Which poll is the most accurate playoff predictor?

[Professor Chaos]

So there was a discussion about this in the STATS poll thread this week so I went back to 2013 (when the field expanded to 24 teams and 8 seeds) to compare the final regular AGS polls with the final regular season STATS and Coaches polls to put some data behind which poll reflects the selection committee's decisions the best. I color coded green for a perfect projection, yellow for a projection that was either 1 or 2 spots off, and red for a projection that was 3 or more spots off.

Here's the data:

https://farm2.staticflickr.com/1935/45477803752_6d6e4c9081_b.jpg



https://farm2.staticflickr.com/1964/30587456057_42b613812c_b.jpg



https://farm2.staticflickr.com/1954/45477803662_c9ab3276ee_b.jpg



https://farm2.staticflickr.com/1925/30587455897_6daa6bf291_b.jpg



https://farm2.staticflickr.com/1977/31655909798_3f857803eb_b.jpg



https://farm5.staticflickr.com/4911/45234504474_149891d9dd_b.jpg



https://live.staticflickr.com/65535/49122460278_1ea19999c3_o.jpg



https://live.staticflickr.com/65535/51127555617_21e9eeed95_c.jpg



https://live.staticflickr.com/65535/51698435244_c344b8a681_c.jpg



https://live.staticflickr.com/65535/52515159365_213bdf727e_c.jpg




Here's the summary of it all. The "bad misses" part refers to any projection that was 3 or more spots off:

https://live.staticflickr.com/65535/52514940434_cdc72c8541_z.jpg



There you have it. Digest. Discuss. Debate.

[BisonTru]

This is good work, PC.

[Daytripper]

Well done.

[Reign of Terrier]

I feel like one could extrapolate the biases or capriciousness of the committee with this data.

I'm more concerned about at large consideration than seeding, because I think seeding is more deductive with SOS.

Good stuff as always.

Sent from my Pixel 2 using Tapatalk

[BisonTru]

One quick thing I noticed is obviously last year none of the polls where close on the at larges and they all were quite different as was the committee, but from 2013-2016 no team missed by AGS was in either the Stats or the Coaches poll either.

[ElCid]

Color coding! A man who knows what he is doing! A spreadsheet/data without color is like toast without butter. Sure I'll eat it dry, but it tastes better with butter.

[Redbird 4th & short]

very nice work ... and not surprising outcome. Would love to see how well each polls ranking predicted playoff winners.

[World]

Really well done

thanks!

[ursus arctos horribilis]

That's pretty damn good PC. I want butter on my toast as well if someone is taking orders since we're talking about it. Eggs too please.

[Go...gate]

That's pretty damn good PC. I want butter on my toast as well if someone is taking orders since we're talking about it. Eggs too please.

Don't forget the orange juice and coffee!

[gofurman]

One guy that I liked last year was nobowls.com - reason being I wondered if Furman would get selected and I think he correctly said we would when many said we would not. *His whole thing is not being a poll but picking the FCS field. It's a different take ... Not are you ranked 20th or 22nd but will you make the playoffs*

http://www.nobowls.com

"Since its inception in 2008 NoBowls has correctly predicted 169 of the 180 teams to participate in the postseason "

[mvemjsunpx]

Great job putting this together. It's clear the AGS Poll is closer to the playoff selections than the other polls.

The question then becomes, "is that necessarily a good thing?" I mean, the goal of the poll isn't to predict the committee's selections, it's to rank the best teams. I do think the AGS Poll is the best of the three, but the committee usually has a few inexplicable selections/seedings each year. That means being too close to the committee's bracket could be a bad thing. Ideally, the committee would base their selections primarily on the AGS poll (or my own personal ballot xsmiley_wix) instead of the other two polls or the weird "Simple Ratings System", but I really doubt they do that.

I guess, to put it another way… the main goal of the poll shouldn't be to predict the playoff field, but to influence it.

[Professor Chaos]

One guy that I liked last year was nobowls.com - reason being I wondered if Furman would get selected and I think he correctly said we would when many said we would not. *His whole thing is not being a poll but picking the FCS field. It's a different take ... Not are you ranked 20th or 22nd but will you make the playoffs*

http://www.nobowls.com

"Since its inception in 2008 NoBowls has correctly predicted 169 of the 180 teams to participate in the postseason "
That's pretty good missing on just 11 teams in 10 years. I was trying to find their final projection from last year to see how they did since the polls kinda got their butts kicked projecting at-larges last year but it appears to be lost to the internet black hole of time.

