2016-2017 College Basketball Thread

[jbergey22]

I agree with the teams but the seeding was interesting to say the least.

Minnesota w a better seed than Michigan or Wisconsin?.
Wichita St should have been much higher(5 or higher)
SMU should have been a 4 seed
Florida should have been lower
St Marys should have been higher
Butler should have been lower.

[Arles]

Why are the Kenpom and BPI so high on St Mary's? They are 0-3 against the top 25 and 2-4 vs the top 50 - they don't really have a great win. Yet, both rankings have them higher than Arizona, UCLA and Oregon. I've seen all four teams play out west and putting them on par with those three PAC-12 teams is laughable. Anyone can win a game or two in the field (they certainly could beat AZ in round 2), but I just don't get that ranking. They got killed by Gonzaga three times (by 20 twice and 10 once). They lost to UT Arlington and their best win was a 3-point win against Dayton in Nov. They beat a lot of marginal to bad teams, but I just don't see why they are ranked that high.

[bronconick]

Syracuse's opening round opponent in the NIT is UNC-Greensboro. Hilarious.

[jbergey22]

Originally Posted by Arles

Why are the Kenpom and BPI so high on St Mary's? They are 0-3 against the top 25 and 2-4 vs the top 50 - they don't really have a great win. Yet, both rankings have them higher than Arizona, UCLA and Oregon. I've seen all four teams play out west and putting them on par with those three PAC-12 teams is laughable. Anyone can win a game or two in the field (they certainly could beat AZ in round 2), but I just don't get that ranking. They got killed by Gonzaga three times (by 20 twice and 10 once). They lost to UT Arlington and their best win was a 3-point win against Dayton in Nov. They beat a lot of marginal to bad teams, but I just don't see why they are ranked that high.

Well if you take the 3 Gonzaga games out(Kenpoms #1 team) they were 29-1. They had a few decent wins(Nevada, Dayton(as you mentioned), BYU(twice). Large margin of victory in a lot of games inflated their Kenpom rating I am sure.

[Arles]

I just don't see how you can be a top 15 team when you didn't beat a team in the top 25. Take Kenpom, St Mary's was the #14 team. Their best wins were #36 (Dayton), #55 Nevada and #74 BYU twice. They got completely blown out twice by Gonzaga and lost to #76 UT Arlington. I don't care if you beat all the teams between 100 and 200 by 20 (which they didn't), that's not a top 15 team. Arizona lost to Gonzaga by 7 (without AZ's best player), Butler (#26) also without Trier, UCLA (#18) and Oregon (#16). They beat #16, #18 twice, #41, #47 twice, #58 twice, #61 twice, #66 and #72 twice. That's 3 better wins than Dayton, 6 better than Nevada and 13 better than BYU. AZ also had the same number of losses and their worst is Butler, who's 50 spots better than UT-Arl. Yet, after all this, St Mary's is 6 spots higher than AZ? You can do a similar exercise with UCLA. It just doesn't make any sense for St. Mary's to be that high. It's almost like getting annihilated by Gonzaga numerous times helped them. What else did they do?

[Atocep]

Originally Posted by Arles

I just don't see how you can be a top 15 team when you didn't beat a team in the top 25. Take Kenpom, St Mary's was the #14 team. Their best wins were #36 (Dayton), #55 Nevada and #74 BYU twice. They got completely blown out twice by Gonzaga and lost to #76 UT Arlington. I don't care if you beat all the teams between 100 and 200 by 20 (which they didn't), that's not a top 15 team. Arizona lost to Gonzaga by 7 (without AZ's best player), Butler (#26) also without Trier, UCLA (#18) and Oregon (#16). They beat #16, #18 twice, #41, #47 twice, #58 twice, #61 twice, #66 and #72 twice. That's 3 better wins than Dayton, 6 better than Nevada and 13 better than BYU. AZ also had the same number of losses and their worst is Butler, who's 50 spots better than UT-Arl. Yet, after all this, St Mary's is 6 spots higher than AZ? You can do a similar exercise with UCLA. It just doesn't make any sense for St. Mary's to be that high. It's almost like getting annihilated by Gonzaga numerous times helped them. What else did they do?

