How do you account for WHEN you play a team? For example, no one would argue that the MSU team we beat by 30 points is the same team they are today. The committee used to give weight to the last 10 games, and that always made sense to me, at least as a tie-breaker when seeding two otherwise equal teams.
3/8 BACATOLOGY: NCAA TOURNAMENT ANALYSIS
- Thread starter bac2therac
- Start date
Body of work
No recency bias
Now your profile will look better if you finish strong and conversely look worse when you struggle at the end
No recency bias
Now your profile will look better if you finish strong and conversely look worse when you struggle at the end
BYU's performance last night will improve their seed line. I think they are at least a 7 but 6 coul be a possibility depending on results this week
I thought conference vs. non-conference was irrelevant? Isnt it overall body of work? That’s Boise’s only win over a sure tourney team.
you think? It’s still their 3rd double digit loss to the Zags and the SDSU win is their only win over a sure tourney team. They have a good record and the WCC is an underrated conference but 6 seems very generous.
The competition around the 6-7 spot isnt all that hot...see Clemson, Wisconsin and Oklahoma futility of late
I’m not sure why we are stressing about our NCAA seeding.
Nobody wants to be in the 8-9 game, but what difference does any other seed make for RU?
Imo playing well against Indiana and playing well in the 1st round of the tournament is far more important.
As a long time RU fan maybe my sights are set too low? Not seeing us as a Sweet 16 candidate this season.
Nobody wants to be in the 8-9 game, but what difference does any other seed make for RU?
Imo playing well against Indiana and playing well in the 1st round of the tournament is far more important.
As a long time RU fan maybe my sights are set too low? Not seeing us as a Sweet 16 candidate this season.
This is another really bad criteria IMO. Why should it matter whether a game is in or out of conference?
I agree. If we get to play a 1-seed in the second round, that means two things: 1) We won a game in the NCAA tournament, and 2) We get to test ourselves on a national stage against one of the top 4 teams in the country.
By the way, if we happen to beat a number 1-seed then we effectively take over their seed, which is nice.
Put a blanket around 6 thru 11. Sit anywhere except the middle and it is all the same
The system the NCAA tournament uses is incoherent IMO. Any coherent rating system needs to have some properties and among them is what I will call, for lack of a better term, symmetry.
Focus on road records is incoherent; in a coherent rating system there must be symmetry between road wins and home losses; i.e. for a team like Rutgers last year the lack of road wins is balanced by the lack of home losses. Home/road splits should be irrelevant.
Focus on good wins is incoherent; in a coherent rating system there must be symmetry between good wins and bad losses. In a coherent rating system (margin of victory aside) you should be able to deduce a team’s rating based on the combination of a schedule and a record with no other info. If I give you a list of 30 games and tell you a team went 20-10 against it, which 20 that was should be irrelevant.
Conference vs non-conference should absolutely be irrelevant.
The entire system is founded on mental gymnastics.
Focus on road records is incoherent; in a coherent rating system there must be symmetry between road wins and home losses; i.e. for a team like Rutgers last year the lack of road wins is balanced by the lack of home losses. Home/road splits should be irrelevant.
Focus on good wins is incoherent; in a coherent rating system there must be symmetry between good wins and bad losses. In a coherent rating system (margin of victory aside) you should be able to deduce a team’s rating based on the combination of a schedule and a record with no other info. If I give you a list of 30 games and tell you a team went 20-10 against it, which 20 that was should be irrelevant.
Conference vs non-conference should absolutely be irrelevant.
The entire system is founded on mental gymnastics.
Wait is this really true? It doesn’t make any sense at all. Just as conference standings don’t matter, and the timing of when a game was played doesn’t matter, I always thought it was about overall body of work.
Thanks. I'm doing well. Looking forward to a meaningful post-season, after having it taken away from us last year. See you at the RAC next year.
did Xavier lead by like 20 at one point. Damaging loss, they have lost to too many of this drek from the Big East. Honestly I think they are getting by on name alone. I had them out coming into today and there they will stay.
meanwhile Duke looks like they are going to stay alive for now but need 2 more wins
Louisville should start worrying. They almost assuredly are heading to the first four games now
meanwhile Duke looks like they are going to stay alive for now but need 2 more wins
Louisville should start worrying. They almost assuredly are heading to the first four games now
winning on the road is important because you are going into a hostile atmosphere. Yes to have them is a big plus....RU last year only had one basically heading into the last week..that was very poor considering they lose to drek like Pitt and St Bonnies
NCAA tournament games are not played on your home court. The committee wants to see that you can win away from home. Makes total sense
Last edited:
I would drive there myself to pick him up if so. But Pike would probably already be there.
