Topic: Ranking of Gold Qualifiers Revised and Completed (Read 4971 times)

Jaggers27 · « **on:** March 05, 2004, 01:59:54 am »

With all the tabs in and my midterms finally done with (

) , here are the final rankings for all the teams that qualified for gold out of their regional:

1) Macalaster (576)
2) Furman (585)
3) Yale (928)
4) Weber State (308)
5) Notre Dame (926)
6) UC Berkley (662)
7) Miami OH (992)
8) Notre Dame (925)
9) Minnesota-TC (501)
10) Furman (586)
11) Penn State-UP (828)
12) USC (345)
13) Geo Washington (757)
14) Drake (489)
15) Rhodes (872)
16) UAB   (317)
17) Miami (OH) (996)
18) UC Berkley (663)
19) Georgia Tech (945)
20) Boston (572)
21) Howard (684)
22) Georgia Tech (944)
23) Holy Cross (657)
24) Furman (584)
25) Notre Dame (927)
26) Weber State (307)
27) Miami OH (993)
28) Dartmouth (472)
29) UCLA   (965)
30) Lewis (888)
31) Rhodes (873)
32) Chicago (697)
33) Kansas (864)
34) Columbia (514)
35) USC (346)
36) Miami OH (995)
37) Iowa   (614)
38) Harvard (797)
39) Tenn-Knoxville (440)
40) Iowa   (613)
41) Columbia (515)
42) St. Thomas (552)
43) Illinois State (980)
44) Wisconsin (784)
45) Florida A&M (532)
46) Virginia (972)
47) UNC   (526)
48) UCLA   (964)
49) Washington (633)
50) Maryland-CP (668)
51) Virginia (973)
52) Notre Dame (924)
53) Brown (957)
54) Rhodes (874)
55) Cincinnati (645)
56) Duke   (304)
57) Miami OH (994)
58) Vanderbilt (350)
59) Michigan (524)
60) Chicago (698)
61) Bellarmine (852)

For the complete results click on the link below:
http://www.nd.edu/~manders8/Comprehensive%20AMTA%20Team%20Rankings.xls

Jaggers27 · « **Reply #1 on:** March 05, 2004, 02:05:41 am »

Again, the above list was computed using a method similar to the one described in the thread "8-0 ranks". One major revision to the entire formula was the way in which bonus bid status was assigned by team. Instead of participating in the nebulous and at times arbitrary process of matching specific teams with bid points, I lumped all of a school's bonus bid points together, and doled them out proportionately by regional. A school's own bonus bid points count only 1/4 of their amount to the regional strength for that team. A bonus point system was added to reward teams for winning their regional. the amount of bonus points is porportionate to the stregth of the regional, specifically they add 1/500 of the overall strength of the region to the total for a team. As always, should you have any questions, comments, oir suggestions, please feel free to share them.

Jaggers27 · « **Reply #2 on:** March 05, 2004, 02:14:59 am »

Finally, an ordinal ranking of regional strength:

1) College Park Regional: 266.38 points
2) Macalaster Regional: 235.08 points
3) St. Louis Regional: 208 points
4) Princeton Regional: 194.29 points
5) Marquette Regional: 193.75 points
6) Peoria Regional: 193.33 points
7) Furman Regional: 185 points
8) Samford Regional: 175 points
9) Notre Dame Regional: 173.25 points
10) Lewis Regional: 157 points
11) Tallahasse Regional:150.75 points
12) Washburn Regional: 145.75 points
13) Manchester Regional: 137.25 points
14) Weber State Regional: 135.92 points
15) St. John's Regional: 132.42 points
16) Indiana, PA Regional: 124.33 points
17) Miami OH Regional: 109.42 points
18) Portland Regional: 97 points
19) So. California Regional: 87 points

Complete results of Regional findings can be found here:
http://www.nd.edu/~manders8/Bonus%20Bid%20Points%20by%20School%20and%20Regional.xls

Enjoy!

unindicted co-conspirator · « **Reply #3 on:** March 05, 2004, 10:02:00 am »

Very impressive. I am shocked you could find the time. Did it take long?

Now you should do the silver teams, hehe.

AlphaTim · « **Reply #4 on:** March 05, 2004, 10:12:04 am »

I'll take care of the silver team rankings as soon as I get a chance. I probably won't get to it before Sunday or Monday though... But when I get them I'll make sure that I take all statistics into account to make sure that the results will be incredibly accurate.

princecaspian · « **Reply #5 on:** March 05, 2004, 12:16:58 pm »

This ranking system is OBVIOUSLY biased, flawed, and entirely inaccurate and unreliable. Columbia is ranked 34 and 41. These rankings clearly are grossly incorrect.

J/K -- good work putting this together. Still can't say I fully get it -- but then I rarely do -- yet it's interesting nonetheless.