You're referencing who beat who which is the point of RPI. Not analytics like KenPom, TeamRankings, Sagarin, ect. KenPom is a mixture of your SoS and overall efficiency and is more about predicting future results.

From a sports analytics standpoint, being able to consistently beat mediocre teams soundly is a better predictor of future results than beating good teams by close margins. This is how a team like Saint Marys can be rated higher than teams with stronger individual wins on their resume.

[Chief Rum]

Originally Posted by Atocep

You're referencing who beat who which is the point of RPI. Not analytics like KenPom, TeamRankings, Sagarin, ect. KenPom is a mixture of your SoS and overall efficiency and is more about predicting future results.

From a sports analytics standpoint, being able to consistently beat mediocre teams soundly is a better predictor of future results than beating good teams by close margins. This is how a team like Saint Marys can be rated higher than teams with stronger individual wins on their resume.

Is it a better predictor of results against the kinds of tourney teams St. Mary's will now face though? I am going to guess those metrics are a lot more accurate predicting results against mediocre opponents than against stronger teams.

Since outside of sub-12 seeded teams, you are facing Top 50 squads in almost any RPI ranking, I have to wonder what good such predictive models are within the context of the tourney.

[Butter]

LET'S GO FLYERS!

Wichita State and Kentucky in a single weekend, should be a piece of cake!

[Atocep]

Originally Posted by Chief Rum

Is it a better predictor of results against the kinds of tourney teams St. Mary's will now face though? I am going to guess those metrics are a lot more accurate predicting results against mediocre opponents than against stronger teams.

Since outside of sub-12 seeded teams, you are facing Top 50 squads in almost any RPI ranking, I have to wonder what good such predictive models are within the context of the tourney.

Compared to something like Rpi? TeamRankings and Pomeroy do a much better job.

And consistently being able to beat mediocre teams by large margins is a better predictor of furniture results against all competition than close games against good teams. That's why margin of victory is so important to metrics and why when it isn't used (BCS and RPI) you have a bad metric.

[Butter]

Originally Posted by Atocep

Compared to something like Rpi? TeamRankings and Pomeroy do a much better job.

And consistently being able to beat mediocre teams by large margins is a better predictor of furniture results against all competition than close games against good teams. That's why margin of victory is so important to metrics and why when it isn't used (BCS and RPI) you have a bad metric.

Yes, but what if ALL YOU EVER PLAY are bad to mediocre teams?

Wichita played 26 of their 34 games against sub-100 teams in the RPI. Every single team in the Missouri Valley Conference except Illinois State is rated 96 or worse in kenpom. They all are also terrible at both offensive and defensive efficiency, only Loyola cracks the top 100 in either, being #90 at offensive efficiency.

I guess by going 26-0 with hardly a close game in the bunch, that does tell you something...

I really think this single game is going to be a confirmation or repudiation of advanced metrics. Wichita State wins easily, you may see a more dramatic shift of the committee to taking those into account. If they lose, you will hear a lot about how they got it right after all by listening more to the body of work/RPI arguments.

[Atocep]

Originally Posted by Butter

Yes, but what if ALL YOU EVER PLAY are bad to mediocre teams?

Wichita played 26 of their 34 games against sub-100 teams in the RPI. Every single team in the Missouri Valley Conference except Illinois State is rated 96 or worse in kenpom. They all are also terrible at both offensive and defensive efficiency, only Loyola cracks the top 100 in either, being #90 at offensive efficiency.

I guess by going 26-0 with hardly a close game in the bunch, that does tell you something...

I really think this single game is going to be a confirmation or repudiation of advanced metrics. Wichita State wins easily, you may see a more dramatic shift of the committee to taking those into account. If they lose, you will hear a lot about how they got it right after all by listening more to the body of work/RPI arguments.

What you've done should matter far more than what you're capable of when it comes to ncaa selections and seeding. This is why, even though RPI needs to be replaced, Pomeroy and other metrics aren't a good replacement on their own. However, basing anything on one game is exactly the thing I'd expect from a bunch of old people that don't understand analytics.