One critique I have of that guy that is that he sometimes ignores the regionalization restrictions to maximize bus trips that the committee is bound to. In the end that's not nearly as important as projecting the right seeds and the right teams in the field though so if he projects those right, which he appears to be doing pretty well, it's not a big deal.

Where I like the poll playoff projection model is it gives everyone a sense for how comfortably in each team is, whether it's for a seed or an at-large invite, as of right now. I prefer looking at it as a snapshot in time (in other words "this is what the playoff field would look like if the playoffs started this weekend") rather than projecting what the field will look like at the end of the season if everything goes to plan (because it never does go to plan). That way you can have a pretty good idea of whether your team needs to kick it up a gear (or needs some help) from that point on in the season to get where they want to go or just maintain.

[Professor Chaos]

Great job putting this together. It's clear the AGS Poll is closer to the playoff selections than the other polls.

The question then becomes, "is that necessarily a good thing?" I mean, the goal of the poll isn't to predict the committee's selections, it's to rank the best teams. I do think the AGS Poll is the best of the three, but the committee usually has a few inexplicable selections/seedings each year. That means being too close to the committee's bracket could be a bad thing. Ideally, the committee would base their selections primarily on the AGS poll (or my own personal ballot xsmiley_wix) instead of the other two polls or the weird "Simple Ratings System", but I really doubt they do that.

I guess, to put it another way… the main goal of the poll shouldn't be to predict the playoff field, but to influence it.
I don't think any poll influences the playoff field unless you want to claim it's subconsciously affecting it by being in the backs of the committee members' minds. Even the SRS is just a tool they use to quantify SOS not a tool to directly influence their rankings (although by that fact it probably has an effect by proxy more than any other poll or ranking system).

I don't necessarily agree that ranking the best teams and projecting the playoff field are mutually exclusive though. It is true that the committee is bound to certain constraints that poll voters aren't (mainly the "having less than 6 D1 wins may give a team less consideration for an at large invite" part) but, as I understand it, how the committee eventually sets the field is they have an anonymous vote where everyone ranks the teams in consideration for seeds and at-large invites. So the merit based ranking methodology is the same in both cases IMO.

[Dukie95]

In 2016 seeds, I don't think I'd give AGS a miss for the 3rd place tie.

[Professor Chaos]

In 2016 seeds, I don't think I'd give AGS a miss for the 3rd place tie.
Yeah, that was a tough decision. I was trying to be unbiased about it. It would probably be more accurate to give JSU and JMU each a half seed line miss. Regardless the AGS poll massively outperformed the other two when it came to the 2016 seed lines.

[Professor Chaos]

Great work!

Not attacking or defending anything or expressing an opinion, really; just wondering...

It seems to me that AGS Voters, of which I am one, probably ‘discount’ the non-Playoff Conferences (Ivy...and more recently, ‘kinda/sorta,’ the MEAC & SWAC) more than other polls might. If this premise is invalid, then no reason to read further, but my (pure) speculation is that if we analyzed just the various polls using some methodology, we might find it to be the case.

Wondering, if it would make sense / be worthwhile to ‘strip-out’ the data (polls AND Playoffs, where applicable) associated with any of those Teams just to see how it would look with that possible ‘hitch’ excluded.
I did exclude playoff ineligible teams from my lists. That includes the Ivies and the MEAC and SWAC champs (MEAC champ only from 2015 on) so the only schools that show up in these lists are the non-Celebration Bowl MEAC teams. I also excluded any autobids from my at-large list so that's why you see gaps in the rankings.

It is accurate to say that AGS seems to generally rank Ivy, MEAC, and SWAC teams lower than the STATS or Coaches polls do. A good example was in 2016 when NC A&T got an at-large bid. According to the AGS poll they would've been the last team in whereas both the STATS and the Coaches polls had them in in front 6 others.

[FUBeAR]

I did exclude playoff ineligible teams from my lists. That includes the Ivies and the MEAC and SWAC champs (MEAC champ only from 2015 on) so the only schools that show up in these lists are the non-Celebration Bowl MEAC teams. I also excluded any autobids from my at-large list so that's why you see gaps in the rankings.

It is accurate to say that AGS seems to generally rank Ivy, MEAC, and SWAC teams lower than the STATS or Coaches polls do. A good example was in 2016 when NC A&T got an at-large bid. According to the AGS poll they would've been the last team in whereas both the STATS and the Coaches polls had them in in front 6 others.yeah - I went back and looked at the data & figured that you did...which is why I deleted my post.

my bad...sorry....nevermind.

