A mistake in design

Last night I went to Snakes and Lattes' Game Designers Night again.  I brought Dot to test and was really pleased with the results:  Everyone who played it enjoyed it and thought the game felt polished and complete though there were a few small suggestions for minor tweaks.  This is a lot better than last time where FMB was played and the reception was not so hot.  I got to see a lot of other people's games there too and I was stunned (again) by how much production and economics goes on in games where the fundamental mechanics aren't even sorted out.  When I don't know what is going on with my games the furthest I get is paper, plasticine and beads but the other folks there had figurines carved by goldsmiths, ingraved metal coins and all kinds of other interesting bits for games where the play style was still totally up in the air.  Lots of comments were made about the feel of the pieces and the boards, on the importance of flexible cards vs. solid and other such details that I naturally overlook.

The biggest example was the last game I played on the night which had the hand carved pieces and for which the creator already had manufacturers and distributors lined up; publishing was set to commence in a month.  Despite this there was a major rules revision just prior to the visit to Snakes and Lattes and the new rules had never been tested before.  I played the game once and at the end I was quite certain that there was a very important flaw:  The game was scored over five rounds and the winner of the last round was nearly guaranteed victory.  This was borne out in our test game where the person in last who had done nothing of note throughout the entire game won the last round by getting a single lucky die roll and consequently defeated everyone else.  I can't imagine this mechanic working out well in the long run since the game is obviously intended to have a lot of long term strategy and what decisions you make seem like they *should* matter but they do not.  I am quite confident in a four player game you could use the algorithm of "Never do anything, if forced to take an action choose randomly" until the fifth round and maintain a solid 24% win rate.  Your decisions in the fifth round are also mostly irrelevant too, though on the last couple rolls there is a "whoever gets the die roll they need and flips over their card first" mechanic to decide the victor in which planning and reflexes do matter.

I just can't fathom having a game in that state and being ready to ship to a publisher.  I can much more readily imagine having a game that has been tested and tweaked and smoothed to near perfection years ago and never publishing it at all which may well show that I am the fool in this scenario since nothing of mine is going to hit a publisher at the rate I am going.  For many people the art and flavour of the pieces, pictures and cards is critical while for me the important part of the game is the mathematics.  I am perfectly content playing with scrap paper and cardboard if the game is awesome and completely uninterested in a beautiful game with a great feel if the gameplay is weak.  The regular gamer is probably somewhere between the two extremes.  Gameplay matters and presentation matters too; if you have neither you get nowhere and if you want a smash hit you must have both.