TenaciousDRB · « **Reply #6 on:** March 05, 2004, 01:24:34 pm »

Quote from: princecaspian

This ranking system is OBVIOUSLY biased, flawed, and entirely inaccurate and unreliable. Columbia is ranked 34 and 41. These rankings clearly are grossly incorrect.

I must respectfully disagree, as we are ranked 1st.

RadicallyLiberal · « **Reply #7 on:** March 05, 2004, 03:33:00 pm »

Quote from: TenaciousDRB

I must respectfully disagree, as we are ranked 1st.

Here here.

ShaphPrime · « **Reply #8 on:** March 05, 2004, 05:25:28 pm »

I see 61 teams on Jaggers' gold list... I thought four came from the two silver tourneys?

Jaggers27 · « **Reply #9 on:** March 05, 2004, 05:50:14 pm »

Shaph, you are correct. The reason why there are 61 teams is because I included every team that *qualified* for gold in the rankings, even if they had to retract their bid. So there are teams on the list, especially from Miami OH, Furman, Rhodes, and ND, that are not going to Des Moines because they had to retract their bid.

ShaphPrime · « **Reply #10 on:** March 05, 2004, 06:23:35 pm »

Ahh I see I see, says the blind man. Gracias, Cecil!

Jaggers27 · « **Reply #11 on:** March 05, 2004, 06:25:42 pm »

Shaph, by referring to me as "Cecil" you move up even higher in my book!

The Voice of Reason · « **Reply #12 on:** March 06, 2004, 09:47:21 am »

Regarding the strength of regional....maybe I am slow, but how do you account for the fact that the College Park regional receives 5 bids to Des Moines and most other regions receive only two or three?
You should rank the regions based something along the lines of (bid bonus points earned from gold)/(# of bids) and avg this over the three years the bid bonus covers. Then do it also for silver and bronze (if it is called bronze). This should account for the discrepency in regional bids per region.

Golden Skull · « **Reply #13 on:** March 06, 2004, 10:17:16 am »

Why would you need a separate formula for that? That would be naturally taken into account by the bonus bid points that each school brings there. If you've got 3 extra bids to gold, you're also going to have teams present there that will bring the # of bonus bid points that earned them those 3 extra bids.

The Voice of Reason · « **Reply #14 on:** March 06, 2004, 06:03:05 pm »

however a region with five teams earning bid bonus points verus a region with only two or three teams is going to give them a higher total as a region. It should be expected that the regions with more bids would have higher scores. A points per bid statistic would more accurately show how well teams from regions do at nationals, instead of simply adding the points together and giving an advantage to the regions with more bids.
for example, if you have 5 teams there and they all get say, three wins, they'll get 15 times whatever the multiplier is...but if you have 3 teams there and they each have 5 wins, you'll get 15 times the multiplier....which region's teams did better? by the current ranking system the regions would have done the same, but if you take points per bid, the region that had three teams would have done better

RadicallyLiberal · « **Reply #15 on:** March 06, 2004, 06:16:43 pm »

Ok, so you're saying that while regionals with teams that have lots of bonus bid points are more competitive, there are more bids, so it's easier to get a bid?

The Voice of Reason · « **Reply #16 on:** March 07, 2004, 08:19:22 pm »

what im saying is that the system to rank regional strength is flawed because not ever region receives the same number of bids. if you have a region that only receives 2 bids to des moines, they simply cant earn as many points as a region with 5, unless they do awesome and the region with 5 bids does poorly.
perhaps i am just misunderstanding how jaggers27 ranked the regionals.

MockDock · « **Reply #17 on:** March 08, 2004, 02:05:20 am »

Well, the number of bids that are earned from each region is entirely dependant on how those teams perform in the nationals. But I would agree that some regionals (namely in the South and West) are just now developing stronger programs, and they have not yet earned enough power bid points. But I think Jaggers did the best with what he could do.

Also, thinking about it, I do think that point differential should be factored in, to some degree. Of course, the strength of opponents should heavily outweigh differential, but I'm sure there's some way to account for some team blowing out other good teams.

FundamentalPrecepts · « **Reply #18 on:** March 08, 2004, 08:20:39 am »

Is anyone working on ranking the Richmond teams? I'd love to see that.

MarshallWythe · « **Reply #19 on:** March 08, 2004, 02:00:05 pm »

Quote from: FundamentalPrecepts

Is anyone working on ranking the Richmond teams? I'd love to see that.

Just finished them, using Jaggers' formula... check them out at http://jdfrew.people.wm.edu

FundamentalPrecepts · « **Reply #20 on:** March 08, 2004, 11:37:50 pm »

Thanks, MW. You rock.

