The Last 4 In from 2013/2014

[jncuse]

This idea came from Pearl, who posted the resume the other day of NC ST who was clearly the last team in last year. The final spot came down to SMU (who 90% had in) or a poo-poo platter of teams (which NC St was given the prize), and SMU's bad losses did them in as I expected.

RPIForecast gives us the following resume entering the ACC tourney at 20-11

Record: 20-11
RPI - 47
1-5 vs top 25 (Louisville *)
2-7 vs top 50 (H Louisville, N Iowa)
6-11 vs top 100
2 Sub 100 RPI Losses (BAD LOSSES) (at Michigan, at Pitt*)


* Assumed as it is our highest win or lowest win probability of those games -
Note (I just picked one of the Big 3)

At 19-12 our Resume may be the same assuming we lose that extra game to a team in that 50-100 range. Although it could be a 3rd bad loss (or no top 25 win) Our RPI may be closer to 50-54.

Let's look at the teams that went to Dayton last year, I had to make estimates of RPI as I saw post NCAA number from Nolan:

Xavier:
20-12
RPI - 52
1-25: 1-4
1-50: 4-5
Top 100: 9-10
Bad Losses: 3 - USC, Seton Hall (2)

Iowa
19-12
RPI - 56
1-25: 1-8
Top 50: 3-10
Top 100: 5-12
Bad Losses - None

Tennessee:

RPI - 45 (Estimate)
20-12
1-25: 1-5 (Virginia a #1 Seed)
1-50 : 2-6
Top 100: 9-8
Bad Losses: 4

NC St (Surprise Pick over SMU -- although I had SMU out, NC ST was in a group of about 4 teams that were all about equal)
21-13
RPI - 53
1-25: 1-7 (Duke)
1-50: 4-9
Top 100: 6-11
Bad Losses: 2

So our resume at 20-11 entering the ACC tourney, and assuming 21-12 after the tourney matches up pretty much with all of these. Not much better or worse. Just basically the same. I guess after doing this I do feel a bit better about our chances if we finish at 21-12. (I was 50/50 before, maybe now I move it to 70-80%)

 

[moqui]

the SU resume is not going to be what RPI forecast says; the forecast is biased by the poor early play and doesn't take into account the steady improvement. The Syracuse team that is playing in January would have gone 7-1 in those early games and would be forecast to be a 24 or 25 win squad.

 

[Cusefan0307]

This idea came from Pearl, who posted the resume the other day of NC ST who was clearly the last team in last year. The final spot came down to SMU (who 90% had in) or a poo-poo platter of teams (which NC St was given the prize), and SMU's bad losses did them in as I expected.

RPIForecast gives us the following resume entering the ACC tourney at 20-11

Record: 20-11
RPI - 47
1-5 vs top 25 (Louisville *)
2-7 vs top 50 (H Louisville, N Iowa)
6-11 vs top 100
2 Sub 100 RPI Losses (BAD LOSSES) (at Michigan, at Pitt*)


* Assumed as it is our highest win or lowest win probability of those games -
Note (I just picked one of the Big 3)

At 19-12 our Resume may be the same assuming we lose that extra game to a team in that 50-100 range. Although it could be a 3rd bad loss (or no top 25 win) Our RPI may be closer to 50-54.

Let's look at the teams that went to Dayton last year, I had to make estimates of RPI as I saw post NCAA number from Nolan:

Xavier:
20-12
RPI - 52
1-25: 1-4
1-50: 4-5
Top 100: 9-10
Bad Losses: 3 - USC, Seton Hall (2)

Iowa
19-12
RPI - 56
1-25: 1-8
Top 50: 3-10
Top 100: 5-12
Bad Losses - None

Tennessee:

RPI - 45 (Estimate)
20-12
1-25: 1-5 (Virginia a #1 Seed)
1-50 : 2-6
Top 100: 9-8
Bad Losses: 4

NC St (Surprise Pick over SMU -- although I had SMU out, NC ST was in a group of about 4 teams that were all about equal)
21-13
RPI - 53
1-25: 1-7 (Duke)
1-50: 4-9
Top 100: 6-11
Bad Losses: 2

So our resume at 20-11 entering the ACC tourney, and assuming 21-12 after the tourney matches up pretty much with all of these. Not much better or worse. Just basically the same. I guess after doing this I do feel a bit better about our chances if we finish at 21-12. (I was 50/50 before, maybe now I move it to 70-80%)

Good Post, but NC State must of had 2 top 25 RPI win's because they beat us in the ACC tournament.

