Where's Mike when you need him?
I am everywhere and in all things, my son. Open your heart to the sky, and I am there.
Its not about the winning ratio, its about if a game MAKES SENSE and is FUN TO PLAY.
"Fun" is a subjective measure, and when you're creating a game you don't try to guess what people will find fun because it's so varied as to be futile. Instead you ask yourself "what will the vast majority of people NOT find fun" and try to eliminate that from your design. Most people don't enjoy playing games they're destined to lose... in fact, most people don't enjoy playing games in which they are genuinely aware that the deck is stacked against them as compared to other players. They'll play games in which their chance of losing is better than their chance of winning if that's true for all the other players too
, such as Blackjack, or the Lottery... but very few people are interesting in volunteering to be the punching bag for other human players based solely on seniority.
So, what MAKES SENSE is to create as even matches as possible in any matchmaking environment... even in terms of likelihood of victory. Why? Because A) we know inducements don't bridge the gap and B) we know most people prefer to test their personal skill against the other coach's personal skill rather than be hammered down by the dice, which is what happens if the other guy has a much better developed team. Some people don't mind, but those people also don't mind playing even games. There are certainly some folks who want to play games in which they are just trashing someone who has very little chance of victory, but very few people are interested in being that trashed team. So, we apply the rawlsian veil of ignorance and say that what will make people happiest, in general, is as even matches as possible. For Cyanide and FUMBBL, the best variable they had for that was TV at the time their MM environments were designed.
Statistical analysis have suggested some superior alternatives, but "just randomly match any two people together and apply inducements by TV difference" is decidedly worse than TV for this purpose.
THAT is the spirit of these rules, assuming that you play a variety of games, sometimes on the giving and sometimes on the receiving end of the inducement system.
The average BB team plays around 5 total (median). They don't play a variety of games. The average TV is in the 1200-1300 range, meaning that, on average, teams will most likely be the underdog in the first 2-3 matches played under the system you propose. In your mind that probably suggests they'll want to play even more games to try to make up the difference, after suffering unfair matches. In reality, the 5 game median would just go down because the coach would be having less fun by being the bitch in the initial matches.
I mean look at it this way: in 33% of all Blood Bowl matches you can expect an upset. Upsets are great, and that's because they are not the usual thing to happen. I mean everyone likes Timo Boll to beat the crap out of Ma Long, right? Because he's not supposed to, but we know he can if things go his way. That's great sports and great entertainment.
So entertaining that we should expect to see heavy underdogs challenging upward in Ranked or Auld right? Turns out people aren't really that entertained by having a worse chance of winning than their opponent, and tend not to choose to be in that position. They'll suffer through it in scheduled league settings, but there doesn't seem to be a ton of interest in voluntarily being in that position.