Many double touch fouls

[g3force]

There were many double touch fouls in recent competitions.
Investigate, why? Is it correct? Or are the autoRefs wrong? Might be too difficult to detect, when ball is not fully visible.

Also, the name is confusing, because bot does not necessarily need to touch the ball twice. Ball just need to move for more than 5cm and still touch the ball.

• Rename the rule?

What is the main purpose of the rule?

• Assumption: To encourage doing a pass after free kicks / kickoffs.
• Do we want to insist on a minimum precision for the kicks, or can be lower it. For example increase the move distance from 5cm to 15cm.

[guyfleeman]

I always assume the rules was to prevent a really good team (like tigers or ZHU) from starting dribble plays literally straight from a kick require a pass and as such require an opening for the ball to be stolen etc. I think that's what the effect has been and I think requiring an opening pass is good.

I obviously haven't done any log analysis, but my gut feeling from the last 5 years or so is that many Div B double touches are the result of a dribbler near miss (off by a cm or so with a narrower mouth) or a faulty break beam. Many of these result in a successful second kick that occurs from within the center circle. Similar behavior on free kicks, but I think what to do about it there is less obvious. I think removing DT and just giving a small manipulation radius, at least for Div B kickoffs, would make the gameplay a lot smoother and not have a meaningful impact on strategy. It's just a more generous execution window.

[tobiasheineken]

I generally agree with @guyfleeman, I wouldn't want to call it "removing DT" to increase the grace distance a little, but I feel that's just semantics. For the record: I think requiring a pass is the primary objective of DT (and is important), but we can probably increase the distance a bit if we want to punish a faulty robot / offset driving less. I'm not fully convinced that this is a good idea - because approaching a standing ball is (relatively) easly. If the "punishment" for failing to do so (the double touch) is aggressive, we give extra incentive to work on that. Which I do like, because I think it is an important stepping stone.

I am also slightly confused about Double Touch because I know that Er-Force has Code to avoid it. We might have the issue Will is describing, where we drive to the ball with an offset or with our break beam not working, but we should call another (different) robot to do the second touch instead, avoiding the double touch entirely. This is almost never successful, so I still have this nagging feeling that the autorefs might actually be wrong there. It's just very hard to see with my human eyes.

[g3force]

We concluded in the meeting today, that someone ideally need to invest some time in analysing the logs to better understand the root cause of the many double touch fouls.