B/X wilderness encounters are bonkers

Last post was about the effort required to roll a single random encounter in B/X.

This post is about how this effort often leads to results that must be immediately discarded or changed.

This impression started when my players actually rolled two dragons in a row last week (in a three day trip). 

The first one (hydra) attacked on sight, but the second (three blue dragons) were indifferent so I decided they'd ignore the party.

If it had been the opposite, I'd have to fudge the dice or get a TPK.

And I do not think you could say I'd be "interpreting" the rolls. 

If you roll an "immediate attack" reaction, not attacking the PCs is simply changing the dice to save them, and even a "hostile" dragon could attack surprised PCs immediately - unless the GM wants to save them.

It is not unlikely that the players would be killed immediately by dragon breath (40 points of damage for each dragon, save for half) - before they could ever think about talking or evading (for example, if the are surprised or lose initiative).

Would it be fair game if that happened?

I am not sure it would.

First, the chances of finding 1d4 dragons in the wilderness are enormous - about 6% for every encounter, much greater if you're in the mountains or if you include hydras and wyverns (they are also under the "dragon" sub-table).

If you make one check per day, you might encounter 1d4 dragons per month in the forest, but the book suggests you could roll three or four times. You'd find 1d4 dragons every couple of days in the mountains!

I have never seen so many random dragons in any story, not even in Dragonlance novel (one random dragon was too much for me in the last one I've read, TBH). Maybe in Fire & Blood.

Also, every dragon has an identical chance of appearing - despite the fact that the book says blue dragons live in deserts and plains (mine appeared in a forest).

The basilisk is both under "dragon" and "unusual" - which means it is three times more "usual" than hawks in the mountains!

Dragons are more common than wolves and hawks!

OTOH, there are no gorillas anywhere, just albino gorillas for some reason.

But should the players know that wilderness adventuring is very dangerous in B/X? 

Again, not sure; we played multiple games (including a recent 5e campaign) and there was nothing like that in any other campaign.

Now I'm  wondering if other groups that play B/X use the tables as written or just change the results to save their PCs from three blue dragons attacking at once.

But fudging and "interpreting" the results in such a manner feels like more work to the GM. I'd prefer having a random table that gave me results that I can actually use. 

The OSRIC tables looks much better at a first glance (I'd have to try them in practice), but they require more complex/expansive sub-tables (including "harlot" tables that I'm unlikely to use).

I'd probably be better off creating my own tables so I can adapt it to my tastes. My ideal would be something like:

- One single table per terrain, no sub-tables.

- Fewer chances of multiple dragons.

- Encounters adequate to the terrain are more likely (e.g., dragons of the appropriate type, but also no salamanders except in extreme heat/cold, etc.).

- Probably adding number of creatures to the encounter table (e.g., "2d4 salamanders" instead of "salamander").

- Probably using a d100 or a bell curve.

- Add undead to forests because I like the idea!

And, of course, no tables "by CR" or "by PC level" - the world does not revolve around PCs' needs. I have a different solution for that, which deserves a post of its own.