[ursus arctos horribilis]

Great job putting this together. It's clear the AGS Poll is closer to the playoff selections than the other polls.

The question then becomes, "is that necessarily a good thing?" I mean, the goal of the poll isn't to predict the committee's selections, it's to rank the best teams. I do think the AGS Poll is the best of the three, but the committee usually has a few inexplicable selections/seedings each year. That means being too close to the committee's bracket could be a bad thing. Ideally, the committee would base their selections primarily on the AGS poll (or my own personal ballot xsmiley_wix) instead of the other two polls or the weird "Simple Ratings System", but I really doubt they do that.

I guess, to put it another way… the main goal of the poll shouldn't be to predict the playoff field, but to influence it.

The comittee does miss a team that we predict each year. We don't try to predict the selection committee picks...we just do it by virtue of paying as close or even closer attention to the teams that deserve to be in.

You've said this before and I am not sure how to understand your mixed up internal feelings on this matter? We do what we do before they do what they do...so don't know how or why you'd think it has anything to do with lining up with them?

When they select some crazy team that ain't on our list that is on them. I think we normally have the best representation of where teams actually should be. They miss one or two sometimes as I've said already and that is reflected in the data that PC put out.

[ursus arctos horribilis]

One guy that I liked last year was nobowls.com - reason being I wondered if Furman would get selected and I think he correctly said we would when many said we would not. *His whole thing is not being a poll but picking the FCS field. It's a different take ... Not are you ranked 20th or 22nd but will you make the playoffs*

http://www.nobowls.com

"Since its inception in 2008 NoBowls has correctly predicted 169 of the 180 teams to participate in the postseason "

NoBowls does a good job. He/they follow us and like the AGS Poll frequently on twitter and I look at that site quite a bit also.

[ST_Lawson]

NoBowls does a good job. He/they follow us and like the AGS Poll frequently on twitter and I look at that site quite a bit also.

Agreed...it's not perfect (nobody is), but it's pretty solid and a hell of a lot better than whatever that crap is that gets pinched out and called the Coaches Poll every week.
He/they tweet and retweet some good (and sometimes funny) stuff throughout the season too...it's a good twitter follow.

[ursus arctos horribilis]

Agreed...it's not perfect (nobody is), but it's pretty solid and a hell of a lot better than whatever that crap is that gets pinched out and called the Coaches Poll every week.
He/they tweet and retweet some good (and sometimes funny) stuff throughout the season too...it's a good twitter follow.

Agreed, been following NoBowls for quite a while now via @AGSFCS and like seeing the stuff that comes across the deck from them.

[TheValleyRaider]

The Coaches poll being as off as they are is a bit surprising to me, given its voters (coaches and/or SIDs) are presumably closer (personally and professionally) to the ADs who actually choose the playoff field.

None of them are particularly far off over the long term though, even if AGS is consistently closer

[Daytripper]

The Coaches poll being as off as they are is a bit surprising to me, given its voters (coaches and/or SIDs) are presumably closer (personally and professionally) to the ADs who actually choose the playoff field.

None of them are particularly far off over the long term though, even if AGS is consistently closer

Reputation, I think, plays a much greater role in the Coach's poll than in others.

[Redbird 4th & short]

The Coaches poll being as off as they are is a bit surprising to me, given its voters (coaches and/or SIDs) are presumably closer (personally and professionally) to the ADs who actually choose the playoff field.

None of them are particularly far off over the long term though, even if AGS is consistently closer
they are too busy and otherwise lazy to look into anything, other than what is most important to their team each week. It is a flawed concept for a poll. They need to introduce a Strength of Schedule Index and put it right next to these polls. So when everyone sees a 9-2 MEAC team ranked 5th and having played the 96th ranked SOS despite having played 2 FBS teams .. guess who everyone .... they'll have some frame of reference for saying, yeah but ... as in yeah, they're 9-2, but they play a very soft schedule compared to some 7-4 team that plays the 3rd ranked SOS, including 6 top 25 teams.

[Mocs123]

Did I somehow miss Chattanooga in the 2015 playoff teams? We beat Fordham in the first round and lost to Jacksonville State in OT in the second round.

Not that this is what this post is about, I just was looking at Chattanooga's rank in each of the polls.