[JonInMiddleGA]

Originally Posted by Atocep

However, basing anything on one game is exactly the thing I'd expect from a bunch of old people that don't understand analytics.

Try to avoid confusing "don't understand" with "not being particularly interested".

If the sheer volume of games was less, I'd pretty much guarantee that you'd hear the same eyeball test factor mentioned at a frequency akin to its appearance in college football discussions.

[JonInMiddleGA]

As for ditching the RPI, that process is apparently underway.

Here's a not-bad article from January's meeting on that subject, which seems enlightening because they have rather lengthy quotes from numerous participants in the meeting (including Sagarin & Pomeroy themselves).

In there is one of the very things that has come up here: whether we want something aligned with performance/results or something that's more predictive.

Also notable (to me anyway): Sagarin's own first proposal isn't even his own formula, Sagarin's extreme example of how an 0-26 team could end up with a high seed if you didn't control for that, and an analytics expert who points out how one criteria might be used for getting in/being left out but different criteria for seeding once the field is determined. Also noteworthy is a comment about how coaches seem to prefer a single metric (ala RPI) be used for the team sheets rather than multiple metrics.

What experts who met with NCAA say about changes to tourney selection process - CBSSports.com

[Butter]

The key argument they talk about is do you want the best team or the "most accomplished" team. I would personally prefer to err on the side of "most accomplished". The 0-28 theoretical was a good illustration of that.

The NCAA is about winning games against good teams. At some point, you should be selecting teams that have shown an ability to do that. I get that not all teams get the same opportunities to do that, but teams that are like Dayton and Wichita State should be able to figure it out.

In thinking more about this, I noticed last year that Wichita was in much the same boat, and ended up destroying their opening round opponents by 20, then beat 6 seed Arizona by 10. I believe they were top 15 in kenpom going into the tourney but still barely got an at-large.

[Thomkal]

March Madness continues for the Chants after all.They are playing at home on Wed, in the first round of the CBI. Their opponent, Hampton, has a losing record, which probably means we have one tourney too many.

[JonInMiddleGA]

Originally Posted by Thomkal

March Madness continues for the Chants after all.They are playing at home on Wed, in the first round of the CBI. Their opponent, Hampton, has a losing record, which probably means we have one tourney too many.

A lot of schools won't touch the pay-to-play tourneys so they have to really stretch to fill their brackets.

I know, for example, that Tennessee -- who is so young that I think there might have actually been a reasonable case to make for them playing in the CBI/CIT -- said beforehand they would not play in any tourney below the NIT.

[digamma]

If you need evidence in this debate, check Vegas lines versus KenPom's predictor on a weekly basis. I'm pretty sure KenPom isn't using the Vegas lines to run his predictor.

[JonInMiddleGA]

Originally Posted by digamma

If you need evidence in this debate, check Vegas lines versus KenPom's predictor on a weekly basis. I'm pretty sure KenPom isn't using the Vegas lines to run his predictor.

Then again (as you obviously know) Vegas lines exist not to predict games but rather to steer money to the casinos benefit.

So I'm not sure how/where they fit into this (yep, I'm a tad confused & trying to improve on that rather than simply trying to kick dirt on your post)

[Butter]

I do realize that there is a lot of debate to be had about metrics, so let's have one. What does everyone think?

I present, a somewhat less absurd case than Sagarin's 0-28 team from the article above... this year's Clemson team.

They go 17-15, 6-12 in the ACC. Yet the predictive metrics love them, ranking #36 in kenpom, #34 in BPI, and #41 in Sagarin. They played a tough but not impossible schedule, but also lost nearly all their close games against good teams. Should a team like that be in over someone like, for example, VCU.

VCU is #40 in BPI, #50 in Sagarin, and #52 in kenpom, despite going 26-8. They also didn't play near the schedule that Clemson did, but performed well against around the #100 SOS in the country.

What should win out? The team that is potentially really good but just had terrible luck? Or the team that has actually won more games, but also has a lower ceiling of "good"?

