Atlantic Region

[IUPbigINDIANS]

Re: Atlantic Region

Totally irrelevant now, but since we followed them all year, here are the final Atlantic Region Performance Indicators (Top 15 including conference tournament results):


Note: Quite the contrast from the 'official' bracket/seeding.

1. IUP
2. Virginia State
3. West Liberty
4. East Stroudsburg
5. West Chester
6. Mercyhurst
7. Shippensburg
8. Notre Dame
9. Fairmont State
10. Slippery Rock
11. Concord
12. Charleston
13. Virginia Union
14. Bowie State
15. UPJ

[IUPTrackMan]

Re: Atlantic Region

Follow the money.

You'll quickly see a trend and common theme.

Said for years and years D2 sports aren't an even playing field.

The disparity from the top 4 teams to the bottom 4 teams in the PSAC is about the same as comparing the Yankees and Red Sox to the Pirates and Devil Rays.

[IUPalum]

Re: Atlantic Region

Ok...Then pick other teams if you want.

I'll always pick WL because they love playing cream puffs.

[PSAC76]

Re: Atlantic Region

Totally irrelevant now, but since we followed them all year, here are the final Atlantic Region Performance Indicators (Top 15 including conference tournament results):


Note: Quite the contrast from the 'official' bracket/seeding.

1. IUP
2. Virginia State
3. West Liberty
4. East Stroudsburg
5. West Chester
6. Mercyhurst
7. Shippensburg
8. Notre Dame
9. Fairmont State
10. Slippery Rock
11. Concord
12. Charleston
13. Virginia Union
14. Bowie State
15. UPJ

Not that many listen to anything that I say, but I have heard that in several regions, the committee can slot teams in different spots to avoid conference matchups. For instance, WL might've been the 2 seed outright, but in theory VSU could've been the 2 and they bumped WL up a line to avoid a conference game in the 1st round. Or conversely, ND could've been a different number and they were bumped down. I know that this is done, but it might provide some clarity as to why the PI is a little different than the actual bracket.

[EastStroud13]

Re: Atlantic Region

IUP-Fairmont
WL-WCU
VSU-ESU
Hurst-NDC

Done and done. Of course, this would mean drastically changing the committee's opinion of Fairmont.

[IUPbigINDIANS]

Re: Atlantic Region

IUP-Fairmont
WL-WCU
VSU-ESU
Hurst-NDC

Done and done. Of course, this would mean drastically changing the committee's opinion of Fairmont.

Aside from the Fairmont love fest ... Ship would have beat them by double digits head-to-head.

Not that it matters now ...

The FSU 5-seed is just baffling ... considering ND beat them 2 out of 3 and ND won the MEC title.

Who knows. Maybe they come up here and wreck havoc.

They have a real tough assignment. Thing with Mercyhurst is they don't change no matter who they play. They make other teams change to that triangle and two. Big advantage against anybody let alone a team like FSU that's never seen it.

FSU has to have patience because Mercyhurst will annoy and agitate the heck out of them.

Fairmont is used to seeing 'bone-crushing' defense so it shouldn't be a big deal, .... right?

[IUPalum]

Re: Atlantic Region

Fairmont is used to seeing 'bone-crushing' defense so it shouldn't be a big deal, .... right?

I see what you did there! hahaha

[boatcapt]

Re: Atlantic Region

I've discovered the criteria that the committee uses for most of it's selection and seedings. It's not PI...It's not overall W/L...It's not record vs ranked opponents. It IS W/L vs In Region Opponents. This criteria got seven of the eight selections correct AND got them ranked in the right slot in six of eight. Here how the region stacked up in W/L vs In Region Opponents:

1. IUP 93.33%
2. WLU 86.67%
3. VA State 85.71%
4. Hurst 75.86%
5. Ship 75.86%
6. NDC 75.86%
7. ESU 74.07%
8. Fairmont 73.33%

9. WCU 72.41%
10. Charleston 67.86%

Clearly Ship should have been in with WCU sitting at home right now.

[Columbuseer]

Re: Atlantic Region

I've discovered the criteria that the committee uses for most of it's selection and seedings. It's not PI...It's not overall W/L...It's not record vs ranked opponents. It IS W/L vs In Region Opponents. This criteria got seven of the eight selections correct AND got them ranked in the right slot in six of eight. Here how the region stacked up in W/L vs In Region Opponents:

1. IUP 93.33%
2. WLU 86.67%
3. VA State 85.71%
4. Hurst 75.86%
5. Ship 75.86%
6. NDC 75.86%
7. ESU 74.07%
8. Fairmont 73.33%

9. WCU 72.41%
10. Charleston 67.86%

Clearly Ship should have been in with WCU sitting at home right now.

You may be correct.
But since I got a call from Old Scratch :-) and he asked me to advocate for him, wouldn't educated guesses by fans who follow this region's teams do just as well in picking 7 out of 8 as well as ranking 6 out of 8 correctly using their own personal heuristics?

[boatcapt]

Re: Atlantic Region

You may be correct.
But since I got a call from Old Scratch :-) and he asked me to advocate for him, wouldn't educated guesses by fans who follow this region's teams do just as well in picking 7 out of 8 as well as ranking 6 out of 8 correctly using their own personal heuristics?

