Posted by steelers821 on 1/19/2024 8:37:00 AM (view original):
This passes the eye test for me I definitely had some really bad divisions. Just ****** me off more when I’m .569 in division and .585 out…

thanks Rob I appreciate all the extra work you put in.

Posted by robusk on 1/19/2024 10:26:00 AM (view original):
My method for accounting for SoS is to assign a replacement level score to each division based on that division's performance, so that a division that kicks *** against other divisions gets a higher score than another division, a conference that kicks *** against the other conference gets a higher score than the other conference. I then calculate how tough each division/conference would have had been in summation if I replaced the respective coach in each of their divisions with their replacement level counterpart. I retabulate the overall totals and sum it all up. After I get the totals I break them down into division averages to make them more digestible.

I also calculate the raw opponent winning percentage of each coach's opponents as well as their raw sums of division/conf scores but both of those are made noisy by their own performance. After 13 seasons the current method is by far the best I have seen and in going back to past seasons it has always passed the eye test. This is superior to other methods like replacing them with a .500 team because that creates a lot of distortion in the numbers that I won't go so far into the weeds on.

Almost always the hard majority of the teams are within a standard deviation of the mean. With that said, I will post the findings below. The second column represents how many games under .500 the adjusted average division for each coach is. For this particular season every coach is within a standard deviation of the mean totals except for the top 3 and the bottom 3. I have evolved the model over time so I don't have all the historical numbers handy to post at the moment. As far as I can tell, the easiest draw this time around is the easiest draw in Savage history (I have no record of a team two standard deviations away besides this one). The hardest draw this time around is one of the hardest ever but there are a few draws that have been slightly harder.

Posted by dh555 on 1/19/2024 6:52:00 PM (view original):
im not sure if any of my teams will win a playoff series but fuuuuucccckkk yea! 5 team bonus! sweeter than a savage world title imo

Posted by jcred5 on 1/20/2024 11:16:00 AM (view original):

Posted by dh555 on 1/19/2024 6:52:00 PM (view original):
im not sure if any of my teams will win a playoff series but fuuuuucccckkk yea! 5 team bonus! sweeter than a savage world title imo

funny you mention that as I was thinking about that recently. 5 in is 20 pts as is a championship. I've had both, personally I think a championship is harder and I would be curious how owners might assemble teams differently if a championship were say 40 points. Probably it would result in more variance as more owners putting together super teams, which means more worse teams and more random benefit to those who get those worse teams in their division. Would be interesting though.

Should we also be discussing any evolution of Savage? I know commish has thrown a couple things out in the past. Just wondering if we think this will stay as is, like ODL I suppose, and the variance comes once a year from new players.

Posted by edsortails on 1/20/2024 4:11:00 PM (view original):
Wait, is getting 5 teams in the playoffs supposed to be difficult?

Posted by steelers821 on 1/20/2024 4:30:00 PM (view original):

Posted by edsortails on 1/20/2024 4:11:00 PM (view original):
Wait, is getting 5 teams in the playoffs supposed to be difficult?

Well the last three champs all had it and the last two won because of it. I’m not a big fan personally. seems like a yardage bonus in fantasy football. If someone’s 5th team makes the playoffs with 39 wins and another teams 5th best gets 44 and misses out who really was more successful at building 5 competitive teams. I assume that’s the purpose of the 5 team bonus, to make it viable to win a title with a balanced team approach.

imo the bonus should be for 5 teams above .500

Posted by 24kpyrite on 1/20/2024 4:33:00 PM (view original):

Posted by steelers821 on 1/20/2024 4:30:00 PM (view original):

Posted by edsortails on 1/20/2024 4:11:00 PM (view original):
Wait, is getting 5 teams in the playoffs supposed to be difficult?

Well the last three champs all had it and the last two won because of it. I’m not a big fan personally. seems like a yardage bonus in fantasy football. If someone’s 5th team makes the playoffs with 39 wins and another teams 5th best gets 44 and misses out who really was more successful at building 5 competitive teams. I assume that’s the purpose of the 5 team bonus, to make it viable to win a title with a balanced team approach.

imo the bonus should be for 5 teams above .500

I think your 39 and 44 argument is the same as 5 teams above .500. It's all about luck of the draw with placement.