games exceeded the time limit, and both games ended in quite different circumstances.

The first game can be viewed on the event VOD at 9h15m. This game was particularly board heavy - with 7 minutes remaining on the clock, one player has almost 3 keys worth of aember captured on their battleline. That player also has an Adaptoid in play, which results in an increased number of triggers to resolve, and substantially extends the length of turns. Ultimately I believe this game was a match-up specific issue where two slow, capture-heavy decks were pitted against one another.

The second game unfortunately is not on the event VOD, but went to time for another reason entirely. Near the end of the game, both players made a rules error resolving a specific card, and entered manual mode in an effort to ‘correct’ it. Having corrected it, then continued to play the game, they soon realised they were both wrong - however, the game state was now in a position which was extremely difficult to repair online. The players attempted to fix the board using manual mode, which took up the last few minutes of the game and resulted in the players exceeding the time limit. This circumstance was very unfortunate and wouldn’t have happened during in-person play.

The Survey

Once players had dropped from or completed the event, I asked them to complete a short post-event survey. The response rate was 42% across both events. The questions on the survey were as follows:

  1. Did any of your games today go to time?

  2. If so, how many games?

  3. If so, how do you feel about the end of game procedures?

  4. Did you win any of your games due to the end of game procedures?

  5. If so, how many games?

  6. Do you feel you had adequate time to complete your games?

  7. Please explain.

  8. Do you feel like your opponents used a much larger amount of time on their turns than you did?

  9. How do you feel about this?

  10. Please explain.

  11. If you suspected slow play, did you raise this with a TO (tournament organiser)?

  12. If so, what was the TO’s response?

The goal of these questions was to measure participant sentiment towards the end of time rules, and observe the difference between respondents whose games had been resolved using tie-breakers and those who had not. Unfortunately due to the fact that none of the four players whose games went to time completed the survey, checking for a difference is not possible. The lack of games which were resolved using end of game tie-breaks also makes the first 5 questions moot. (I should note here that plenty of participants gave their opinions on the end of game procedures in the question 3 response area, but as none of them were affected by end of game procedures in this event, I will not be examining the responses to this question.)

I will be examining the rest of the data in the surveys to evaluate participant sentiment.

Question 6: Do you feel you had adequate time to complete your games?

98.7% of respondents answered ‘yes’ to this question; 1 participant responded with ‘no’, but the follow-up response indicates this was because they had “too much time” rather than too little. Most of the responses to the follow-up question are similar, commenting that 45 minutes was too much time in each round.

Several respondents remarked that they feel more time pressured in an Archon environment with a 35 minute time limit - this is definitely an area for further investigation.

Below is a word cloud which shows the most common terms used in the responses to this question.

Question 8: Do you feel like your opponents used a much larger amount of time on their turns than you did?

The majority of participants (62.7%) did not feel that their opponents used a significantly larger amount of time on their turns than they did. Only 14.7% of respondents felt as though their opponents used significantly more time than they did the majority of the time.

Question 9: How do you feel about [whether your opponent used significantly more time on their turns than you did]?

Answers to this question were rated on a 5-point scale, where 1 was “I felt very bad about it,” and 5 was “I felt very good about it”. Most respondents (94.6%) either felt good about their opponent’s use of time or didn’t care.

The responses to the follow-up question are mostly positive about other players' conduct and the time allowance for the games. There was a clear difference between players who are slow (which most respondents were supportive of), and players who deliberately played more slowly to manipulate the end of game procedures. The clear sentiment seems to be that as long as games did not resort to using the end of game tie-breakers, players did not mind if their opponents used more time than them to resolve their turns, particularly if their opponents’ decks were visibly more time-intensive (e.g. having a larger board).

Below is a word cloud which shows the most common terms used in the responses to this question.

Question 11: If you suspected slow play, did you raise this with a TO?

For context, ‘slow play’ refers to when a player purposefully procrastinates on their turn to waste time remaining on the game clock. This can become a particular issue when a game is close to being resolved with end of game tie-breaks.

85.3% of respondents responded with ‘not applicable’ for this question, meaning that the majority of participants did not experience any slow play. This is supported by the positive responses to the previous question.

The respondent who raised the issue of slow play with a TO had an immediate response. The TO then observed the remainder of the game, which completed before time was called. The respondent was happy with the way the situation had been treated by the tournament organiser.

Those who suspected slow play overwhelmingly did not raise this with a tournament organiser - one follow-up response explains that this is because they have never seen a judge sanction anyone for slow play.

Conclusions and Lessons for the Future

Games going to time in a sealed format online is a non-issue - a tiny minority of games went to time, and the two that did did so for quite different reasons. An uneven use of time in a game is also not an issue for the overwhelming majority of players, as long as the game is completed within the time limit. Judging by this, I think the case for changing time distribution within KeyForge games (e.g. by introducing Chess clocks) is extremely weak and almost unjustifiable.

The fact that the majority of comments concerning time use were conditional on tie-breakers not being used points to the fact that the main issue is, and in my opinion always has been, the tie-breakers. Finding a better way of resolving tie-breaks should be the goal of any discussion concerning end of game resolutions.

The most concerning issue that has arisen from this is a lack of confidence in resolving genuine slow play. 13.3% of players felt there was an issue with slow play and did not approach a judge about it. The reasons for this should be explored in more depth. In response to this, I believe there are a number of measures which could be implemented to increase player confidence in calling judges about slow playing opponents:

  1. Implement timestamps on the TCO game log - This would allow players and judges to see the amount of time spent resolving individual actions, and would allow the judge to make a more informed call on the pace of play despite not viewing the entire game. This is the most important asset that needs to be added for judges to action slow playing complaints.

  2. Set player expectations - Tournament organisers and judges should set the expectations for a reasonable pace of play in a pre-event briefing, and outline the sanctions for not behaving in line with these expectations.

  3. Meet expectations - Consistently apply sanctions according to the information given out in the pre-event briefing.

  4. Raise player confidence in judge calls - This is more complicated and is something that will need more time. Meeting player expectations by sanctioning slow playing participants will contribute towards this, as will outlining the expectations in pre-event briefings. There should be a range of judges who are clearly identifiable that players feel comfortable and confident in calling upon for quick and fair responses.


  • 0.4% of games went to time.

  • 94.6% of players were not upset if their opponent used more time on their turns than they did, as long as the game was completed within the time limit.

  • The majority of players are only concerned about time use if they are threatened by tie-breakers.

  • Players who experience slow play from their opponents do not report it to judges. Raising confidence in reporting should be the priority going forwards.


The full anonymised responses to the open-ended survey questions are included in the appendix here.

Kate Dunstone, also known as Muffins on Discord and lotsa_muffins on TCO, is a long-time KeyForge player and administrator for the KeyForge Premier League. She tweets as @KeyForgeLeeds and @Lotsa_Muffins, proudly reps Team Archimedes, and her favourite card is Zyzzix the Many.

317 views0 comments