It’s important to win games against at large caliber teams - preferably ones that end up in the top half of the bracket. For smaller conference teams, that almost always means winning big games OOC. For major conference teams, a soft OOC slate could be an issue (V-Tech teams of the past that ended up with weak unbalanced conference schedules and no wins over the better ACC teams) or it could turn out fine to do poorly OOC when you turn things around in conference against good teams (see this year’s Purdue). The common denominator is having good wins on your resume and a decent overall SOS. Individually, I don’t think Purdue would be better served to replace the Clemson loss with an OSU loss. Would’ve been helpful for the conference as a blended whole but not when thinking about Purdue’s resume relative to another team from a different conference individually.
Why?
I mean are we just talking about having a minimum # of games? (I agree that playing a schedule like we did this season should be frowned upon in a normal year) Or are we talking about putting some sort of special weight on having good wins out of conference specifically? If it’s the latter I don’t get it at all.
I disagree with this for a few reasons:
#1 - NCAA tournament games are not home games, true, but they are not road games either. If the logic is that they want to see how teams perform in settings similar to the NCAA tournament they should be looking for neutral wins specifically.
#2 - You are severely straining an already small sample size of team performance by attempting to split it into subsets. The signal / noise ratio in home / away record splits is very low.
#3 - Even if I agree that road wins are predictive of tournament success, I don’t actually think the committee should be trying to predict tournament success. They should be selecting and seeding teams simply on the strength of the teams’ results looking backwards.
Most important
- Games by quadrant, listing results and upcoming games
- Records by quadrant, away and neutral
- Non-Conference Strength of Schedule (SOS)
- Overall SOS
- Overall road and neutral records
- Non-Division I losses
- Average NET win and loss
- Overall record
- Non-Conference record, road record
- NET and other computer rankings
- Overall home records, non-conference and by quadrant
- Game scoring margins
- Conference records and standings
- AP Top 25, Coaches Poll
- Tournament history
a team has serious flaws if they have a poor record, get at least 3 in a normal year. its not that hard
because schools in mid majors and below do not get the opportunities. You think it doesnt matter if they schedule Wisconsin and beat them
I see Fremantle shot 60% from the charity stripe this year so he might fit right in.
It’s virtually impossible to have better than a mediocre profile without any road wins (if a team plays 30 games and 13-14 are road/neutral, obviously the best you could do is 19-11 if you are 2-11 on the road).
But I simply disagree that a team that is 16-1 / 2-11 is fundamentally more flawed than a team that is 14-3 / 4-9.
I don’t really get what you’re saying. Of course it matters if they beat Wisconsin. But it doesn’t matter extra just because Wisconsin is not in their conference.
Rutgers last year an example...they beat Nebraska big deal, what if they didnt beat Purdue, their resume was strong throughout except for the road record. If you have an outstanding home record but go 1-11 on the road its a sign that something is wrong
You keep saying it but I just don’t agree that there would have been anything more or less wrong with our team last year if they beat Michigan at home and lost to Purdue on the road.
I love it, in theory. But I can live with each of these incoherencies:
I can live with a bit of OOC bias because it incentivizes the scheduling of games that help differentiate between the top ~48 teams. More important in football, granted.
I can also live with bias toward quality wins. The tournament is played among quality teams and so I'd be slightly more interested in how a team performs against top competition than how they perform against bad teams.
Finally, but least of all, I can live with a very small amount of bias toward road record. The way it was applied to RU last year pissed me off, a team that was picking off top teams left and right at home and consistently just barely missing on the road against a tougher road slate (against the 11-14th seeded teams last year, RU had 4 home games and 1 road game). A little different for, say, the 2002 RU team that was just as good at home but complete ass on the road. I just think that once again, the tournament setting asks teams to win in unfamiliar confines and you'd like to see that they can do that over the course of a season.
I readily concede that none of these are justifiable in the context of a "body of work" with no recency factor, so it's tricky, but that body-of-work principle is flawed in itself because it could diverge from anticipated tournament performance, so I see these (at least the last two) as acceptable little hedges against that flaw.