[WestCoastAggie]

they are too busy and otherwise lazy to look into anything, other than what is most important to their team each week. It is a flawed concept for a poll. They need to introduce a Strength of Schedule Index and put it right next to these polls. So when everyone sees a 9-2 MEAC team ranked 5th and having played the 96th ranked SOS despite having played 2 FBS teams .. guess who everyone .... they'll have some frame of reference for saying, yeah but ... as in yeah, they're 9-2, but they play a very soft schedule compared to some 7-4 team that plays the 3rd ranked SOS, including 6 top 25 teams.

Thank goodness you aren't anywhere near the playoff committee boardroom.

[Professor Chaos]

Did I somehow miss Chattanooga in the 2015 playoff teams? We beat Fordham in the first round and lost to Jacksonville State in OT in the second round.

Not that this is what this post is about, I just was looking at Chattanooga's rank in each of the polls.
They were the SOCON auto in 2015. If they weren't seeded (or ranked above a team that was seeded) I didn't include any autos in the list. Just would've cluttered it up when the point was to see how well the polls were at projecting the at-large selections.

[Mocs123]

They were the SOCON auto in 2015. If they weren't seeded (or ranked above a team that was seeded) I didn't include any autos in the list. Just would've cluttered it up when the point was to see how well the polls were at projecting the at-large selections.

Thanks

This is an interesting look at the polls

[Redbird 4th & short]

Thank goodness you aren't anywhere near the playoff committee boardroom.

I'd give you guys a fair shot xsmhx

[mvemjsunpx]

The comittee does miss a team that we predict each year. We don't try to predict the selection committee picks...we just do it by virtue of paying as close or even closer attention to the teams that deserve to be in.

You've said this before and I am not sure how to understand your mixed up internal feelings on this matter? We do what we do before they do what they do...so don't know how or why you'd think it has anything to do with lining up with them?

When they select some crazy team that ain't on our list that is on them. I think we normally have the best representation of where teams actually should be. They miss one or two sometimes as I've said already and that is reflected in the data that PC put out.

I'm not really as ambivalent as I made it sound. It's just that I've seen posters on here occasionally "brag" about how close the AGS Poll is to the final playoff field. I guess the concern is that voters might—even subconsciously— turn their vote into an attempt to predict what the committee's gonna pick rather than ranking how good the teams truly are. I don't really think that's a problem right now, but it's something voters should at least be cautious of.

[cx500d]

I'm not really as ambivalent as I made it sound. It's just that I've seen posters on here occasionally "brag" about how close the AGS Poll is to the final playoff field. I guess the concern is that voters might—even subconsciously— turn their vote into an attempt to predict what the committee's gonna pick rather than ranking how good the teams truly are. I don't really think that's a problem right now, but it's something voters should at least be cautious of.


That may be true...I know I'm gonna have UNH on mine...

[ursus arctos horribilis]

I'm not really as ambivalent as I made it sound. It's just that I've seen posters on here occasionally "brag" about how close the AGS Poll is to the final playoff field. I guess the concern is that voters might—even subconsciously— turn their vote into an attempt to predict what the committee's gonna pick rather than ranking how good the teams truly are. I don't really think that's a problem right now, but it's something voters should at least be cautious of.

Well, I can see that type of a concern I guess but it isn't based in reality from what I've seen and as you probably know I see every vote every week and if I ever saw someone suddenly moving things around without any real reason for doing so in the last week I would probably send an email asking why. It just doesn't happen. What does happen is some of us do like that what we see and do with our overall poll is reflected for the most part by others that should be more in the know than anyone else out there since they are a group selected by the NCAA to put together the playoff grouping.

We all put in a serious effort and take a serious pride in what we do even if it just our hobby. xthumbsupx

[mvemjsunpx]

Well, I can see that type of a concern I guess but it isn't based in reality from what I've seen and as you probably know I see every vote every week and if I ever saw someone suddenly moving things around without any real reason for doing so in the last week I would probably send an email asking why. It just doesn't happen. What does happen is some of us do like that what we see and do with our overall poll is reflected for the most part by others that should be more in the know than anyone else out there since they are a group selected by the NCAA to put together the playoff grouping.

We all put in a serious effort and take a serious pride in what we do even if it just our hobby. xthumbsupx

Oh, I know most all of us take the vote seriously (I know I do). This isn't something I'm real worried about in the short term. It could be more of a concern if, say, the AGS Poll ends up becoming a bigger deal (e.g., the STATS Poll goes defunct), and there's a big surge in voters or something.

