Weird thing happened while driving

[marksi]

If as you are suggesting, the merge happens at the last possible point, you would be saving no more time.

So I don't see the point really.

You're not considering how the situation initially arises.

Lane is closed. As long as there is sufficient traffic, at some point shortly afterwards a small queue will form in the approach to that single lane. It may be a queue of, say, 10 cars. If another car joins at the point the lane is coned off, it is jumping 10 people who are already waiting to pass the obstruction, which is why the driver of that car merges at the back of the queue. If traffic flow heading towards the obstruction continues to be greater than that passing it, the queue will lengthen and people in the closed lane will have to merge at some point; if a merge point (10 cars back) already exists, that is where the one merge will happen. If a person jumps the merge point and tries to merge at the closed point, he/she will cause a secondary stop-go merge to happen and cause further delay to the traffic queue.

[cdd]

Aha, an interesting point there. I'm referring to situations where the first merge point is an 'informal merge' (no queue, just individual cars 'randomly' migrating rightwards). If there is a queue and you go AROUND it, that is clearly juming ahead and is unfair.

My objection is where that queue forms, i.e. earlier than necessary. The current system most people follow results in a one mile queue over one lane, rather than a half-mile queue over two lanes. The merge has to happen at some point, I still think the later the better.

[marksi]

My objection is where that queue forms, i.e. earlier than necessary. The current system most people follow results in a one mile queue over one lane, rather than a half-mile queue over two lanes. The merge has to happen at some point, I still think the later the better.

How will you get through it quicker if that is the case, and if you aren't getting through it quicker, why is that better?

[cdd]

Cars merge X metres earlier than necessary.

Only Y cars can flow through a road in any length of time (Y rate).

If two lanes are available then Y*2 cars can flow through the road in the same length of time.

So, for the length "X", the flow is Y rather than Y*2.

The scale of that effect depends on the length of X, and quite frequently X is surprisingly large.

[marksi]

The maximum flow of cars through the single lane obstruction is a constant.

[cdd]

Indeed, hence the effect only applies for the above length variable of X - the distance after cars have merged, but before the one lane region begins.

[marksi]

Indeed, hence the effect only applies for the above length variable of X - the distance after cars have merged, but before the one lane region begins.

This does not get you through the obstruction any faster.

[cdd]

Of course not, but it avoids an unnecessary queue leading up to the obstruction - again, for the length of X.

(EDITed to make this make a little more sense): I undersatnd where you're coming from: the rate of cars ("Y") is reduced, because merging at the end reduces the speed of everyone. But because this slower speed takes place over two lanes instead of one, you are still doubling the road capacity over that crucial lead-up length of "X". The speed would have to be halved or more for there not to be a gain, and I don't think that joining at the end halves the speed - it only reduces it by a small amount.

[davidmcg]

Don't even know if this is even relevant but I feel that I lost 3 hours on Call of Duty tonight, switched off the console and thought it was 8 when it was 11, very wierd experience. Does anyone else tend to get this when gaming?