Warrior General Games

Originally posted by Corndog
Except I asked Bort, again, tonight, and he said that he didn't do anything special with them and they used whatever the default ELO was in the database. He said he thought about making adjustments but never did.

Yeah unfortunately this definitely wasn't the case. For proof just look at Iowa Rogues at the start of S2. I think they had lost a grand total of 2 games and were outside of the top 20 and were behind teams with fewer wins then them let alone win differential. You only have to look at the records of mid tier teams with equal numbers of wins and losses and it's clear that teams from the start of S1 are ranked higher than those in mid season S1.

It's not particularly important now however it is interesting that this wasn't intentional!

Originally posted by Corndog
Except I asked Bort, again, tonight, and he said that he didn't do anything special with them and they used whatever the default ELO was in the database. He said he thought about making adjustments but never did.

Well, I can't tell you what value they started with, because you won't publish ratings. All I can tell you is that, absolutely, for certain, every S1 midseason team at creation time started tied at the bottom of the rankings and had to work their way up - including below teams that had losing records or even lost every game. Given that you are using a stock elo rating formula, this means one of two things.

1) They started with a lower rating than the teams created on S1D1 started with
2) Your elo calculations are not working right

It has to be one or the other. The claim that the elo is working fine and the S1 midseason teams started with the same rating that the S1D1 teams started with isn't possible.

Maybe the "Default ELO in the database" is corrupt somehow, or loading teams with the "default elo" isn't working right?

Are we supposed to play the semi-finals best of three?

That might be a bit ambitious given that we are getting close to the end of the season?

?