[Professor Chaos]

Here's how 2018 broke down.

https://farm5.staticflickr.com/4911/45234504474_149891d9dd_b.jpg


And here's the updated summary.

https://farm5.staticflickr.com/4838/45909453862_2f450e4217.jpg

The Coach's poll (the "official" poll in the toolbox of the selection committee) was definitely the 3rd best in terms of reflecting what the selection committee did. I'd say it was about a draw between the AGS and STATS polls in terms of which reflected the committee the best. The AGS Poll did much better reflecting the seeds and the STATS poll did much better reflecting the at-large teams. Overall though for the 6 years of the 24 team playoff format I think the AGS poll is further distancing itself as a better reflection of what the selection committee will do. Something pretty remarkable, in my opinion at least, is that the AGS Poll hasn't been further than 2 spots away on any seeded playoff team since 2013 (that's a streak of 5 years and counting).

In another thread there was a comment that a poll shouldn't strive to reflect what the selection committee does and I agree to a certain extent and I don't think most AGS poll voters vote with the "what is the selection committee going to do?" in the back of their minds but what I think this goes further to prove is that having an open forum and debate to criticize a poll leads to a better product at the end of the year. Of these 3 polls only the AGS Poll has that and I think it shows when it comes to which poll reflects the selection committee (who obviously comes to their decision via criticism and open forum debate) the best.

[kalm]

Here's how 2018 broke down.

https://farm5.staticflickr.com/4911/45234504474_149891d9dd_b.jpg


And here's the updated summary.

https://farm5.staticflickr.com/4838/45909453862_2f450e4217.jpg

The Coach's poll (the "official" poll in the toolbox of the selection committee) was definitely the 3rd best in terms of reflecting what the selection committee did. I'd say it was about a draw between the AGS and STATS polls in terms of which reflected the committee the best. The AGS Poll did much better reflecting the seeds and the STATS poll did much better reflecting the at-large teams. Overall though for the 6 years of the 24 team playoff format I think the AGS poll is further distancing itself as a better reflection of what the selection committee will do. Something pretty remarkable, in my opinion at least, is that the AGS Poll hasn't been further than 2 spots away on any seeded playoff team since 2013 (that's a streak of 5 years and counting).

In another thread there was a comment that a poll shouldn't strive to reflect what the selection committee does and I agree to a certain extent and I don't think most AGS poll voters vote with the "what is the selection committee going to do?" in the back of their minds but what I think this goes further to prove is that having an open forum and debate to criticize a poll leads to a better product at the end of the year. Of these 3 polls only the AGS Poll has that and I think it shows when it comes to which poll reflects the selection committee (who obviously comes to their decision via criticism and open forum debate) the best.

Great work here, PC!

Spot on especially with that last paragraph.

[Professor Chaos]

Here's the 2019 breakdown along with the updated summary:

https://live.staticflickr.com/65535/49122460278_1ea19999c3_o.jpg


https://live.staticflickr.com/65535/49123095687_3f455ae7ef_o.jpg


All 3 polls did well with the at-larges but all 3 struggled on the seeds. This was the first time the AGS Poll has been more than 2 seeds lines off any seed since 2013. I still have my doubts about the Coaches Poll having UCA seeded since that poll didn't release until this morning and UCA (and UNI for that matter) somehow jumped Kennesaw St and we all know how much the Coaches Poll loves Kennesaw St and how much they loathe moving teams up above other teams that didn't lose the previous week.

I'll call it a draw between the 3 polls this year although STATS may have the tiniest of edges since they were cumulatively one seed line closer than the AGS Poll was.

[MacThor]

If ursus would just let me submit my poll on Sunday afternoon, it would be balls on accurate.

[Bisonoline]

Thats pretty impressive.

[Redbird 4th & short]

impressive but only find it mildly surprising. I would have guessed AGS has been the best, but maybe not by this much in terms of "bad" picks.

[Professor Chaos]

Here's how the polls stacked up with the playoff selection committee with this spring:

https://live.staticflickr.com/65535/51127555617_21e9eeed95_c.jpg


And the updated totals (a "bad miss" is something that was more than 2 spots off):
https://live.staticflickr.com/65535/51127555607_b11a47ffac.jpg


I'd say it was pretty close but the AGS Poll takes the bragging rights for this year slightly over the STATS Poll with the Coaches Poll lagging in third. It's also worth pointing out that only the AGS Poll this year was released before the selection shot... the STATS Poll was released yesterday (and JSU made a curious one spot rise over EWU even though they were both idle last week) and the Coaches Poll was released today. Overall though the bad misses and seed alignment with AGS Poll is much better than the other two since I started tracking this in 2013.

