Top 25 talk

[kiko]

Thanks to deiscanton and Dave for the above.

I guess my question, even before the 2nd loss, is why is a team with only 7 wins ranked as high as #4 when, at the same time, there were several teams with 8, 9, 10, or 11 wins. Were those 7 wins against such high quality teams that they would carry more weight than teams with up to 4 more wins than Middlebury? If a team with, lets say, 10 wins at the same time Middlebury has only 7, and that team is good enough to be ranked 7th or 8th, it's hard to understand how they couldn't be ranked higher than a team with 3 less wins. Did Middlebury's 7 wins come against teams that, combined, were that much better than the 10 teams that others had beaten? And no, I am not picking on Mid here. Just using them as an example as I noticed they had so many fewer wins. Again, a team with 7 wins is better than a team with 10 or 11? 
Just hard to understand unless the 10-11 win teams had played all their games against teams with only 1 or 2 wins. That's all I'm asking. 

The simple answer to this is "because I believe the 7-win team is better than the 11-win team".  I'm not sure why you would consider the absolute number of wins as a factor.  Strength/impressiveness of results would play a role, sure, but just because I am impressed with Team A's win doesn't automatically mean I think they are better than a Team B who gorged on a cupcake.

[AndOne]

Thanks to deiscanton and Dave for the above.

I guess my question, even before the 2nd loss, is why is a team with only 7 wins ranked as high as #4 when, at the same time, there were several teams with 8, 9, 10, or 11 wins. Were those 7 wins against such high quality teams that they would carry more weight than teams with up to 4 more wins than Middlebury? If a team with, lets say, 10 wins at the same time Middlebury has only 7, and that team is good enough to be ranked 7th or 8th, it's hard to understand how they couldn't be ranked higher than a team with 3 less wins. Did Middlebury's 7 wins come against teams that, combined, were that much better than the 10 teams that others had beaten? And no, I am not picking on Mid here. Just using them as an example as I noticed they had so many fewer wins. Again, a team with 7 wins is better than a team with 10 or 11? 
Just hard to understand unless the 10-11 win teams had played all their games against teams with only 1 or 2 wins. That's all I'm asking. 

The simple answer to this is "because I believe the 7-win team is better than the 11-win team".  I'm not sure why you would consider the absolute number of wins as a factor.  Strength/impressiveness of results would play a role, sure, but just because I am impressed with Team A's win doesn't automatically mean I think they are better than a Team B who gorged on a cupcake.

This argument that "I believe the 7 win team is better than the 11 win team" doesn't hold much water. Why? Because Middlebury's 7 wins before tonight came against teams with a combined record of 36-35. One game over .500. Not exactly numbers that would strongly support a #4 national ranking. And tonight, they beat a team that that is now 5-7. Middlebury's now 8-2 record has been compiled against teams with a combined losing record of 41-42! This doesn't even sound like the record of a top 20 team, let alone a top 10 team.

[Dave 'd-mac' McHugh]

AndOne... go through the rest of the Top 25... Middlebury isn't the exception in that case. I am not voting for Lycoming because their opponent's record is low (far lower than Middlebury's if memory serves) yet they are 6th in the poll. That argument isn't going to work because you are going to find a LOT of teams in similar situations. If anything, it makes me feel better about my selection of Middlebury in my Top 25 and maybe higher, because I have been leaving out teams like Lyco, Juniata, Whitman, and others with their opponent's records being far lower (in some cases FAR lower).

And again... making an argument about a team's current ranking based on now two games worth information after the poll was voted on isn't fair, either. You are adding data and information to an argument that the voters haven't had the chance to decide on as of yet including a loss. Stick with their resume prior to the latest poll or it just gets too convoluted.

[Pat Coleman]

You're moving the goalposts here, AndOne, so I'm only going to address the original argument -- a variation of a couple of wins in the raw number in the 'W' column is a pretty poor way to determine a ballot. Everyone is going to end up with 24 or 25 regular season games, so as a voter, I wouldn't have cared so much if they were played in December or January.

[sac]

Knightslappy has Middlebury at #7 on his RPI chart.   They're in the ballpark of where they should probably be ranked.  Good teams sometimes lose, it happens.

[kiko]

Thanks to deiscanton and Dave for the above.

