Slitherine

There's been some confusion over how retreat works in this game. As far as I know, a unit will always retreat in the exact opposite direction of attacker. If that hex is taken, then the unit will select one of the next two adjacent hexes* to retreat to, and if those are taken, one of the two hexes adjacent to the attacker. If all possible retreat hexes are occupied, the unit will surrender. A unit must be fully suppressed and have an entrenchment** of 0 after combat for retreat to occur.

*See Part 2 for the "ambiguous" cases, which turns out to be completely predictable.
**The entrenchment level of the defender can be 1 or 0 at the start of combat, since combat automatically reduces the entrenchment of the defender by 1.

Part 1: The Simple Cases


Figure 1: The defender retreats directly opposite of the attacker.


Figure 2: Since the hex directly behind the defender is blocked, it selects the only open adjacent hex to retreat to.


Figure 3: See Figure 2.


Figure 4: There is only one hex open, so the defender retreats in the attacker's direction.


Figure 5: All adjacent hexes occupied, the defender surrenders.

Part 2: "Ambiguous cases"

What about a case like Figure 6 below?


Figure 6

According to Rudankort, in theory, there should be a 50/50 chance of selecting either hex A or hex B. I believed this for the longest time. However, when doing testing, I did not find any randomization. I propose that these cases are 100% predictable. The defender will always retreat to the hex that is "right-handed" if seen from the attacker's perspective. Examples are illustrated Figures 7-16 below.


Figure 7


Figure 8


Figure 9


Figure 10


Figure 11


Figure 12


Figure 13


Figure 14


Figure 15


Figure 16

Thanks deducter. I'll copy this link to the FAQ.

Thank you very much for this information.
The easy cases I already knew, but that right handed thingie was news to me.
Thanks again.

entirely plausible

Good job finding out about this deducter. This explains why my MP opponent and I witnessed a P26 of his retreat into a city hex while there was one other legit retreat hex available. The poor P26 got whacked of course.

deducter wrote: According to Rudankort, in theory, there should be a 50/50 chance of selecting either hex A or hex B. I believed this for the longest time. However, when doing testing, I did not find any randomization. I propose that these cases are 100% predictable.

For the record, I never said there was randomization in this case, I just did not want to complicate the explanation with such edge cases. You are absolutely correct that there is no randomization at all in retreats. The game builds a list of "retreat candidate hexes" and then just takes the first which is available. You can always know 100% where a unit will go if it retreats.

This thread is a very good explanation, thanks

retreating units always move clockwise when a direct backward retreat route is blocked. That has been the case from Panzer General through this game. There is nothing random about this.

Deducter, thanks for this, it explains a lot.

Let me posit a slightly different deployment to your figure 14. Assume there is no 10-strength infantry unit in the right-most hex. My experience and what I've complained about, is that the retreat you indicate will take place DESPITE what is now (an unoccupied) 'safe' retreat hex to the lower right. Thus the retreating unit ignores a 'safe' hex and retreats forward to danger.

Is that congruent with your tests?

It would retreat into the hex that has the 10 str inf in it. Set it up in the editor and test it.

Herding enemy units into an ambush is a tried and true tactic of warfare. I would argue that in such cases where units apparently retreat into more danger exactly represents such occasions.

Although in this case, I believe your modified version of example 14 was already explained in example 2 and 3.

Oh this is huge

HUGE

I wish I would have known this long, long ago...

Thanks Pete for pointing it out on my latest video.

thanks for the detailed explanations, I didnt know these mechanics and they help a lot in understanding how surrenders works.

One question, does unit nationality or scenario parameters add anything on these rules?
For example during late scenarios in Grand Campaign (us based) I had a lot of difficultiesin force surrenders.

faos333 wrote: ↑Sat Mar 09, 2019 12:53 pm thanks for the detailed explanations, I didnt know these mechanics and they help a lot in understanding how surrenders works.

One question, does unit nationality or scenario parameters add anything on these rules?
For example during late scenarios in Grand Campaign (us based) I had a lot of difficultiesin force surrenders.

Not sure re unit nationality or scenario parameters, I don't think so per se... maybe the units were just stronger (higher defense values) and/or deeper entrenched. Playing at the setting AI 2 makes units leave their entrenched positions more often/easily as the AI is playing a bit more aggressively...