[JonInMiddleGA]

In that example, no, I would not put Clemson in the field over VCU.

The gap between them in the various metrics is not enough to overcome the disparity of their actual record.

[digamma]

Originally Posted by JonInMiddleGA

Then again (as you obviously know) Vegas lines exist not to predict games but rather to steer money to the casinos benefit.

So I'm not sure how/where they fit into this (yep, I'm a tad confused & trying to improve on that rather than simply trying to kick dirt on your post)

I fully recognize this, but I think the analytics serve as a starting point in setting the opening line, which of course will be adjusted to shade to the public market. During the tournament you are going to see the public effect magnified and find spots where the line is off by several points from the analytics. Over the season though, particularly in the small conferences, these are going to correlate pretty heavily. At least that's what I've seen from a few seasons of making things more interesting.

[JonInMiddleGA]

Originally Posted by digamma

I fully recognize this, but I think the analytics serve as a starting point in setting the opening line, which of course will be adjusted to shade to the public market. During the tournament you are going to see the public effect magnified and find spots where the line is off by several points from the analytics. Over the season though, particularly in the small conferences, these are going to correlate pretty heavily. At least that's what I've seen from a few seasons of making things more interesting.

So what is the actual correlation that exists (I think maybe I'm missing something that's in the thread somewhere already?)

Does Vegas seem to kinda match up to KenPom for predictives? Or is there a lot of disparity between them?

(And is this simply ultimately going back to the debate about whether we want performance metrics to rule vs predictive metrics to rule?)

[Arles]

OK, so I have an analysis program for college football, so I instead used KenPom as the ratings source and looked at W/L, margin of victory and ranking for both St. Mary's and Arizona. Now, bear in mind this is only using KenPom's rating - no RPI or anything else. So, you have the KenPom bias in the ratings (which plays a small part given I'm questioning the KenPom rating system for both teams). Still, here are the results:

top 50
------------Record, Total Margin, AVG per game
Arizona 5-4, -6, -0.7
St. Mary's 1-3, -47, -11.75

51-100
------------Record, Total Margin, AVG per game
Arizona 7-0, 57, 8.14
St. Mary's 4-1, 61, 12.2

101-150
------------Record, Total Margin, AVG per game
Arizona 8-0, 148, 18.5
St. Mary's 6-0, 104, 17.3

151-200
------------Record, Total Margin, AVG per game
Arizona 6-0, 84, 14.0
St. Mary's 3-0, 58, 19.3

200+
------------Record, Total Margin, AVG per game
Arizona 4-0, 85, 21.3
St. Mary's 14-0, 320, 22.8

I still don't see how St. Mary's is higher than AZ's ranking from this. Unless you really weigh those 14 games against 200+ crap opponents high, Arizona looks better across the board. They were 12-4 against the top 100 with a better record and overall margin per game (+3.8 vs +1.5) than St Mary's and won 7 more games (SM was 5-4). Arizona averaged facing the #38 team for it's 16 games while St. Mary's averaged facing the #43 team. Granted they had 3 games against #1, but AZ had 5 games against #16 and #18, one game against that same #1 team and another top 30.

From 100-200, Arizona was almost exactly the same at 14-0 at around +17 with St. Mary's at 9-0 at +18. Arizona averaged facing #135, while St. Mary's averaged #137. So, again, a pretty close comparison. Arizona was also at +21 for it's 200+ (but it only played 4 games) compared to St. Mary's +23 (but for 14 games).

I just don't get a system that rewards 200+ ranked team blowouts more than wins in the top 100 (even if it's a smaller margin). Essentially, if Arizona would have played 10 fewer top 100 teams (who they beat at a +4 margin) and instead beat 10 200+ teams by 22 - they would be ranked higher. That just seems stupid.

[digamma]

They are almost lock step. (And so, in the tournament, you can find value in games where they fall out of touch.)

[JonInMiddleGA]

Originally Posted by Arles

I still don't see how St. Mary's is higher than AZ's ranking from this.

