Surveys & Samples: Charm Person Redux

A few weeks ago, I shared the poll on charm person that I posed to the big 1E AD&D group on Facebook. Shortly after I did that, I also thought to ask the exact same question on the ODD74 forum ("What can a charm person force on an enemy fighter?", i.e., when ensorcelling an enemy in combat), thinking that the opinions might be very different. This actually got more responses there than any of my prior polls, I think (N = 32) -- and more importantly a really valuable ongoing discussion. (Link; account required.)

The results are given in the table above. Given that respondents could pick multiple options, the percentages shown aren't exactly right. Here are the corrected numbers, showing what percent of voters approved each option: 

1. Attack former allies: 10 votes; 31% approval
2. Defend the caster: 25 votes; 78% approval
3. Surrender and disarm: 17 votes; 53% approval
4. Flee the encounter: 16 votes; 50% approval
5. Nothing: charm fails in combat: 2 votes; 6% approval

Now, the first thing that occurs to me is how surprisingly similar these results are the poll of AD&D players. Again in that case: there was clear majority support (around 80%) for "defend the caster"; around 50% support for fleeing or surrendering; just a minority (20-30%) that support "attack former allies"; and almost no support (6-8%) for the "nothing" option. 

That said, it's a bit awkward that the fleeing/surrender options consistently get around 50% support -- making that an issue of ongoing contention and no clear consensus. Personally, when I first created those options, I assumed that those were clearly weaker possibilities than "defend the caster" (participating in combat at all seems like more power to the spell, and seemingly more risk to the victim), and that therefore anyone picking the latter option would surely also pick the other two. Clearly I was incorrect. (Thanks to the Discord advance comments that coached me not to make the assumption that the options are all well-ordered.) 

On a personal note, it's fascinating to find out that I've been well off the reservation for most of my gaming career, because the option to "Attack former allies" was something I always enforced, reading the O/AD&D language as clearly permitting that (as well as almost any other direct control desire). But simultaneously, it did always bug me a bit as making for overly swingy combats. This is a case where I'm very happy to hear the voice of community experience. 

Also, one of the great parts of the ODD74 conversation was the observation of a fine distinction in the OD&D magic items of control. To wit: the potions of human, giant, and dragon control each refer back to the charm person/monster spells for their effect. But on the facing page, the ring of mammal control does something different: it says, "Control is complete, even to having the controlled mammals attack the others with it which are not controlled." See it seems like a compelling argument that the latter capacity ("complete control") is not included in the basic charm spells, or else it would not be so called out in this one case. (Big thanks to SebastianDM for picking up on that detail!)

So the next time I edit my custom Book of Spells, I'm pretty likely to edit in the limit against attacking former allies when charmed (or at least, you know, add a footnote on the issue). That would certainly have helped me on numerous occasions over the years in the past.

What's your justification for why fleeing/surrendering are considered by many to be less achievable than the "defend the caster" option?