Posted by ike1024 on 12/4/2011 11:31:00 PM (view original):

Posted by hackerhog on 12/4/2011 10:55:00 PM (view original):
No team should EVER play a member from their own conference in the NT before the Elite 8 ever!

Ever!

It's simply that simple.

What are you talking about?

It happens in real life.  They try not to do it, but it happens.  Marquette and Syracuse played in the second round of the NCAAs last year.

True, but that was a rule change made in response to the Big-ger East.

Since all of our conferences have 12 teams that shouldn't be as much of a problem.

Then again, if this new selection seems to be favoring SOS like you guys said, some of the top conferences might be sending 10+ teams to the tourney and you still might need to play conference mates early.

There's absolutely no way to prevent intra-conference matchups until the E8.  None.

Sounds like I'm in the minority here, but I do believe margin of victory should be included here -- not to the extent that it seems to be in the new formula, but to some extent. I went 22-7 this year in D2 Naismith. My avg margin was 9.9 points, vs a top-40 schedule.  My losses were by 19 (to bow2dacowz' juggernaut), 8, 1, 3, 5 in OT, 2 and 2. Six of those seven losses were on the road. Meanwhile, I won 9 gms vs top-100 teams by double digits. Maybe I'm a really bad gameplanner, maybe I just had weird luck, but the point is my wins were much more decisive than my losses.

I understand a win's a win and a loss is a loss, and whatever other cliche, but if you want to assess quality when picking / seeding teams for the NT, margin of victory (at least against decent teams) does say something valuable.

So I should be penalized for getting my bench in for the second half of a 20-point win?

I understand the point, but including mov discourages developing players or experimenting with lineups/strategies.  I think that's bad for the game.

Posted by ike1024 on 12/5/2011 12:03:00 AM (view original):
So I should be penalized for getting my bench in for the second half of a 20-point win?

I understand the point, but including mov discourages developing players or experimenting with lineups/strategies.  I think that's bad for the game.

That's pretty easy to solve. Put a cap on margin of victory - so a 40-point win counts the same as a 20-pt win. (Or 10, or 15, whatever.) But to say that a bunch of 3-pt wins say the same thing about the quality of your team as a bunch of 13-pt wins just doesn't ring true.

And as far as developing young players, sure, I do that too, but I understand that there's always been a tradeoff there - I'm taking the risk of weakening my overall results this year for future gains. I don't think moderately applying MOV really changes that, though maybe it moves the needle just a little in the direction of now vs later.

Sure, I understand the point.  But I'm not talking about a risk.  I've never once lost a game when I was trying something new or attempting to get more playing time for my young kids.  But I may have only won by 15 instead of 40.

A cap is fine, I guess, but it's still pretty arbitrary.  There are just a lot of random, unimportant events that can make a 10-point win a 16-point win, and vice-versa.  In real life, do I care that my guy missed the front end of a 1-and-1, the other team gets the rebound and hits a halfcourt three to end the game when I'm up 10?  Not really.  Does it mean my team is any worse or better than if I had hit the FT or they hadn't hit the three?  Not really.

All I'm saying is that any use of MOV has to be pretty minimal or you run the risk of placing too much importance on irrelevant late-game events.

Posted by kujayhawk on 12/4/2011 11:40:00 PM (view original):

Posted by ike1024 on 12/4/2011 11:31:00 PM (view original):

Posted by hackerhog on 12/4/2011 10:55:00 PM (view original):
No team should EVER play a member from their own conference in the NT before the Elite 8 ever!

Ever!

It's simply that simple.

What are you talking about?

It happens in real life.  They try not to do it, but it happens.  Marquette and Syracuse played in the second round of the NCAAs last year.

True, but that was a rule change made in response to the Big-ger East.

Since all of our conferences have 12 teams that shouldn't be as much of a problem.

Then again, if this new selection seems to be favoring SOS like you guys said, some of the top conferences might be sending 10+ teams to the tourney and you still might need to play conference mates early.

there are no 253 team conferences in HD like there is in real life Big East

Naismith put all Big 6 teams with the exception of one into the second round. Is this what will make the game better?

Posted by ike1024 on 12/5/2011 12:27:00 AM (view original):
Sure, I understand the point.  But I'm not talking about a risk.  I've never once lost a game when I was trying something new or attempting to get more playing time for my young kids.  But I may have only won by 15 instead of 40.

A cap is fine, I guess, but it's still pretty arbitrary.  There are just a lot of random, unimportant events that can make a 10-point win a 16-point win, and vice-versa.  In real life, do I care that my guy missed the front end of a 1-and-1, the other team gets the rebound and hits a halfcourt three to end the game when I'm up 10?  Not really.  Does it mean my team is any worse or better than if I had hit the FT or they hadn't hit the three?  Not really.

All I'm saying is that any use of MOV has to be pretty minimal or you run the risk of placing too much importance on irrelevant late-game events.