The short answer appears to be a higher "adjusted offensive efficiency". (there's only a 0.5 diff in their adjusted defensive efficiency, the gap in offense is several times higher)

The longer answer might be found in his Oct 2016 explanation of recent changes in how he computes those.

Ratings methodology update | The kenpom.com blog

If I've followed the voluminous amount of words correctly, St. Mary's basically beat those 200s they played by a greater amount over expectations than AZ beat those 100s vs expectations.

[dfisher]

Also, the KenPom ratings are based on efficiencies and a per possession metric. Arizona plays at a faster tempo than St. Mary's so they would have to win by larger margins to account for the discrepancy.

[Atocep]

Originally Posted by Arles

OK, so I have an analysis program for college football, so I instead used KenPom as the ratings source and looked at W/L, margin of victory and ranking for both St. Mary's and Arizona. Now, bear in mind this is only using KenPom's rating - no RPI or anything else. So, you have the KenPom bias in the ratings (which plays a small part given I'm questioning the KenPom rating system for both teams). Still, here are the results:

top 50
------------Record, Total Margin, AVG per game
Arizona 5-4, -6, -0.7
St. Mary's 1-3, -47, -11.75

51-100
------------Record, Total Margin, AVG per game
Arizona 7-0, 57, 8.14
St. Mary's 4-1, 61, 12.2

101-150
------------Record, Total Margin, AVG per game
Arizona 8-0, 148, 18.5
St. Mary's 6-0, 104, 17.3

151-200
------------Record, Total Margin, AVG per game
Arizona 6-0, 84, 14.0
St. Mary's 3-0, 58, 19.3

200+
------------Record, Total Margin, AVG per game
Arizona 4-0, 85, 21.3
St. Mary's 14-0, 320, 22.8

I still don't see how St. Mary's is higher than AZ's ranking from this. Unless you really weigh those 14 games against 200+ crap opponents high, Arizona looks better across the board. They were 12-4 against the top 100 with a better record and overall margin per game (+3.8 vs +1.5) than St Mary's and won 7 more games (SM was 5-4). Arizona averaged facing the #38 team for it's 16 games while St. Mary's averaged facing the #43 team. Granted they had 3 games against #1, but AZ had 5 games against #16 and #18, one game against that same #1 team and another top 30.

From 100-200, Arizona was almost exactly the same at 14-0 at around +17 with St. Mary's at 9-0 at +18. Arizona averaged facing #135, while St. Mary's averaged #137. So, again, a pretty close comparison. Arizona was also at +21 for it's 200+ (but it only played 4 games) compared to St. Mary's +23 (but for 14 games).

I just don't get a system that rewards 200+ ranked team blowouts more than wins in the top 100 (even if it's a smaller margin). Essentially, if Arizona would have played 10 fewer top 100 teams (who they beat at a +4 margin) and instead beat 10 200+ teams by 22 - they would be ranked higher. That just seems stupid.

As mentioned, it's about expected margin of victory against your competition and efficiency on a per possession basis. What Pomeroy's ratings are saying is that on a per possession basis Saint Marys has been a bit better than Arizona relative to its schedule.

Arizona was +.15 point per possession better than its opponents in conference play this year (3rd in the PAC12) while Saint Marys was +.25 better in conference play. That's a rather sizable difference. Enough to make Saint Marys better considering the competition? Hard to say, but Pomeroy's ratings think so.

FWIW, TeamRankings has Arizona 20th and Saint Marys 22nd in its predictive ratings.

[Arles]

I see the logic used in KenPom, but I just don't think it makes a lot of sense. The WCC had 3 teams in the top 75, 3 from 100-200 and 4 from 240+. That means 14 of their 18 conf games were against sub 100 opponents (8 against sub 240). The Pac-12 had 7 in the top 75, 4 from 100-200 and 1 over 200 (OSU). Arizona only played OSU once because of the unbalanced schedule, so 9 of their 18 games were against top 75 teams and 8 were 100-200. Arizona also had a much tougher nonconference by facing 4 more top 75 teams in nonconference compared to St. Mary's 2.