[HensRock]

That may be true...I know I'm gonna have UNH on mine...

Thank you cx!
I just spit my coffee all over myself!

That was a good one right there.

[geaux_sioux]

Fun to look at. Nice work! Basically confirms what we all knew already. AGS is the gold standard.

[FCS_pwns_FBS]

Fan polls, they work in practice but not in theory. xnodx

[Professor Chaos]

Fan polls, they work in practice but not in theory. xnodx
Yeah, if we knew that AGS Poll had a negligible influence on the selection committee how impartial can you expect fans of teams on the playoff bubble or in seeding territory to be?

The value I find in showing the AGS Poll's superiority over the other polls in aligning with the selection committee is it shows where to get the best gauge earlier in the season as to where teams sit either in the seeding discussion or in terms of the at-large bubble.

[TheKingpin28]

Yeah, if we knew that AGS Poll had a negligible influence on the selection committee how impartial can you expect fans of teams on the playoff bubble or in seeding territory to be?

The value I find in showing the AGS Poll's superiority over the other polls in aligning with the selection committee is it shows where to get the best gauge earlier in the season as to where teams sit either in the seeding discussion or in terms of the at-large bubble.

Yep. The reason why we are accurate is we watch the games, we debate the games, and we are not afraid to be challenged by our beliefs on teams. The media cannot be bothered to do 2 of the 3 if not 3 of the 3. Also, if slot voting is occurring, we call it out real quick and ask people to defend their position as to why they do it.

[Gil Dobie]

3 out of 4 in the final four, Delaware was #5.

[Chalupa Batman]

3 out of 4 in the final four, Delaware was #5.

Nice. It also correctly predicted 7 of 8 first round matchups and all 4 quarterfinal matchups.

[FUBeAR]

3 out of 4 in the final four, Delaware was #5.

FUBeAR had the Hens right, but fell victim to the Weber hype, pushing the Dukes 1 spot out of the Top 4 & ruining a perfect semi-finals boxed superfecta…

Hello FUBeAR,

We have received your AGS Top 25 vote on 4/17/2021 22:44:10

Your vote is listed below.

1: Sam Houston State Bearkats
2: Weber State Wildcats
3: South Dakota State Jackrabbits
4: Delaware Fightin' Blue Hens
5: James Madison Dukes

[Professor Chaos]

My final regular season poll ballot:

1: Sam Houston State Bearkats
2: South Dakota State Jackrabbits
3: Delaware Fightin' Blue Hens
4: James Madison Dukes

Even got the matchups and homes teams right (just switched around the seeds). Should I do some victory laps a la caribbeanhen???

[FUBeAR]

My final regular season poll ballot:

1: Sam Houston State Bearkats
2: South Dakota State Jackrabbits
3: Delaware Fightin' Blue Hens
4: James Madison Dukes

Even got the matchups and homes teams right (just switched around the seeds). Should I do some victory laps a la caribbeanhen???Yes. Yes, you should.

That’s a straight up superfecta, which would have paid you out at $9,456.40 for a $1 bet at Churchill Downs this past weekend.

https://thumbs.gfycat.com/CompassionatePeacefulAmericanriverotter-max-1mb.gif

[Chalupa Batman]

My final regular season poll ballot:

1: Sam Houston State Bearkats
2: South Dakota State Jackrabbits
3: Delaware Fightin' Blue Hens
4: James Madison Dukes

Even got the matchups and homes teams right (just switched around the seeds). Should I do some victory laps a la caribbeanhen???

Ditto.

1: Sam Houston State Bearkats
2: South Dakota State Jackrabbits
3: Delaware Fightin' Blue Hens
4: James Madison Dukes

https://media2.giphy.com/media/l0ErFafpUCQTQFMSk/giphy.gif

[Professor Chaos]

Yes. Yes, you should.

That’s a straight up superfecta, which would have paid you out at $9,456.40 for a $1 bet at Churchill Downs this past weekend.

https://thumbs.gfycat.com/CompassionatePeacefulAmericanriverotter-max-1mb.gif
Might just be evidence that even a blind mouse finds a cookie every once in a while. xlolx

[ursus arctos horribilis]

My final regular season poll ballot:

1: Sam Houston State Bearkats
2: South Dakota State Jackrabbits
3: Delaware Fightin' Blue Hens
4: James Madison Dukes

Even got the matchups and homes teams right (just switched around the seeds). Should I do some victory laps a la caribbeanhen???