 

[JOC44]

the SU resume is not going to be what RPI forecast says; the forecast is biased by the poor early play and doesn't take into account the steady improvement. The Syracuse team that is playing in January would have gone 7-1 in those early games and would be forecast to be a 24 or 25 win squad.

Yeah, but RPI Forecast also doesn't take into account that McCullough is out for the rest of the season. While he wasn't playing well, we're now down to a rotation of no more than six players who JB, per his substitution pattern over the last month, feels comfortable playing in conference games.

 

[jncuse]

the SU resume is not going to be what RPI forecast says; the forecast is biased by the poor early play and doesn't take into account the steady improvement. The Syracuse team that is playing in January would have gone 7-1 in those early games and would be forecast to be a 24 or 25 win squad.

Yeah, but RPI Forecast also doesn't take into account that McCullough is out for the rest of the season. While he wasn't playing well, we're now down to a rotation of no more than six players who JB, per his substitution pattern over the last month, feels comfortable playing in conference games.

Part of the reason I take the forecast is because its shows where we are at 20-11, which is a common "where would be stand" if we had that record.

 

[moqui]

Yeah, but RPI Forecast also doesn't take into account that McCullough is out for the rest of the season. While he wasn't playing well, we're now down to a rotation of no more than six players who JB, per his substitution pattern over the last month, feels comfortable playing in conference games.

even without Chris, I believe that in any game in which Roc is not in foul trouble, SU is capable of beating everyone this side of Duke & UVA. Of course his absence changes things, but SU played its best full game of the season mostly without him on Sunday night

 

[jncuse]

the SU resume is not going to be what RPI forecast says; the forecast is biased by the poor early play and doesn't take into account the steady improvement. The Syracuse team that is playing in January would have gone 7-1 in those early games and would be forecast to be a 24 or 25 win squad.

I am a little confused as to your post, although I suspect I am just overlooking something. It's predicting a record of 20-11, and showing our resume at that record. It does use our year to date performance to predict our final performance.

If you are saying we are a 24 or 25 win team, no worries we are getting in (I hope you are right!). I am just assessing the situation if we ended up at 21-12.

 

[jncuse]

Good Post, but NC State must of had 2 top 25 RPI win's because they beat us in the ACC tournament.

I guess I didn't fully compare apples and oranges. That NC St record includes the ACC Tourney win over Syracuse.

I am equating 20-11 with 21-12 (one meh win, one loss to a good team in the ACC Tounrey)

So say we were 21-12
Our RPI would about the same
Our Top 25 Record is probably 1-6
And our Top 100 Record may now become 7-12 or 6-12.

 

[moqui]

I am a little confused as to your post, although I suspect I am just overlooking something. It's predicting a record of 20-11, and showing our resume at that record. It does use our year to date performance to predict our final performance.

If you are saying we are a 24 or 25 win team, no worries we are getting in (I hope you are right!). I am just assessing the situation if we ended up at 21-12.

what I'm saying is that the forecast is not dynamic and does not account for change. the Syracuse team playing tonight - even without Chris - is much better than the one that played in November & the first week of December. But the forecast doesn't see that, it just dumps all the data in and treats it all as equal. The reality is that this squad will outperform the forecast because the team that is playing is not shackled to those early performances, but the forecast model is.

 

[jncuse]

what I'm saying is that the forecast is not dynamic and does not account for change. the Syracuse team playing tonight - even without Chris - is much better than the one that played in November & the first week of December. But the forecast doesn't see that, it just dumps all the data in and treats it all as equal. The reality is that this squad will outperform the forecast because the team that is playing is not shackled to those early performances, but the forecast model is.

You are correct in that the model purely predicts the future based on the past. I tend to like to use this because it puts what you have done year to date, in a 30 game package which is easier to assess.

I'm not quite as confident in our ability to easily surpass 11-7 in the ACC as you are, even with our upward climb on the floor. We still have a few recent negatives -- shorter bench going forward and the recent stinker against VTech.

 

[two3zone]

We need top 50 wins. We don't get in with 0.

 

[jordoo]

Did Pitt have any top50 wins last year? They beat no one in the non conference and only beat mediocre and poor teams in conference if memory serves. Its hard to say right now what will get us in and what won't because we will be going against this years teams and their resume's not last years or any other years. There is a lot of junk out there this year . . . . . so much so that I pitty the voters having to come up with the last 5 teams for their top25, that must be brutal.