You can also look at common opponents. Both played Gonzaga at home, UC Irvine at home, Santa Clara on the road and Stanford on the road. In those 4 games, Arizona was +62 while St. Mary's was +48. I know it's a small sample size, but those are exact copies with Arizona playing all four without its best player (Trier). You can torture the numbers all you want, but I just can't fathom a ranking system that would have St. Mary's favored against Arizona given the season both have had. Arizona had 24 games against the top 150 and finished with a net margin of +8.3 per game. St. Mary's finished with 15 games against the top 150 and a net margin of +7.8. Now, St. Mary's did have 10 more games against sub 200 teams, but that shouldn't overshadow Arizona's performance in the top 150 compared to St. Mary's (esp the top 50).

[Chief Rum]

IMO, what stands out from what Arles said and makes everything crystal clear to me is that Arizona played top opponents much better than St. Mary's did, while they did roughly the same against the other groupings. That tells me Arizona is a better team than St. Mary's and any ratings system that says otherwise must be considered quite questionable.

[Atocep]

Originally Posted by Chief Rum

IMO, what stands out from what Arles said and makes everything crystal clear to me is that Arizona played top opponents much better than St. Mary's did, while they did roughly the same against the other groupings. That tells me Arizona is a better team than St. Mary's and any ratings system that says otherwise must be considered quite questionable.

You could take 2 teams and make an argument questioning the validity of any ranking system by that measure.

[digamma]

KenPom himself speaks: How the NCAA abuses statistics to stack the deck against small schools like Illinois State.

[JonInMiddleGA]

So what's our conclusion here then?

That we don't like RPI and we don't like KenPom?

Why am I suddenly reminded of how the whole BCS ratings thing went?

[molson]

If Marquette and Illinois State swapped schedules, the Golden Eagles would almost surely lose some games to teams outside the top 100. If you put Illinois State in the Big East, it would have earned some quality wins

And that's the crux of it. Should selection committees assume that teams with weaker schedules would win theoretical games against better teams, or assume that teams with stronger schedules would lose theoretical games against worse teams?

I can understand why selection committees and poll voters give more credence to actual wins than theoretical wins. Even if that means teams from smaller conferences have a smaller margin for error.

[JonInMiddleGA]

re: the whole "oh poor mid-majors, P5s don't schedule them" thing

Unless you're a high profile mid-major, doing so likely carries risks at the gate.

Take this quick anecdote from a non P5 school who had a pretty weak OOC schedule. Memphis averaged 7,100 turnstile fans per conference game, vs 6,300 for all home games overall. Not "announced" attendance, but actually honest-to-goodness turnstile count.

Best I can figure quickly is that they had 9 conference games at home and 9 non-conference games at home this year (18 total home games). Taken at the point where the article was written (with 2 conf. home games left) the math seems to be 5,677 for non-conference games vs 7,100 for conference games. That's a 20% difference, not insubstantial.

Now if a non-P5 team has trouble drawing as well with non-conference opponents, how much better is it going to get with a P5 school? Illinois State barely filled half their own house last year (NCAA hasn't published the current attendance numbers).

They were just 18-14 last year before a much better 2017 ... but would have looked like a P5 team scheduling a middling mid-major opponent when the schedules came out.

It's a largely moving target that P5s would be expected to hit AND they would be doing so to the detriment of their own program & their own gate. Where's a rational motivation for them to do that?

[JonInMiddleGA]

Originally Posted by molson

And that's the crux of it. Should selection committees assume that teams with weaker schedules would win theoretical games against better teams, or assume that teams with stronger schedules would lose theoretical games against worse teams?

I can understand why selection committees and poll voters give more credence to actual wins than theoretical wins. Even if that means teams from smaller conferences have a smaller margin for error.

Which puts us right back at performance vs projection arguments.

[Arles]

I think it comes down to philosophy. When comparing teams, I think you need to look at teams that have a puncher's chance to beat you. No sub-200 team was beating either Arizona or St Mary's this year. So, if you look at their top 150 opponents, that seems more reasonable. The question is should a 22-point win against a sub 240 team be worth the same as a 9-point win against a top 75-150 team. It seems that KenPom values that 22-point win more, which I just disagree with.

