Gotcha then this is where we disagree. That “slim difference” between 1 pt win and loss is the only difference that has ever mattered before the nerds took over (just having some fun 😃). Maybe on a spreadsheet those numbers are barely different and you are correct the opposite outcome was a play or 2 away, but that’s the difference between a winning team and a losing team and all that matters in March. I think these “victory points” you refer to should be weighted much higher than they currently are, with Bonas efficiency spreadsheet game counting just a wee bit
2023–24 NET Ratings
- Thread starter SFspidur
- Start date
Your individual game is obviously huge, but the NET is also driven by how all of your opponents do. So, you could have a real nice Q2 win, but if a lot of your opponents lose their games, or drop in the NET, you aren't gonna move up much, if at all.
Yes. And you don't really have to consider wins / losses explicitly because Margin accounts for it anyway. + Margin (win) / - Margin (loss)
As stated, Margin is indeed a predictive measure.
This is the part we can agree on. I think margin in the NET counts too much. That's why when I built my system I placed a cap on how much it counts or to what extremes it counts. I do think margin or efficiency preferably matters in ranking, but the question is how much it should count.
In the case of the Bona vs URI game, margin probably counted too much. I don't think there's any value in winning by 35 or whatever they won by vs winning by 20 or so. It doesn't prove anything about your rank to win by that additional 15 points or so in my view.
How about this. Our win over Bona helped us even more because of their large margin of victory over URI.
You may not be able to pinpoint exactly how it influenced our rank precisely, but it's accounted for in the equation.
so quick question: we seamlessly moved in this discussion from "efficiency" to "margin of victory" - is MOV a criteria? I had assumed that when discussing efficiency, we were discussing things like shooting %, points per possession, etc.
So in the Bona example, if they went 9-10 FG in the last 6 minutes and ran it up against URI, would that account for metrics in their favor (as opposed to going 3-10, but URI 0-10 and Bona wins by a similar amount?)
Are the two intertwined? I'm losing the thread a bit.
So in the Bona example, if they went 9-10 FG in the last 6 minutes and ran it up against URI, would that account for metrics in their favor (as opposed to going 3-10, but URI 0-10 and Bona wins by a similar amount?)
Are the two intertwined? I'm losing the thread a bit.
Efficiency margin is just MOV/possessions, most people have an easier time understanding MOV so it is easier to talk about.
Of course they wouldnt since ch they were already ranked 5th. In your explanation it seems that the outcomes are already determined and no movement in rankings will occur.
But you know what, Im too old school for this. You are likely right but it will never make sense to me.
So a team that loses 20 games by only 1 point should be ranked nearly as high as a team that wins 20 games by 1 point? Joking of course but trying to understand.
Its crazy to judge a teams season of success or failure on anything more than wins and losses, with a measure of strength of schedule added. You play to win the games, not to have “efficiency” or MOV.
Crazy IMO but Im old.
A team with 20 losses by 1 point would obviously be ranked lower than a team with 20 wins by 1 point against similar competition.
If I were a betting man, though, I would use the margin of victory predictive measure as a clue that the losing team might just pull one off soon.
Their rank would probably not be great, but it would likely be higher than you might expect if you just glanced at their record.
If you only cared about wins, the rankings would clump all the undefeated together, followed by all the 1 loss teams, followed by all the 2 loss teams, etc. That obviously wouldn't make sense because of strength of schedule as you pointed out, right? So, margin is just another metric in addition to strenth of schedule to determine the rank.
Let me give an example:
If the hypothetical 20 win team with one point victories had only beaten teams ranked 350 or worse, would you pick them to advance far into the post season when they start playing teams in the top 50? What if that same team playing that same schedule won each of those games by 20+ points? They'd be a bit of an unknown for sure, but I might be more likely to pick them as a Cinderella.
On the flip side, what if that 20 loss team had lost all their games to top 10 teams by only 1 point? They are certainly not as bad as their record indicates, right?
