Has anyone tried this? Does it break the game? Does it make it easier/quicker?
Assuming a 1 gave no resources to anyone and a 7 still invoked the robber how different a game would this make Settlers?
Significantly. The thing about 2d6 is that you've got a range of numbers that have a certain probability of them showing up.
When you're rolling 2d6, your available numbers are 2-12. You're dropping off the number 1. Also, you have an escelating scale of percentage probability. 2's and 12's will happen 2.77% of the time (each). 3's and 11's will happen 5.55% of the time. 4's and 10's happen 8.33% of the time. 5's and 9's happen 11.11% of the time. 6's and 8's will happen 13.88% of the time. And 7's will happen 16.66% of the time.
As you can see, the closer you get to 7, the greater chances you have of something on those hex's from happening. Because of this, the game has been weighted so that certain places will not show up as often as others. That's part of the resource management and placement. It's what makes the game interesting.
By using a 1d12, you have two things working against you. One, you have a dead number that you won't be using. And two, each other number will only show up 8.33% of the time (as much as 4's and 10's used to show up). And they all have the same chance, so strategy would be very different. I would think it would make it a whole new game.
First thing is, that it would make the numbers more or less unimportant, since they all have the same probability now (that of 4 or 10 before). This means good building spots are nothing you have to fight for anymore. Settlements will be more spread out from the beginning, because a starting settlement with a 2 or 12 is no problem anymore. So there will be more space for building.
Another effect of equal probabilities is, that you would not have important and unimportant settlements anymore. So it might take more time until buying your first city pays off.
More important on the long run is, that no resources can be scarce just because all of their fields have bad numbers in your variant. This probably makes all games the same, since differences in the island layout do not matter much anymore.
In the rules as you suggested them there will also be more cards in the game. You reduced the probability for the robber and (more importantly) for halving the cards from 1/6 to 1/12. This means having many cards is likely to go unpunished. Better use 1 and 7 for the robber events.
But if you really want to find out: Just try it! And maybe report the results here...
In my house, we only ever play with a d12. We combine it with a number of different house rules. I don't like sleeping on the couch, so my house doesn't use the robber. Instead on a 7, you simply grab any one card from the bank. You can use it for a robber if you want to though, for sure. But once we started with the d12, we never went back to 2d6.
On a 1, you get a development card. Why? Because we need something cool for the 1, and development cards didn't see as much play in our style. This also means that you get an extra card of some kind on exactly 1/6th of the rolls, which is the same chance of rolling a 7 with 2d6. What you do with 1 and 7 is up to you. I recommend experimenting, as there is a lot of design space available.
We always play with the 'roll again on doubles' house rule. (Or is it an official rule? I forget, they're all such a blur!) With a d12 there are no doubles, of course, so we use the 2 and 12 for doubles. This again makes for 1/6th of the rolls being doubles, the same with 2d6. This changes strategy slightly, as stacking on 2 and 12 can give you massive chain-rolls. Of course, you stand the same chance to have a drought for a while any time you stack numbers, so it balances it out.
The obvious big change is that there's less interest in 5,6,8,9. It allows you to balance your picks based on what you need, not just what the numbers are. It also reduces crowding around the 6 and 8 (and to a lesser extent 5 and 9, as expected). Because of the lower crowding, there is more room to grow uninterrupted. This leads to games with longer roads and faster expansions (made even faster by having equally good choices on whatever resources you need), and ultimately faster games. This wasn't planned when we started doing it, it just happened that way.
We've used the d12 several times, and I enjoy the change of pace. When using the d12, we've always treated the one as equivalent to the seven. That seemed like the most obvious choice. The idea of picking up a development card is intriguing.
We have the event card expansion that is intended to replace dice. It's goal is to keep the probability curve similar to the 2d6 game play while smoothing out streaks. Within a given shuffle you'll eventually hit all the numbers. It also adds earthquake events that just serve to extend game play. Overall when trying to simplify the game i prefer using the d12 to the event cards.
The strategy with the d12 becomes more about settling on the resource types you need for your preferred style of play and possibly competing for ports instead of just competing for the numbers that come up the most often. It greatly decreases the chance for one player to get into a situation where they're really not competing with everyone else. I don't think it diminishes the fun of the game at all.