Another point is that when Arizona played the better teams like UCLA (18), Utah (47) and Cal (58) - they had to face USC (61), Colorado (72) and Stanford (101) right before or after. So, they couldn't really rest guys in preparation for the big opponent. When St Mary's played the 2 best teams in Gonzaga (1) and BYU (74), they got to face Portland (#278) and Pepperdine (#299) right next to them. That's an enormous advantage to have essentially a BYE before or after a bigtime game. Before their loss to Gonzaga, they beat a hapless Portland team 74-33 and essentially benched their starters down the stretch. When they beat BYU in two games, they beat Pepperdine by 29 and that same Portland team by 23 right next to the BYU games. Again, they got to rest everyone in prep of their big game. When Arizona lost to UCLA, they had to play their starters 30+ min two days before to fend off USC. When they beat Utah by 10, they had to play their starters 30+ minutes against Colorado in the other game. Again, had they had weaker opponents, they may have been more rested/prepared for the other teams. I'm not sure how to quantify that factor, but I think it plays a part.

The only time St. Mary's ever had to play 2 top 100 teams in the same weekend, they beat BYU and then lost to Gonzaga by 20. They had their lowest win margin against both BYU and San Francisco (#105) when they played them over the same weekend. That can't be a coincidence (the BYU game was even at home). But that's how the NCAAs are. They will be facing #52 VCU and then #20 Arizona over a 3 day period, it will be interesting to see how they do playing their starters big minutes in both games.

[JonInMiddleGA]

Originally Posted by Arles

No sub-200 team was beating either Arizona or St Mary's this year.

Is that non sub-200 KenPom?

Cause I noticed that Washington (#210 RPI) lost to Arizona by 11 on the road & by 8 at home.

[Scarecrow]

Massey Ratings

Fun little website to compare 2 teams. I just randomly picked 2 teams for show - just a coincidence that it's K-State (the last team in) and Illinois State (the small school that you think should be in)

[Arles]

Originally Posted by digamma

KenPom himself speaks: How the NCAA abuses statistics to stack the deck against small schools like Illinois State.

I think he has a point about some "bad losses" hurting smaller school, but I don't think the Tulsa loss hurt a ton (they were #143). What hurt Illinois State is they played a terrible non conference schedule. They played one top 100 team (TCU) by KenPom's own rating and lost by 9. They played another team at 105 (SF) and lost by 8. They had one nice win (home against #8 Wichita State), but lost by 40 to them on the road and 20 in the tourney. Their other top 100 game was a home one against Loyola Chicago (#96) and they won by 2. So, when you lose your only 2 non conference games against teams with a pulse (TCU and SF) and get beat by a combined 60 points in two other games against the best team you beat, it's a hard sell to get in. That's not even counting the losses to Murry State and Tulsa.

I take issue with this statement:

If Marquette and Illinois State swapped schedules, the Golden Eagles would almost surely lose some games to teams outside the top 100. If you put Illinois State in the Big East, it would have earned some quality wins. No doubt, though, the Redbirds would do much worse than their 17-1 Missouri Valley Conference record when facing the tougher competition. But consider that Xavier went 8-10 against Big East teams not named DePaul and easily earned an at-large bid. The standard for small-conference teams is incredibly high, while the standard for major-conference teams is not as high as you think.

Let's look at Xaiver first, they scheduled Baylor (13) and Colorado (72) on the road and lost both (Colorado was only by 2). They beat Utah (#46), Wake Forest (#30) and Clemson (#35) in other nonconference games (Utah was a neutral site). Marquette played Michigan (21), Wisconsin (23), Vanderbilt (34), Pitt (75) and the same San Francisco team that beat Illinois State (Marquette won 78-65). They went 2-3 against those teams and also beat the top team in the country (Villanova) at home while only losing by 9 on the road (not 40).

