Hi Brian: Here's a link to another thread on a similar topic: http://www.pokertda.com/forum/index.php?topic=227.0
Personally I look at each of these situations from the point of view of "least damage" to the game (i.e. in the best interest of the game). As TD I don't want to have to enforce a statement that is not complete as to the intended action. Nor do I like to put the action of Player B (the out of turn declarant), in the hands of Player A in the case of "I'll do whatever you do". On the other hand, some of these statements are unambiguous and complete (ex "I'm all-in regardless of what you do").
They are however out of turn (which we don't want), and so what do we do with the rule that allows for releasing an O-O-T player if the action changes (TDA Rule 29)? This rule clearly sets up the possibility that the player will not be bound if the action changes, so does a player have the right to elect to terminate a rule when they see fit by declaring "I'm all in no matter what you do"? On the other hand rule 29 by one interpretation makes O-O-T declarations absolutely binding if action doesn't change but it may give the TD latitude in deciding when to release a player from their declaration in that it doesn't appear to absolutely guarantee the right to release if action changes, just to absolutely bind them if the action doesn't change... Sometimes it just doesn't seem to be in the best interest of the game to allow a player to so completely and deliberately influence another players thinking and then be able to weasle out of it...
So I think as the rules currently stand you have to interpret each of these conditional out-of-turn situations on a case-by-case basis, looking for what ruling is in the best interest of the game, or perhaps more accurately is least damaging to the game. And always warn or penalize these players enough that they won't repeat it. This all said, as to your specific examples: Example 1 I'd most likely treat as trash talk because it's so incomplete... Player B doesn't say what he'll do if Player A bets.... Example 2 I'd probably hold the player to a raise, or a bet if Player A checks. Example 3 I'd probably hold him to the all-in. Example 4 I'd probably hold him to calling Player A's bet... or to checking if Player A checks. I don't really like any of these rulings, but they are the least damaging to the game under the circumstances, IMO.
These rulings all presume the players are adjacent. Then you have the issue of a player who makes a conditional statement way out of turn (which was one of Dave's questions on the other thread)... with numerous players to act before him, it's just very damaging to the game... Example 4 doesn't make any sense in that case, how can player in seat 7 "pre-commit" to call any bet by seat 3 when there's 3 seats to act between them? What if one of those seats raises?... is he released because the action changed or is he held to calling 3 but can then fold ?? ... it just becomes nonsense. Keep in mind these are DELIBERATE declarations, not innocent mistakes out of turn when the player didn't recognize there are players before him... Example 2 is also iffy if there are seats to act between the players... Even example 3, despite unambiguous action still sets up a situation where we would have a player all-in to the left of several seats yet to act, and that's just not how we normally want the betting action to proceed... . so every one of these under current rules have to be considered individually IMO, as to whatever is best for the game at the time. The key is prevention by discouraging these statements in the first place. Fortunately, I don't see incomplete conditional statements very often, probably because most players recognize the jeopardy they put themselves in when they start making such rambling declarations.
Looks like this issue is on several people's minds so perhaps it's a topic for membership review at a future Summit.