Now, the real question was what if the 20 win team with one point victories and the 20 loss team with one point losses played an equal mixed schedule (some tough, some easier for both teams). The 20 win team would be ranked higher for sure, but both teams would be somewhere in the middle of the bell curve. Neither would be ranked extremely high or extremely low. There would, however, be reasonable separation between them because they've each made habits of either winning or losing.
That 20 loss team that somehow won their conference tourny looks like a fun Cinderella pick. Their seed will be low, but they have proved they can hang around in their games, and I might just pick them in my bracket to pull off an upset. The ranking system including margin would help predict them as a possible upset alert.
Thank you for the lengthy explanation but I am lost with this probably because of my age and not so geared toward formulas and algorithms like the newer generations.
The idea that metrics and offensive efficiency ratings, etc. now plays a determining factor in determining a teams success rather than wins and loses is mind blowing to me. Sorry.
There have been many “great” “talented” teams that underachieved and had mediocre records, maybe due to taking nights off, not being prepared, etc. And also some less talented teams that overachieved through sheer will, fight and strategy, and ended up with great records. Now the underachieving “talented” team with a mediocre record is said to have a successful season because they occasionally play amazing and their efficiency numbers are high? Im not buying it. They may be talented and may be able to win games in the dance but they lost too many games and dont deserve to be there.
The ranking should not be trying to determine which teams are the most talented, or which can go far in March, they should be determining which teams are having the best seasons, and which deserve a chance in March.
Your example of teams ranked by record alone was unfair because of course there has to be a strength of schedule factor involved, and that has always been factored in rather routinely in the past. The systems of old worked fine imo. Controversy of who doesnt make it is always going to happen. This system, with its many derivatives, is trying way too hard to be perfect.
The talented team that underachieved is not going to make the dance. Unfortunately, this is often us. Hopefully not this year, but we did hurt ourselves early with a few losses that could have been wins. Of course, we have a few wins that could have easily been losses, too. In the end, you do have to win. That 0-20 team with all 1 point losses is not going dancing.
The only way that underachieving talented team makes the dance is by winning their conference tourny. And in that case a predictive ranking such as this may suggest they have a reasonable chance at an upset.
The only real change is adding margin to the already existing strength of schedule component.
Beat a bad team by a little = unimpressive
Beat a bad team by a lot = doesn't hurt you
Lose to a bad team by a little = bad
Lose to a bad team by a lot = very bad
Beat a good team by a little = impressive
Beat a good team by a lot = very impressive
Lose to a good team by a little = doesn't hurt you
Lose to a good team by a lot = unimpressive
Just playing a good team shouldn't be enough by itself to boost your resume.
Likewise, just playing a bad team shouldn't automatically hurt your resume.
My biggest gripe is always when a power conference team with a record something like 18-14 who is also under 0.500 in conference gets a bid.
Would they be favored in a neutral site game against a mid-major who went 26-6 against a weaker schedule? Probably.
But the reason we play the games is to win - and if you have a season long body of work that says you'll barely win more than half of your games, I don't like rewarding that with a tournament bid.
Would they be favored in a neutral site game against a mid-major who went 26-6 against a weaker schedule? Probably.
But the reason we play the games is to win - and if you have a season long body of work that says you'll barely win more than half of your games, I don't like rewarding that with a tournament bid.
This system actually helps counter that scenario of an 18-14 "power" conference team. This says playing a tough schedule is not enough. It still probably helps because you have more chances at good wins, but you better win some and you better make sure the ones you lose are close, or it will hurt you a bit.
It also gives that 26-6 team a chance to prove its worth by handling its weaker schedule easily.
BTW, I quoted "power" because I hate separating conferences into power and mid-major. So derogatory.
Also remember, the NET is not the be all end-end all decider for the the NCAA's. It is just one objective tool the committee uses to evaluate who should be in and who shouldn't.