Had Illinois State scheduled 5 top 75 nonconference opponents (maybe 1-2 at a neutral site tourney a la Xavier and Marquette), they may have made the field. But they scheduled 1 and lost. You can't be in a crap conference and then schedule patsies in nonconference, lose 7 games and get in. It just won't work. Had they beaten San Francisco or TCU or Tulsa or not gotten drubbed twice by Wichita State, they would probably be in. That's not the same as the underseeded situation with Wichita State and St. Mary's. Both those teams had good wins, only lost 4 games and did well against top 150 opponents. They deserved seeds in 5-6 range, not in the 7-10 spot. That's where I think analytics do need to get better overall. But I also think that going overboard with large margin crap wins like KenPom does to have Wichita and St. Mary's in the top 15 is just as wrong.

[Arles]

Originally Posted by JonInMiddleGA

Is that non sub-200 KenPom?

Cause I noticed that Washington (#210 RPI) lost to Arizona by 11 on the road & by 8 at home.

Yeah, I'm using only KenPom ratings since I'm arguing his system. It's only fair, IMO. Washington is 168 in his system.

[JonInMiddleGA]

Women's bracket looks like drunken chimps handled the location assignments, and took over the seeding for part of the process too.

[GoldenEagle]

I am heading to Vegas tomorrow and would like to place bets somewhat intelligently (even though it is difficult to beat the point spread). My general philosophy is to bet against the blue bloods as usually a lot of money comes in on them and drive the spread up. But what else should I look at? I have looked at KenPom and the Sagarin ratings but that is about it. I think someone mentioned last year using 538, but not sure how to do that and account for the spread.

[cuervo72]

Originally Posted by JonInMiddleGA

Women's bracket looks like drunken chimps handled the location assignments, and took over the seeding for part of the process too.

If they put drunken chimps on the court too I might actually watch...


(In fairness, I watched 0 men's games this season. But the opening here was too easy.)

[muns]

538 has it's own spread so you can compare the 2 to see if there might be value on any given game you are looking at

Sent from my SM-G730V using Tapatalk

[murrayyyyy]

Am I the only one with a new cable provider and have no clue where truTV is?

[jbergey22]

Originally Posted by GoldenEagle

I am heading to Vegas tomorrow and would like to place bets somewhat intelligently (even though it is difficult to beat the point spread). My general philosophy is to bet against the blue bloods as usually a lot of money comes in on them and drive the spread up. But what else should I look at? I have looked at KenPom and the Sagarin ratings but that is about it. I think someone mentioned last year using 538, but not sure how to do that and account for the spread.

So many people use Sagarin and KenPom that the market advantage has already gone away.

One early thing I have found is Virginia Tech is top 10 in the nation in 3 point percentage while Wisconsin is in the bottom 20 percent at guarding the 3 pointer. Virginia Tech might be worth looking into.

[JonInMiddleGA]

Watching GT vs IU.

Tums or Rolaids oughta put spots in this game, some real heartburn inducing action by both.

[digamma]

Originally Posted by jbergey22

So many people use Sagarin and KenPom that the market advantage has already gone away.

I think the first round of the tournament is actually the one time where KenPom still has value on a game to game basis.

[BishopMVP]

Originally Posted by GoldenEagle

I am heading to Vegas tomorrow and would like to place bets somewhat intelligently (even though it is difficult to beat the point spread). My general philosophy is to bet against the blue bloods as usually a lot of money comes in on them and drive the spread up. But what else should I look at? I have looked at KenPom and the Sagarin ratings but that is about it. I think someone mentioned last year using 538, but not sure how to do that and account for the spread.

I haven't bet seriously in years, but I always preferred money lines during bowl season & March Madness. You think Virginia Tech has a shot vs Wisconsin? I'd rather take the +200 than +5.5. VCU/St. Mary's +170 vs +4. Oklahoma State/Michigan +125 vs +2. Dayton/Wichita State +220 vs +6. UNCW/Virginia +290 vs +7. Just never liked depending on which team hits desperation shots/FT's at the end of a close one, or which 1/16 seed tries during the last 10 minutes of a 25 point game.

[JonInMiddleGA]

Yeah, it's the NIT. But at least this GT team manages to come up with moments to cheer about, and the fans are trying to believe in at least a little something good.

That's progress.