Pause game timer during Stop

[g3force]

Motivation:

• the referee does not have to remember to halt the game during discussions
• the time needed by the automatic ball placement procedure does not influence the net game time
• the net game time is roughly the same for all games

Disadvantages:

• the total game time may vary a lot more

[g3force]

Here are the stats of all the games from RoboCup 2017: match halt stop running avg 31.06 9.00 11.84 10.21 std 12.31 6.16 7.82 2.97

Match is the time taken in first and second half (and extra) in minutes. Stop is when the STOP command was issued, until another command except ball placement/goal was issued. The three game state times sum up to the match time.

Here are the relative times (game state to match time): halt/match stop/match running/match avg 27.70% 37.63% 34.65% std 14.54% 9.80% 9.88%

Here for stop vs. running (without halt) stop+running running/(stop+running) 23.09 48.13% 9.09 8.44%

When I remove all matches with a team that is not in the Top 8, I get:

match halt stop running avg 34.43 10.71 11.95 11.78 std 12.66 9.09 6.32 2.33

halt/match stop/match running/match 27.63% 35.08% 37.29% 16.05% 8.92% 11.23%

stop+running running/(stop+running) 23.73 51.13% 7.91 7.53%

Does not make a big difference.

When I interpret this correctly, the actual playing time (running) varies between 25% and 45% in 95% of the cases. That's quite a lot. If we would not count STOP to the game time, we would need to half the total time, according to this numbers.

You can find the sources here: https://github.com/g3force/ssl-log-analyzer/blob/master/gameStateAnalyzer.go

And the non-aggregated results are attached: matchTimeStats.ods.tar.gz

[g3force]

Suggestion: Apply the rule change and reduce the game time to 5 minutes per half

[mahi97]

Since the field size is increased, I think more time is needed in game-on for teams to reach the other goal and also new rules about chip and ball speed, avoid direct games, so passing is really necessary. IMO, reducing time will make the game too short to decide a genuine winner.

Plus 10 minutes game in a quad-field makes a new challenge for teams to manage robot's battery and try to optimize battery usage to keep the quality of their play in the whole game. Furthermore, we can limit the number of changes to keep battery usage as a real challenge.

btw, we should consider that there's a lot of game in the tournament, so if a game takes too much time scheduling will be a problem.

[g3force]

the overall time needed for a full match will not change significantly. The main change here is, that we do not count time during STOP state. According to the games from last year, the STOP state makes up 50% of the game time. So we reduce the game time by 50% to get approximately the same total time.

[g3force]

Depending on the number of games we will actually play, we could consider raising the game time a bit. Unfortunately, the number of fields is not fixed yet and thus the tournament schedule is not fixed either.