Looking at this objectively, beating a 10-7 Duquesne team by 2 points on the road, probably shouldn't move the needle that much. Yes, internally it was a great win for us, but really we are talking about beating a pedestrian Duquesne team. If any other team got a 5-6 Net jump on that (say VCU), we would have been the GTFO.
Looking at this objectively, beating a 10-7 Duquesne team by 2 points on the road, probably shouldn't move the needle that much. Yes, internally it was a great win for us, but really we are talking about beating a pedestrian Duquesne team. If any other team got a 5-6 Net jump on that (say VCU), we would have been the GTFO.
I think I am to blame for starting this conversation (or to credit, I guess!) but as mostly a casual observer of NET, it just has seemed to me that when a team wins a road game against a team ranked very close to it, it almost always moves up at least a little bit, which is why I was surprised that we didn't. It makes sense logically why we might not, given the close score, but it was just something I hadn't really noticed happening before.
And as usual, we have vcu sitting there not too far behind us despite losing like five home games and a couple games to lousy teams, and I have a feeling that they will soon go on the road, beat a team ranked close to them by a few points and jump 10 spots in the NET.
And as usual, we have vcu sitting there not too far behind us despite losing like five home games and a couple games to lousy teams, and I have a feeling that they will soon go on the road, beat a team ranked close to them by a few points and jump 10 spots in the NET.
Yeah, I have not idea how VCU is sitting that close to us. They have a Quad 4 home loss and didn't leave the Siegel Center until conference play started. Both of which should be very detrimental to the NET.
As usual, you're hilarious how you phrase this! Love you, man.
Looking at vcu's results, I don't see anything special, but I don't see anything special about their rank either. They have a few questionable losses, a few convincing wins, a few close losses to good teams. Kinda blah.
At a glance, it seems we should be ranked ahead of them but not by light years. We have a couple losses we would like to have back, a few convincing wins, a couple close losses to good teams...
I like our team right now, though. We are playing well and confident with high energy.
Not takings sides. Actually, my mind is about to explode reading this. Tedious. Very.
Just win the next game! Davidson away. Any rankings will take care of themselves with a season of data. Either we prove we belong or not. It is on the team, not the relative rantings' metrics.
Just win the next game! Davidson away. Any rankings will take care of themselves with a season of data. Either we prove we belong or not. It is on the team, not the relative rantings' metrics.
Spiders moved back up to 81 overnight. Nice job on travel day, guys!
(VCU again showing the importance of margin...moved up 5 spots by handily beating the lowest-rated team in the A-10 without needing to go on the road.)
Yeah. Margin should be a factor, but it appears to be too much of a factor in the NET system in my opinion.
SF, where do we need to be from a net perspective to have a realistic shot at an at large? What was the three highest net rankings that received at large bids since the new system was implemented?
I rest my case...NET makes no sense. We beat Duquesne on the road by 2 when they are a top-100 team, we stay the same. We beat Davidson on the road by 5 in OT when they are ranked 126 and we move up eight??
Our Loyola road win is getting better, as they are 5-1 in league.
Rutgers got in at #77 a couple years ago, but they had 6 Q1 wins. Without a NET better than 50, our issue will be a lack of Q1 wins to convince the committee we belong. So beating Dayton is a must if we’re hoping for an at-large.
But, it doesn’t just look at our game individually, it looks at every game played, both that day and for the season. If all of the teams we have played lost yesterday, we would not have jumped up 8 spots. And, same if all the teams in front of us won yesterday.
I don't think it is even close to a must win. Would it be a huge win and help our chances? Of course. But, the A-10 did well OOC this year, has 7 teams in the top 92 of the NET right now, and could definitely be a 2 bid or more league. If we go 14-4 or better A-10, finish 2nd to a top 20 NET, maybe even top 10 NET Dayton team ( they are 15 right now), and lose to Dayton in the A-10 finals, our resume should look pretty good at 24-10 or better compared to other bubble teams.