That dude gets a phone number right he does a victory lap so I'd go for it if I were you.

[caribbeanhen]

My final regular season poll ballot:

1: Sam Houston State Bearkats
2: South Dakota State Jackrabbits
3: Delaware Fightin' Blue Hens
4: James Madison Dukes

Even got the matchups and homes teams right (just switched around the seeds). Should I do some victory laps a la caribbeanhen???

I approve of victory laps all day every day Professor

if you had not posted that I would’ve never known, and its a feat very worthy of knowing

[caribbeanhen]

That dude gets a phone number right he does a victory lap so I'd go for it if I were you.

I’m glad you noticed, Ive been hot this year. Gotta piss everybody off before I cool off

[caribbeanhen]

Ditto.

1: Sam Houston State Bearkats
2: South Dakota State Jackrabbits
3: Delaware Fightin' Blue Hens
4: James Madison Dukes

https://media2.giphy.com/media/l0ErFafpUCQTQFMSk/giphy.gif

Hello caribbeanhen,

We have received your AGS Top 25 vote on 4/17/2021 20:44:33

Your vote is listed below.


1: Sam Houston State Bearkats
2: South Dakota State Jackrabbits
3: Weber State Wildcats
4: Delaware Fightin' Blue Hens
5: James Madison Dukes



I’m pretty sure I cursed out Weber St on another thread somewhere

Should’ve known better to go with a big fluffy team

[Chalupa Batman]

Nice. It also correctly predicted 7 of 8 first round matchups and all 4 quarterfinal matchups.

With the Bearkats victory yesterday, the AGS poll correctly predicted the winner in every game but one in the playoffs. Only one fourth down play from being perfect as the only loss was Weber States last second loss to SIU in the 1st round.

[Professor Chaos]

Updated for the 2021 fall playoffs.

https://live.staticflickr.com/65535/51698435244_c344b8a681_c.jpg


So, despite the fact that the STATS and Coaches polls didn't release until today AGS blew them both away this year when it came to the seeds. I find it a bit fishy that UNI jumped EKU in the coaches poll to get into that last at-large spot just for beating WIU considering the committee's selection of UNI as an at-large could've impacted voting.

Here's the updated summary:

https://live.staticflickr.com/65535/51698628005_ce9e096e7f.jpg

When it comes to predicting who the seeded teams are AGS drew even with the Coaches Poll who had somehow led in that category until now. Other than that the AGS Poll has the lead in every other metric (although the predicted at-large teams are very close).

[Chalupa Batman]

Great work as always PC!

[ursus arctos horribilis]

I always love seeing these. It's pretty cool to get a look at the overall.

[Professor Chaos]

Here's the 2022 edition:

https://live.staticflickr.com/65535/52515159365_213bdf727e_c.jpg



I guess the STATS Poll takes the prize this year for most closely matching the selection committee although both they and the Coaches Poll didn't release until this morning so put a big 'ol asterisk on that. :)


Here's the summary updated with this year's numbers:
https://live.staticflickr.com/65535/52514940434_cdc72c8541_z.jpg

[kdinva]

Great job and thanks for all that effort.

[ngineer]

What an excellent job done by AGS. It shows what happens when people who closely watch a lot of the teams and knows what they're talking about. We all know the coaches' poll is crap because the the SIDs do most of the voting. Coaches don't have the time to know what's going on in other conferences besides their own, and the few crossover games played. While the few media who "seriously" follow FCS level football may have broader knowledge of the landscape, they still don't have the time to drill down and see what is clicking in the various locales. What I find very nice, is that despite almost everyone on this Board having a 'dog' in the fight, we have shown a great amount of objectivity when evaluating teams. Congratulations voters!

[OhioHen]

Only the AGS poll is predictive as the others are published after the field is announced and some reactive voting can take place.

[Chalupa Batman]

Great work as always PC! One note, it looks like AGS & the Coaches Poll should only have 1 "bad miss" each instead of 2.

[Professor Chaos]

Great work as always PC! One note, it looks like AGS & the Coaches Poll should only have 1 "bad miss" each instead of 2.
I counted it both ways. AGS had a bad miss having Mercer in and another bad miss having Delaware out. Had Mercer been one of the final 2 in or Delaware been one of the first 2 out it would've only been one bad miss.

It's a pretty arbitrary measure since I just picked anything more than 2 spots away as being "bad" but it was a way to illustrate which polls have more of a tendency to miss by quite a bit *cough*COACHESPOLL*cough*.