But those random events cut both ways, as you said. . so that randomness would tend to even out over the course of a season>  one 15 point win becomes a 14, sure.  Another 14 point win becomes a 15.

. 

Posted by ike1024 on 12/4/2011 11:42:00 PM (view original):
There's absolutely no way to prevent intra-conference matchups until the E8.  None.

The real-life selection committee has a hard-and-fast rule that no teams from the same conference are allowed to play one another earlier than in a regional final... unless a conference puts 9 or more teams in the tournament. If real-life does it, WIS should make every attempt to include it in their programming.

Another real-life rule is that the top three teams (based on S-curve position) from a given conference must be placed in different regions.

Also in real life, the committee has a guideline to be adhered to if possible, that regular season rematches are to be avoided in the 1st and 2nd rounds of the tournament.

WIS should make every attempt to include these rules and guidelines in their programming.

Posted by a_in_the_b on 12/4/2011 11:08:00 PM (view original):
. . you are using the BCS as an example to be FOLLOWED?

Yes, because this is a computer game, and the BCS is a computer system for determing rankings. Its logical to use it as a basis/starting point for a system in HD.

Posted by a_in_the_b on 12/5/2011 8:45:00 AM (view original):

Posted by ike1024 on 12/5/2011 12:27:00 AM (view original):
Sure, I understand the point.  But I'm not talking about a risk.  I've never once lost a game when I was trying something new or attempting to get more playing time for my young kids.  But I may have only won by 15 instead of 40.

A cap is fine, I guess, but it's still pretty arbitrary.  There are just a lot of random, unimportant events that can make a 10-point win a 16-point win, and vice-versa.  In real life, do I care that my guy missed the front end of a 1-and-1, the other team gets the rebound and hits a halfcourt three to end the game when I'm up 10?  Not really.  Does it mean my team is any worse or better than if I had hit the FT or they hadn't hit the three?  Not really.

All I'm saying is that any use of MOV has to be pretty minimal or you run the risk of placing too much importance on irrelevant late-game events.

But those random events cut both ways, as you said. . so that randomness would tend to even out over the course of a season>  one 15 point win becomes a 14, sure.  Another 14 point win becomes a 15.

. 

They might, they might not, but it's pretty small sample size to assert that they will. 

And if they don't, which I would bet happens plenty often, then seeding is affected by a factor lacking true substance.  It would be one thing if you have a seclection committee that can actually look at the games and make a judgment on whether the game was closer than the final score indicated, but it's a completely different problem when it's a computer-programmed bright-line rule.

Posted by professor17 on 12/5/2011 8:53:00 AM (view original):

Posted by ike1024 on 12/4/2011 11:42:00 PM (view original):
There's absolutely no way to prevent intra-conference matchups until the E8.  None.

The real-life selection committee has a hard-and-fast rule that no teams from the same conference are allowed to play one another earlier than in a regional final... unless a conference puts 9 or more teams in the tournament. If real-life does it, WIS should make every attempt to include it in their programming.

Another real-life rule is that the top three teams (based on S-curve position) from a given conference must be placed in different regions.

Also in real life, the committee has a guideline to be adhered to if possible, that regular season rematches are to be avoided in the 1st and 2nd rounds of the tournament.

WIS should make every attempt to include these rules and guidelines in their programming.

Without going in and changing it by hand, my guess is that it would be almost impossible to program something like that, especially when there will need to be exceptions for the numerous super-conferences that exist in HD.

Posted by bow2dacowz on 12/4/2011 8:30:00 AM (view original):
in naismith D2 i have the top overall seed in the tournament and have a potential second round matchup with the 8 (17rpi)/9(22rpi) winner...nice reward for having the best season...

Not saying that the new system is necessarily any better, but that sort of stuff happened under the old system, too. In Phelan a few seasons back, my Michiigan St. team had a 1-seed, and we played a 12 or 13 RPI, 800+ overall UNC team that was the 1st or 2nd most talented team in the country in the 2nd round. We lost, and UNC won the national title.

Posted by bow2dacowz on 12/4/2011 2:49:00 PM (view original):
Theres a reason margin of victory is not a factor in real life beyond the impact it may have on the voter polls which do factor in. the transative property does not work in sports.

Transitive properties don't work when looking at single game results or comparing one team to another based strictly on common opponents, because the sample size is so tiny. You can't conclude anything whatsoever based on one game result. But when you're dealing with the game results of 4500 games over the course of an entire season, when all the teams are fully linked, then absolutely using margin of victory is a more accurate method of evaluating team strength than not. The key is making sure the person who's setting up the model knows what they're doing, and properly accounts for diminishing returns with increasing margins of victory, correcting for differences in tempo (i.e. using game score ratios instead of absolute margins of victory), and ensuring that teams don't get undue credit for *playing* a tough schedule as opposed to actually *winning* games against tough opponents.