King Flevance
Deity
I agree that too much rubber banding is bad game design as well.I think that too much rubber banding is also bad game design. If the game won't let you fall behind during the first 150 turns, those will not really matter. If the only thing that matters is playing the perfect last 50 turns, why play the first turns at all?
In my opinion, at every point in the game, you should be able to get an advantage. Depending on the game, this could be either by skilled moves or just sheer luck. Even if you are ahead, you should be able to increase your lead. It might just be harder to do so. It only becomes a snowballing problem when being ahead in the game increases the amount of advantage you are able to get, so you get exponential scaling.
But that also means that there will be a point in the game when victory is inevitable. Which is a price I am willing to pay for having the entire game matter.
I also agree that a game should offer many opportunities for players to gain an advantage, which is why snowballing is bad. If the game design allows one player to get a massive lead, then the other players have no opportunities to gain an advantage so the design fails to offer on its promise. In order for that premise to hold true then it means there are constant ways to gain a disadvantage and this does not hold true in that scenario. Unless your argument is that YOU should be able to gain an advantage at all times, not everyone. Either you allow a runaway leader OR you implement ways for all players to gain the opportunity for an advantage so the leader must fight to retain the lead. There are many ways to do this and "rubber banding" is too broad of a term beyond the discussion of the broad topic of "snowballing".
Specifically, the conversation was primarily in regards to crises as a "punishing for winning" mechanic. However, I see it more like friction or resistance for the leaders in specific areas. You recall the old lists in Civ "Pliney the Elder compiled a list of the largest civilizations" Now, with Crises these lists could have more relevance. If each age gave you a list of largest, wealthiest, etc. then when the crises hit, you would have a good idea of who would probably be impacted most. But I wouldn't let this list telegraph which crises was coming. Instead, show them all mid age.
This does not attack you for "winning". The largest empire could have a weak economy. The most advanced could be militaristicaly weak, and so on. Crises should be predetermined and select targets based on how well a civ is doing in that area. The plague should target players with the most population. Smaller nations should be able to stop trading with them to minimize risk of spread. It is possible that player might lose the lead for a bit, but they could also have a strong enough economy to hold out against it. Beyond this, we get into the details but the overall idea is simply to add resistance to any player(s) who would be dominating in a particular area of the game. As well, this is only 2 times per game.