...
Stuart referred to RROP S3-Irreg #5,6. These rules certainly suggest that the chips could be carried over in some situations. On the other hand, #4 (fouled deck) and #5 (".. the right to a refund..") also clearly suggests that a "refund" of bets in other situations is possible. And both situations, it turns out, are quite different from the scenario that we are talking about.
I understand the "freeroll" situation described in RROP #5 to refer to the act of an individual, knowing that the deck is fouled, attempting to win the pot by forcing others to fold. Then, in the event of a failed attempt, he points out that the deck is fouled hoping for a refund. That is the "freeroll" attempt. This rule is not about collusion. This is about an individual using private knowledge to try to get two chances to win the pot, and the chips forfeited act essentially as a penalty for that unethical behavior. RROP provides that the forfeited chips may subsequently be used to "juice" the pot for the next hand.
My problem with the carryover method generally, is that it can create its own unique strange incentives. In a tournament setting, with a carried-over pot of any reasonably large size, you can guarantee that changing the effective pot odds for any decision made in that next hand is going to affect the natural course of how that hand is played out, and this affects the tournament as a whole. For example, the carried-over part can be so significant, that it can make it correct for one or more players to move all-in to try to take it down, when they might not have otherwise done so. Moreover, the free 'dead money' could potentially give a player (e.g. one that is now all-in in the blind) to now get a huge windfall.
In this regard, I don't see how it could ever be fair -- in a tournament setting -- to potentially allow a player to win more chips from a pot than (his contribution * the number of active players in the pot + antes/blinds). In a cash game, it is fine if one player happens to get a "bonus" from a juiced pot -- the extra cash, however obtained (from the house or another player) goes in his pocket. But in tournaments, IMO, relative chip balances always matter and carried-over pots can throw off that relative balance. For example, if the blinds are 500/1000, and the Big Blind is all in for 200, but there is a carried-over portion from the last hand of 20,000 in the center of the table, how can it possibly be right that the Big Blind could win and come out with more than 2K in chips at a ten-handed table, let alone 20K+? This certainly is going to be unfair to other short stacks left in the tournament at other tables.
Also, if you apply the carryover method, what happens if the players that were eligible for the side pot were all-in in that last hand? Is it right to eliminate all three of them from the tournament because they are left with no chips at the end of the hand, even if we know that at least one of them would have survived had the dealer kept the hands out of the muck? Or do we deem them still in despite having no chips from the last hand, and deal them in as if they were all-in for antes? A possible solution, but highly unconventional.
And finally, if we are OK with allowing "forfeited" chips to carry over, why wouldn't we extend this concept to situations where a player goes all-in (remember that chips bet in turn are supposed to stay in the pot), but his hand gets accidentally mucked by the dealer? Isn't the existing rule that says we return the 'uncalled portions' of the bets inconsistent with the carry-over approach? If the carry-over approach was ideal for tournaments, why don't we then just eliminate the player, and carry-over the uncalled portion in the pot to the next hand instead? (Maybe we should!)
I don't expect all of us to agree, and I would respect Stuart's decision if I was playing in his venue. From a player's perspective, it is my fault for not tabling my hand, so I'd be prepared to take my lumps, whatever they may be. But from the POV of someone trying to keep things fair to all players not just at the table in question but in the tournament as a whole, while I am fine with the carryover method for cash games, it does not make sense for me in tournaments. The carried-over amount referred to in RROP #5 and #6 is basically a penalty that can go to the winner of the next hand as a "bonus"... in a cash game, this is fine. In tournaments, IMO this can be problematic especially when the carry-over amount is large. I would rather take the chips out of play than carry-over chips, as that also solves the chip dumping and collusion concerns. We do take chips out of play when players are disqualified, so this wouldn't be completely foreign. But I also can't criticize the Original Poster's approach of splitting the pot, effectively allowing the players eligible for the side pot to play the board.
With respect, I think the "ganging up" scenario is unrealistic in the OP's particular scenario, because the situation is not completely controllable by the players -- it requires a total procedural failure on the dealer's part to not force the all-in hands to be shown. The participants in the "ring", to succeed, would need to be certain that they could collectively get around the fact that the dealer/floor will insist on their hands being turned face-up, and they need to be certain that the Dealer will not intervene by physically preventing at least one hand from getting mixed into the muck. If somehow they manage to avoid the dealer from intervening (and if the dealer is in on the gig, you have bigger problems), practically they are not going to be able to get away with this 'play' looking accidental more than once, and if they do try, then the case for disqualification for collusion is clear.
I guess this is my usual long-winded way (and I apologize for that) of saying -- I think we need to debate this further, and perhaps take a poll. Ultimately, I am willing to defer to the majority on this one.