[FUBeAR]

I counted it both ways. AGS had a bad miss having Mercer in and another bad miss having Delaware out. Had Mercer been one of the final 2 in or Delaware been one of the first 2 out it would've only been one bad miss.

It's a pretty arbitrary measure since I just picked anything more than 2 spots away as being "bad" but it was a way to illustrate which polls have more of a tendency to miss by quite a bit *cough*COACHESPOLL*cough*.
AGS had it right. It was the Committee that ‘missed.’

[Chalupa Batman]

I counted it both ways. AGS had a bad miss having Mercer in and another bad miss having Delaware out. Had Mercer been one of the final 2 in or Delaware been one of the first 2 out it would've only been one bad miss.

It's a pretty arbitrary measure since I just picked anything more than 2 spots away as being "bad" but it was a way to illustrate which polls have more of a tendency to miss by quite a bit *cough*COACHESPOLL*cough*.

That makes sense.

[caribbeanhen]

That's pretty damn good PC. I want butter on my toast as well if someone is taking orders since we're talking about it. Eggs too please.

I like my toast burnt on one side...

Would be great to see the Coaches and the Stats voters look at the Excellent work that Professor Chaos has compiled here

[ursus arctos horribilis]

I like my toast burnt on one side...

Would be great to see the Coaches and the Stats voters look at the Excellent work that Professor Chaos has compiled here

Pretty easy to see, it is not behind a pay wall.

[Chalupa Batman]

Pretty easy to see, it is not behind a pay wall.

Only way to get them see it…

https://media3.giphy.com/media/UMa2NDPK5oXFm/giphy.gif

[Professor Chaos]

Here's this year's addition:

https://live.staticflickr.com/65535/53344885593_a6c16b1c9f_c.jpg

AGS Poll is the only one to get Richmond projected as an at-large. All 3 polls had a miss for NDSU being seeded. The Coaches Poll broke it's own record (since I've been tracking this) in terms of tthe worst seed line miss with Albany being 9 spots off. If you can believe it they would've been more off I think had they not waited until the day after to release their poll since Albany moved from #23 to #15 ahead of multiple teams in front of them that also won and everyone knows that's not how the Coaches Poll typically works. Projecting Incarnate Word as the 5th at-large into the field is also the worst miss any poll has had on putting an at-large into the field since I started tracking this.

I'd say AGS took the prize this year for most closely reflecting the selection committee with the STATS Poll a close second and the Coaches Poll a very distant third. And, of course, AGS was the only poll of these three to release before the selection show.


Here's the summary going back to 2013 when the 8 seed/24 team format started:

https://live.staticflickr.com/65535/53344653681_8ef8c585e8_z.jpg


A bad miss is anything that's more than 2 spots off. AGS widens or maintains its lead in all the metrics listed here except for somehow the Coaches Poll has been able to collectively get one more seed accurately projected in these last 11 playoffs.

[Preferred Walk-On]

I do enjoy this, especially when AGS's poll was out BEFORE the selections. Not saying there was any malfeasance with the others. Just that AGS has the receipts for being before the show.

Now I wonder how many committee members are actual AGS members/voters? ;)

[Professor Chaos]

I do enjoy this, especially when AGS's poll was out BEFORE the selections. Not saying there was any malfeasance with the others. Just that AGS has the receipts for being before the show.

Now I wonder how many committee members are actual AGS members/voters? ;)
Absolutely - I mentioned it in my post above but Albany moving from #23 to #15 in the coaches poll for a 41-0 win over 4-7 Monmouth is very curious since they jumped over SIU (who also won convincingly over a bad team), Lafayette (who won comfortably in a rivalry game), and Chattanooga (who lost to Bama). That never happens in that poll - so I'm assuming at least some of their voters were influenced by the committee seeding Albany at #5 or they would've been even more substantially off on that one.

It's also pretty incredible how bad of a miss UIW was for them because I'm pretty sure they weren't even listed in the first four out on the selection show (I believe that was UTM, UC Davis, WCU, and Holy Cross) and the Coaches Poll would've had them as the 5th at-large team in.

It's also pretty convenient how Furman landed right at #7 in the STATS poll released this morning after that awful loss they took last Saturday...

[ursus arctos horribilis]

Well done as always to the group, PC, PWO, etc. for all the help with this endeavor each year.

[TheKingpin28]

As usual, not one other group or poll comes close to this group. Excellent work everyone.

Sent from my SM-A546U1 using Tapatalk