Hahahaha! Yes, yes it would!!!

I just find it funny that people cleave to the notion that their is some particular criteria that the committee uses over all others for their selection and seeding. We are told with great authority that THIS particular criteria has been 100% accurate for the last two years so that PROVES that that is the criteria the committee uses. Usually the very next year after the grandeous statement of accuracy, the criteria is proven to be not such a great predictor!

I came to the conclusion long ago that the whole selection process is completely subjective with each committee member using what ever criteria he/she wants to validate the way they want to vote. If they want to have WCU in, there's probably a criteria that says they should be it...If they want to keep Ship out, there's probably a criteria that let's them vote that way.

Now that said, just because I believe the process is totally subjective doesn't mean I believe that is the correct and most valid way of selecting playoff teams. I still believe that a totally objective process that ranks each team against the same exact criteria (formula) each year is the most fair to all teams. Any time you have a subjective system like we have you introduce personal prejudices and opinions into the process and that by definition is not a fair, level playing field for all teams.

[Columbuseer]

Re: Atlantic Region

Maybe someone can come up with a formula, but I would guess it would be complex. Here are some heuristics that some folks might use ( I am not saying they are the correct ones, but are just examples):
1) Goal is to get the teams who are playing the best at the end of the season, but not to the exclusion of their season-long performance.
2) Losses due to injury to key players are not as damaging if that player is back at full strength now.
3) Losses to weak teams late in the season are more damaging
4) Victories against strong opponents are a plus, especially in other conferences because it provides a frame of reference (with a lot of caveats of course)
5) Close losses in tournament to high seeds is not as damaging, especially if the winning team is peaking in performance (i.e., Notre Dame in MEC tourney)
etc.
etc.

I would probably take an AI program to do a better job than human heuristics and judgment. Not sure a pure math formula could do it.

[EastStroud13]

Re: Atlantic Region

That criteria STILL doesn't explain Fairmont State.

As for Ship, I'm guessing the "mystery criteria" is head-to-head. It's not really controversial at all to have them out and WCU in.

[Chuck Norris]

Re: Atlantic Region

Hahahaha! Yes, yes it would!!!

I just find it funny that people cleave to the notion that their is some particular criteria that the committee uses over all others for their selection and seeding. We are told with great authority that THIS particular criteria has been 100% accurate for the last two years so that PROVES that that is the criteria the committee uses. Usually the very next year after the grandeous statement of accuracy, the criteria is proven to be not such a great predictor!

I came to the conclusion long ago that the whole selection process is completely subjective with each committee member using what ever criteria he/she wants to validate the way they want to vote. If they want to have WCU in, there's probably a criteria that says they should be it...If they want to keep Ship out, there's probably a criteria that let's them vote that way.

Now that said, just because I believe the process is totally subjective doesn't mean I believe that is the correct and most valid way of selecting playoff teams. I still believe that a totally objective process that ranks each team against the same exact criteria (formula) each year is the most fair to all teams. Any time you have a subjective system like we have you introduce personal prejudices and opinions into the process and that by definition is not a fair, level playing field for all teams.

[boatcapt]

Re: Atlantic Region

Maybe someone can come up with a formula, but I would guess it would be complex. Here are some heuristics that some folks might use ( I am not saying they are the correct ones, but are just examples):
1) Goal is to get the teams who are playing the best at the end of the season, but not to the exclusion of their season-long performance.
2) Losses due to injury to key players are not as damaging if that player is back at full strength now.
3) Losses to weak teams late in the season are more damaging
4) Victories against strong opponents are a plus, especially in other conferences because it provides a frame of reference (with a lot of caveats of course)
5) Close losses in tournament to high seeds is not as damaging, especially if the winning team is peaking in performance (i.e., Notre Dame in MEC tourney)
etc.
etc.

I would probably take an AI program to do a better job than human heuristics and judgment. Not sure a pure math formula could do it.

I disagree with the need for complexity. While I think PI has some obvious flaws, at least it is consistent and easily applied. Formula using PI, along with something that takes margin of victory AND opponent total points into account and a small "kicker" for W/L over the last 10 games would probably get you pretty close to picking the 8 best teams from a region.

The thing I like in an objective formula based system is the absolute fairness of it. Every team, good and bad...past champion or past door mat would be rated against the same exact formula that is devoid of personal opinion. It would also give coaches a target to shoot at when they are preparing their schedule each year. As it stands right now, scheduling is a total crap shoot...Is it more important to get W's or is it more important to perhaps take a loss to a good team? Is it better to play at home or on the road? Should I schedule a tough game against a top tier but out of region team at our house or would it be better to schedule a bad but in-region team at their house?? Right now with the apparently unequally applied criteria, I would imagine coaches throw their hands up and say 'what ever" more times than we might care to admit.

[boatcapt]

Re: Atlantic Region

But if you have the "human element" thrown in, you really don't have a fair system. I agree there is no perfect formula but at least if you have a set formula that doesn't differ year to year and that is applied equally to all teams with out regard for intentional or unintentional prejudice or opinion, you at least level the playing field for all AND you give coaches a set standard to shoot at.