If all of that gets our NET up to the ~50 range, then yeah, I could see it. I'm just not sure how likely that is.
Three of our past four games have been Q2 wins, which has obviously been great. But as it stands right now, we have Q1 Dayton and then two other Q2 games on the entire rest of the schedule. Everything else is Q3 with a Q4 against La Salle, so outside of a few select opportunities, we're in the "should win, doesn't help résumé a whole lot" zone.
Yes, the ratings/values can still shift (both previous and upcoming games, in either direction), and the A-10 tourney potentially adds a couple more solid games, so yeah, should be possible. Again, just not sure how likely.
Whatever, I'm enjoying the winning streak, and the NET is gonna do what it does. One game at a time.
Three of our past four games have been Q2 wins, which has obviously been great. But as it stands right now, we have Q1 Dayton and then two other Q2 games on the entire rest of the schedule. Everything else is Q3 with a Q4 against La Salle, so outside of a few select opportunities, we're in the "should win, doesn't help résumé a whole lot" zone.
Yes, the ratings/values can still shift (both previous and upcoming games, in either direction), and the A-10 tourney potentially adds a couple more solid games, so yeah, should be possible. Again, just not sure how likely.
Whatever, I'm enjoying the winning streak, and the NET is gonna do what it does. One game at a time.
Perhaps it's the VMI effect as GK suggested. They moved up to 355 from 360 with their first D-I win! Think that moves them up to a Q5 win for us.
Maybe the NET didn't actually believe its own rank putting Duquesne in the top 100!
Torvik has his teamcast feature where you can play with results, and it does include a NET column. Obviously the actual formula isn't public, so I'm not sure how he's generating those numbers and how much faith to put in them.
Caveats aside, his predictions have us going 14–4 in A-10 leading to a NET of 51, but the 16th team out with a 2.1% chance, likely due to the quality win issue.
This obviously doesn't include any A-10 tournament games, but you can add them against whatever opponents you want if you like, and also play with the results of the remaining games.
Edit: Playing around with this, adding wins isn't bumping our NET up much, although we do move up the "first teams out" list. Giving us wins over Dayton, VCU, URI, and Mason, while still leaving the rest as probabilities moves us to a predicted 16–2 A-10 record and third team out, but NET is still just 49. So yeah, I'm suspect of his ability to predict NET, unsurprisingly.
Caveats aside, his predictions have us going 14–4 in A-10 leading to a NET of 51, but the 16th team out with a 2.1% chance, likely due to the quality win issue.
This obviously doesn't include any A-10 tournament games, but you can add them against whatever opponents you want if you like, and also play with the results of the remaining games.
Teamcast for Richmond - Customizable College Basketball Tempo Free Stats - T-Rank
T-Rank College Basketball Ratings and Rankings. And fun.
barttorvik.com
Edit: Playing around with this, adding wins isn't bumping our NET up much, although we do move up the "first teams out" list. Giving us wins over Dayton, VCU, URI, and Mason, while still leaving the rest as probabilities moves us to a predicted 16–2 A-10 record and third team out, but NET is still just 49. So yeah, I'm suspect of his ability to predict NET, unsurprisingly.
I think the A10 is a one bid league if Dayton wins the regular season and the tournament. The one “chance” of this not being true is if the 2nd place team beats Dayton in the regular season and makes it to the championship and loses to Dayton. So I agree with you SF.
Otherwise, the Spiders have to win the A10 tournament to get in. Sure the A10 did “better” in OOC overall, but conference affiliation isn’t going to push UR over the top. The committee is going to look at what Richmond did and there simply isn’t a Q1 win to justify inclusion unless they beat Dayton. Not having a “bad loss” is nice, but that means nothing to me in terms of the NCAA tournament. I want a mid-major who has proven they can beat the best (the teams in the dance), not that they can avoid losing to a team they should beat (teams not in the dance).
Dayton is a must win. I’ll die on that hill.
