Posted by mullycj on 1/29/2021 11:32:00 AM (view original):
"Also of note, mully’s opponent was #39 in the PR, so really should have been a 10 seed. And the winner is up against the lowest #1 seed (a very overrated one that is due for a second round exit), "

Don't you hate being right most of the time?

Yeah, they beat us in exhibition, too. I knew what was coming. Not like we were this loaded team, we were preseason #15, and I thought that was too high, with one real senior, and fully half the roster is role player/insurance caliber. I’m not nearly as disappointed this season as the last two, to be honest, it was a smoke and mirror type season.

Posted by shoe3 on 1/29/2021 1:03:00 PM (view original):

Posted by mullycj on 1/29/2021 11:32:00 AM (view original):
"Also of note, mully’s opponent was #39 in the PR, so really should have been a 10 seed. And the winner is up against the lowest #1 seed (a very overrated one that is due for a second round exit), "

Don't you hate being right most of the time?

Yeah, they beat us in exhibition, too. I knew what was coming. Not like we were this loaded team, we were preseason #15, and I thought that was too high, with one real senior, and fully half the roster is role player/insurance caliber. I’m not nearly as disappointed this season as the last two, to be honest, it was a smoke and mirror type season.

How much exactly is "fully half"?

6 of 12. Fully half, you know, as opposed to nearly half (5 of 12) or a third (4 of 12) or a quarter (3 of 12).

Posted by mullycj on 1/28/2021 3:49:00 PM (view original):
I agree cub.

But what are your thought on the system stretching the adjustment all the way out to the 3rd round?

unrelated to the debate about fairness, just wanted to mention the game actually goes further than 3 rounds even. the game will always place the 4 highest seeds from a conf in different brackets, if it is possible to do so without displacing and already-placed team. it will do the same for the 8 highest teams in the conf, and the semi-region or whatever, the elite 8 sized bracket. this goes beyond real life significantly, which i think has changed over the decades but last i looked, i think it would do the top 3 teams from a conf in different brackets always, and never let conf teams meet in the first 2 rounds unless necessary.

in case anyone is wondering, all the mystery referenced about seeding is actually super straight forward logic wise, its just not readily apparent (to put it lightly). the sim engine takes the S-curve, which is exactly the projection report, and takes the top team off the list, and places them. the way it places them is, it tries to place them in the highest available spot, with a conference constraint of the whole region (meaning if there is a conf in the region, the school cannot be placed). then it tries the next highest available spot. if this fails (can't place in any available spot), then it tries again from the top with a half-region constraint, and then a quarter-region constraint (and perhaps so on).

Posted by mullycj on 1/28/2021 10:37:00 AM (view original):
I recently got fuked over in the NT with a #8 seed while at 27 on the projection report. The ONLY thing I can assume is the engine wanted to avoid a sweet 16 matchup vs a conf mate. (The team that got the #7 seed got a SIM 1st round instead).

Personally , I say screw the drop in seed for a POTENTIAL game 2 rounds out and place me where I belong. I can see doing this for 1st round and maybe potential 2nd round matchups, but that's it. Others say 1 seed spot is no big deal. Maybe so. But it could mean a difference in next season's prestige which affects recruiting which then snowballs into something that IS meaningful.

Agree/disagree?

So .. 4x8=32 and 32-3=29.

So theoretically, an 8 seed would be positions 29-32. Getting an 8 seed at 27 is not out of the realm of possibility for just about any reason. If you were at 24 and got an 8 seed, that would be getting fuked over :)

Posted by mlitney on 1/29/2021 12:39:00 PM (view original):
I'm glad I don't have to write the game code that formulates the NT seeds. Although I don't know much about programming, I'd imagine it to be a difficult task.

Former low-level coder here. I don't think it would be that tough. It just goes by straight logic, a bunch of if/thens.

We have all the conference tourney winners, then the remaining teams straight from the projection report. We rank them 1-64 going by order of projection report, might throw in some weight for conference tourney winners/last 10, wins against top 100, top 50 and top 25, etc (I'm not sure how much projection report weighs these items), then I get to seeding.

There are 16 blocks of 4 - everyone hopefully fits into their appropriate block of 4, but we add the steps of avoiding conference matchups as far out as possible, but I would not move any team more than 1 seed away from their ideal seeding so a different team could avoid a conference member.

I could probably do it in a day and I was far from great during my PHP/PERL days. I'm sure there are complexities I'm not thinking of at the moment.

Posted by vegaskevin on 2/2/2021 10:57:00 AM (view original):

Posted by mlitney on 1/29/2021 12:39:00 PM (view original):
I'm glad I don't have to write the game code that formulates the NT seeds. Although I don't know much about programming, I'd imagine it to be a difficult task.

Former low-level coder here. I don't think it would be that tough. It just goes by straight logic, a bunch of if/thens.

We have all the conference tourney winners, then the remaining teams straight from the projection report. We rank them 1-64 going by order of projection report, might throw in some weight for conference tourney winners/last 10, wins against top 100, top 50 and top 25, etc (I'm not sure how much projection report weighs these items), then I get to seeding.

There are 16 blocks of 4 - everyone hopefully fits into their appropriate block of 4, but we add the steps of avoiding conference matchups as far out as possible, but I would not move any team more than 1 seed away from their ideal seeding so a different team could avoid a conference member.

I could probably do it in a day and I was far from great during my PHP/PERL days. I'm sure there are complexities I'm not thinking of at the moment.

The projection report is the exact S curve. It has already controlled for everything.

Let’s scratch the S-curve and draw for spots in the tournament like tennis does.

The top 16 teams are seeded 1-16 in designated spots on the bracket. Teams 17-64 are unseeded; a random draw places them anywhere on the bracket.

Would definitely make for an interesting tournament!

Posted by gillispie1 on 1/29/2021 9:44:00 PM (view original):

Posted by mullycj on 1/28/2021 3:49:00 PM (view original):
I agree cub.

But what are your thought on the system stretching the adjustment all the way out to the 3rd round?

unrelated to the debate about fairness, just wanted to mention the game actually goes further than 3 rounds even. the game will always place the 4 highest seeds from a conf in different brackets, if it is possible to do so without displacing and already-placed team. it will do the same for the 8 highest teams in the conf, and the semi-region or whatever, the elite 8 sized bracket. this goes beyond real life significantly, which i think has changed over the decades but last i looked, i think it would do the top 3 teams from a conf in different brackets always, and never let conf teams meet in the first 2 rounds unless necessary.

in case anyone is wondering, all the mystery referenced about seeding is actually super straight forward logic wise, its just not readily apparent (to put it lightly). the sim engine takes the S-curve, which is exactly the projection report, and takes the top team off the list, and places them. the way it places them is, it tries to place them in the highest available spot, with a conference constraint of the whole region (meaning if there is a conf in the region, the school cannot be placed). then it tries the next highest available spot. if this fails (can't place in any available spot), then it tries again from the top with a half-region constraint, and then a quarter-region constraint (and perhaps so on).

Alright Gil, but what happens in the unusual event that a conference gets 9 teams in the tourny? Seeding logic goes out the window?

Ha. Devil’s advocate here.