I guess my question, even before the 2nd loss, is why is a team with only 7 wins ranked as high as #4 when, at the same time, there were several teams with 8, 9, 10, or 11 wins. Were those 7 wins against such high quality teams that they would carry more weight than teams with up to 4 more wins than Middlebury? If a team with, lets say, 10 wins at the same time Middlebury has only 7, and that team is good enough to be ranked 7th or 8th, it's hard to understand how they couldn't be ranked higher than a team with 3 less wins. Did Middlebury's 7 wins come against teams that, combined, were that much better than the 10 teams that others had beaten? And no, I am not picking on Mid here. Just using them as an example as I noticed they had so many fewer wins. Again, a team with 7 wins is better than a team with 10 or 11? 
Just hard to understand unless the 10-11 win teams had played all their games against teams with only 1 or 2 wins. That's all I'm asking. 

The simple answer to this is "because I believe the 7-win team is better than the 11-win team".  I'm not sure why you would consider the absolute number of wins as a factor.  Strength/impressiveness of results would play a role, sure, but just because I am impressed with Team A's win doesn't automatically mean I think they are better than a Team B who gorged on a cupcake.

This argument that "I believe the 7 win team is better than the 11 win team" doesn't hold much water. Why? Because Middlebury's 7 wins before tonight came against teams with a combined record of 36-35. One game over .500. Not exactly numbers that would strongly support a #4 national ranking. And tonight, they beat a team that that is now 5-7. Middlebury's now 8-2 record has been compiled against teams with a combined losing record of 41-42! This doesn't even sound like the record of a top 20 team, let alone a top 10 team.

This is rather unique... I guess it is supposed to be logic.  That 8-2 record sounds like a team that lost to two very good teams who sit within shouting distance of Middlebury in the rankings.  (At the time the rankings were made, the only loss was in OT on the road to a team that is undefeated and sits just outside the top ten).  And it sounds like they have beaten a lot of teams that I would expect the #4 team in the country to beat.  Those wins don't tell me that Middlebury is the #4 team in the country, but they also don't tell me that Middlebury is *not* the #4 team in the country.  They are simply games that I would expect #4 to win, which Middlebury did.

You seem to be penalizing them for not beating teams that are not actually on their schedule.

[AndOne]

No, I'm not penalizing them for not beating teams that aren't on their schedule. What I am doing is just not rewarding them with the distinction of being the #4 team in the country when, as of the date of the most recent poll, they only had 7 wins against teams with a combined record of 36-35.

Now consider, let's say, York. Not only did they beat Middlebury head to head, but, as of the date of the most recent poll, they had 11 wins (Vs 7 for Midd), against teams with a combined record of 64-62. Thus, York's wins came against teams that were a combined 2 games over .500 while Midd's opponents were a combined 1 game over. Not a big difference, but York had 4 more wins (or are you going to penalize them for scheduling more earlier games?), than Midd at the time, AND beat Midd head to head. Given this, it just doesn't seem either logical or right that Midd was ranked #4 while York was down at #13. Sorry, but given this information, there is just no way you can say Midd was  better than York, as of the date of the most recent poll.

[Ryan Scott (Hoops Fan)]

Who cares about the combined record?  Most of the teams on everyone's schedule aren't very good.  That's just fact.  When I'm voting, I'm looking for the good teams they beat and the bad teams to whom they lost.  Especially early in the season, it's all about one or two games that might (repeat: might) tell us something about them.  At this point in the season, I also rely more heavily on seeing a team play.  Things change when you get to the end of the season, when conference strength enters into the equation and the separation between "good" schedule and "weak" schedule come into play.

I've had both Middlebury and Swarthmore in my Top 5 from almost the beginning of the season.  That game was slightly surprising, because I didn't think Swat would have as easy of a time of it as they did.  I'm sure Dave and I will discuss it a bit on Sunday (if our segment is still on the schedule).

Honestly, though, every team has its weaknesses - more stark, in most cases, than a typical year - I tend to vote on what teams seem strongest in the moment.  I considered switching my #1 vote to Augie last week and I'm glad I didn't, because we saw some new weaknesses exposed.

It's not all the games that really matter, but the few that are most telling - at least right now.

