09 Rules: Best Sim Results or Any Given Sim

[Goodell]

I can run test games over and over to see how things would go MOST of the time, but that's not always what scenario is generated for a specific game. Once I push record and then run the game, that's it. I can probably alter the game simulator to run a matchup 3 or 5 times and then pick the BEST result that most closely matches what SHOULD happen. That would limit upsets somewhat (less so with 3 versus 5) but ensures stronger game results more in line with the best individual stats expected.

I think many want to preserve the upset option, but also get ticked when their team gets the shaft on an odd sim result. In the end, I got to consider my accountability in it too to put out the best results and sometimes that's difficult with just 1 game short samples. Do a 1-game sample in Diamond Mind for baseball (or any sim) and sometimes a great team will lose unexpectedly to a bad one. Sometimes a great hitter will strike out 4 times. In football those are exaggerated with only 1 game a week and 1 game series to decide playoffs.

For each game there are some possibility for either team to win usually. Even if one should win 90% of the time, no guarantee that one of the 10%-ers will pop up when it counts. Upsets happen in reality too. For each sim player, if run a matchup 100 times there would be a range of outcomes and sometimes the game that counts comes up on the lower or higher end. Sometimes those 1-time odd games get over-emphasized in meaning when if the same game run again something very different usually happens.

It may not be something I have time to get done for next year, but looking for some general vibes toward getting the BEST results versus keeping things somewhat unpredictable with just the 1-game roll of the dice. And if most want to keep just the roll of the dice, then more understanding about sometimes odd results come out of sims with just one trial. If we want less odd results, we could try to have the sim get the best results for more satisfying conclusions possibly.

That might be something tied into overall sim improvement for the long off-time after the draft into the summer months more than now, and even if we did that it would be something totally transparent where you'd see no difference in the game results but know that the game simulator ran that matchup a couple times and picked out the best one in terms of both overall winner (whoever won the 3- or 5-game internal series in which there could be an upset too but less than 1-game trial) and the best individual game stats.

[Ben C.]

I actually think there should be more upsets in these leagues. How many times in the real NFL have we seen the better team on paper lose? Think Giants-Patriots in Super Bowl, Packers-Falcons in Wild Card one year (first home playoff loss at Lambeau Field), Falcons-Vikings in the 98 or 99 NFC Championship... these things happen a LOT more than you would think and it's part of what makes sports so interesting.

We need to preserve the underdog's chances if we want the game to remain interesting long-term.

[sportznut]

One game and one game only.

Well, except if I should lose.

Then re-sim until you get it right.

[Goodell]

We will work on the game simulator in the summer months and attempt to get better results without creating a series of games and picking the best outcome of those, but keep in mind that there may always be some strange things with 1-time rolls of the dice even if great improvements made in simulator generally.