[Titan Q]

The toughest thing about voting in the Top 25 is comparing a team w/ a weak schedule (and a great record) to a team w/ a strong schedule (and more losses).

Use Illinois Wesleyan as an example.  IWU is 10-3 vs this schedule - http://www.d3hoops.com/teams/Illinois_Wesleyan/men/2017-18/index.

What would Whitman's record be vs that schedule?  Middlebury's? UWW's? Lycoming's? Swarthmore's? Juniata's?

[kiko]

No, I'm not penalizing them for not beating teams that aren't on their schedule. What I am doing is just not rewarding them with the distinction of being the #4 team in the country when, as of the date of the most recent poll, they only had 7 wins against teams with a combined record of 36-35.

Now consider, let's say, York. Not only did they beat Middlebury head to head, but, as of the date of the most recent poll, they had 11 wins (Vs 7 for Midd), against teams with a combined record of 64-62. Thus, York's wins came against teams that were a combined 2 games over .500 while Midd's opponents were a combined 1 game over. Not a big difference, but York had 4 more wins (or are you going to penalize them for scheduling more earlier games?), than Midd at the time, AND beat Midd head to head. Given this, it just doesn't seem either logical or right that Midd was ranked #4 while York was down at #13. Sorry, but given this information, there is just no way you can say Midd was  better than York, as of the date of the most recent poll.

This is nonsensical.  York has eleven wins against teams that, collectively are roughly .500.  Middlebury has seven wins against teams that, collectively, are roughly .500.  I am at a loss for how you extrapolate these four extra wins, which were over teams that I would expect either school to beat, as somehow providing a difference-maker from a ranking standpoint.  Especially, as the landlord pointed out, since everyone is going to play 24 or 25 in the end anyway.

[smedindy]

The toughest thing about voting in the Top 25 is comparing a team w/ a weak schedule (and a great record) to a team w/ a strong schedule (and more losses).

Use Illinois Wesleyan as an example.  IWU is 10-3 vs this schedule - http://www.d3hoops.com/teams/Illinois_Wesleyan/men/2017-18/index.

What would Whitman's record be vs that schedule?  Middlebury's? UWW's? Lycoming's? Swarthmore's? Juniata's?

There are measures out there that normalize all of that noise and make a strong signal.

[AndOne]

Fine, throw out the the combined records of the opponents both Midd and York beat.
By the time the most recent poll came out, York had BEAT Midd and had 4 more wins than Midd.
I think that's a fair argument supporting the assertion that Midd should not have been ranked higher than York.
By the end of the season, sure, Midd could easily end up being the better of the two. But as of 1/2/18 the facts showed otherwise.

[kiko]

Fine, throw out the the combined records of the opponents both Midd and York beat.
By the time the most recent poll came out, York had BEAT Midd and had 4 more wins than Midd.
I think that's a fair argument supporting the assertion that Midd should not have been ranked higher than York.
By the end of the season, sure, Midd could easily end up being the better of the two. But as of 1/2/18 the facts showed otherwise.

Sure, it is reasonable for a voter to rank York above Middlebury.  It's also reasonable for a voter to come to the reverse conclusion since York's win came by three points in OT at home.

But that's not what you were arguing.  You were wailing about the unfairness of a team that had beaten seven teams being ranked over a bunch of teams that had beaten eleven.  Which is a really embarrassing argument to make if you think about if for more than four seconds.  You're essentially arguing that if Middlebury had scheduled and beaten four additional .500ish teams just prior to Christmas, it would have somehow proven something to you about their place in the top 25 pecking order, which entirely consists of teams a lot better than .500?  Unless we've somehow teleported back to the 1950s, when this passed for sound analytics, that is some pretty astonishing reasoning.

[FCGrizzliesGrad]

While I agree you can argue that York should be ahead of Middlebury, the fact that they had 4 more wins is not the reason. Similar SoS, with a H2H win, and 1 fewer loss is why. It's not like football when everyone has played roughly the same amount of games at any point during the season. How can a team that's played 8 games possibly have 11 wins?

[me]

Middlebury about to lose again. Maybe they'll fall out of the top 5.

One look at the poll vs. the Massey ratings and it seems to me that it is very clear that the poll is a complete joke this year.

Me thinks some folks are mailing it in this year.