MarshallWythe · « **Reply #21 on:** March 08, 2004, 11:39:17 pm »

Quote from: FundamentalPrecepts

Thanks, MW. You rock.

No problem... just running the stuff derived by you geeks

olu orange · « **Reply #22 on:** March 09, 2004, 01:01:21 am »

MarshallWythe · « **Reply #23 on:** March 09, 2004, 01:07:59 am »

actually, something that would be interesting to test out would be an RPI-like system. For those unfamiliar with the term, the RPI is a rating system relied on in College Basketball, and is a pretty good determiner of a team's strength. The formula works as follows, with some tweaking, I believe:

your win % * .25+ your opponent's win%*.5+your opponent's opponent's win%*.25

in MT terms, that would come out to something like this, taking only the regionals into account:

# of ballots won/32+ CS/128 + Avg. CS of opponents/256

I can try to work on that tonight/tomorrow, but this one could take a looong time to work... I'll see if I can add it to the Richmond rankings on my site.

EDIT: Correction to the math

MockDock · « **Reply #24 on:** March 09, 2004, 01:43:36 am »

The RPI could be effective, but I don't think it would be able to account for strong/weak regionals. In NCAA, teams will have inter-conference match ups, which will demonstrate one conferences superiority over another. Jaggers' method tried to take this into account. A 5-3 team in the Los Angeles region is different from a 5-3 team in St. Paul. Under the RPI, they would be treated similarly.

MarshallWythe · « **Reply #25 on:** March 09, 2004, 02:03:44 am »

Quote from: MockDock

The RPI could be effective, but I don't think it would be able to account for strong/weak regionals. In NCAA, teams will have inter-conference match ups, which will demonstrate one conferences superiority over another. Jaggers' method tried to take this into account. A 5-3 team in the Los Angeles region is different from a 5-3 team in St. Paul. Under the RPI, they would be treated similarly.

Point noted... we'll see how this goes.

I've finished the Birmingham/College Park RPIs, and I'll continue them sometime tomorrow... hopefully I can have them up before I leave for Richmond on Thursday.

Jaggers27 · « **Reply #26 on:** March 09, 2004, 02:30:36 am »

MW, fantastic job on the Richmond rankings and interesting point on the RPI system. However, with a MT RPI system, you wouldn't have to worry about %s since every team has the same amount of ballots, the same possible CS, etc. The %s are used in NCAA BB because teams play different amounts of games. Although, in the end I don't think an RPI based system would operate too differently than the win qualifier, as the only difference would be that after accounting for wins and CS, the RPI would take opponent CS while the win qualifier takes regional strength. However, I think the results ought to be very interesting and I'll be looking forward to your work.

MarshallWythe · « **Reply #27 on:** March 09, 2004, 11:31:28 am »

Quote from: Jaggers27

MW, fantastic job on the Richmond rankings and interesting point on the RPI system. However, with a MT RPI system, you wouldn't have to worry about %s since every team has the same amount of ballots, the same possible CS, etc. The %s are used in NCAA BB because teams play different amounts of games. Although, in the end I don't think an RPI based system would operate too differently than the win qualifier, as the only difference would be that after accounting for wins and CS, the RPI would take opponent CS while the win qualifier takes regional strength. However, I think the results ought to be very interesting and I'll be looking forward to your work.

True. If I did it that way, I could modify the formula, and it would be scaled to a maximum value of 8. Since I've already started with the %s, I'm going to stick with them, but it doesn't really effect the rankings by RPI, since they'd be the same values either way, just multiplied by 8.

Also, I've thought of a way to throw the regional points in there. I scaled all of the strength of region scores by deriving the Log of them, then dividing them by the highest result (which would be from Pitt 610 and WM 336, as they provided no bid pts to College Park, the highest rated regional) *2, so that the maximum value was .5. This number was added to the RPI to determine a "Power points" value. I'm still working on the numbers, but I will be checking back here to see if you guys have any thoughts.

Lastly, I wanted to share something I found somewhat surprising... in the Indiana, PA regional, the byebuster 222, who finished with 2, had a CS of 43, one the highest I've seen so far. Who woulda expected that out of a byebuster?

MarshallWythe · « **Reply #28 on:** March 10, 2004, 03:04:13 am »

After much tedious work, I have finished the RPI for Richmond teams, and I have added an additional factor to it...

the RPIs are as per the formula I mentioned earlier (Ballots won/32+CS/128+Opp. Avg. CS/256). They ranged from a 0.650390625 for MTSU 660, who was also #1 in Jaggers' Index, to a 0.442382813 for NYU, who was #33 in Jaggers' Index.

I also added the Strength of Region to the RPI to create a Power Points index. What it does is scale the Strength of Region (SoR) down, using the following formula:

Power of Region= (Log(SoR)/2.425501615)*.25

I then added the Power of Region to the RPI, creating the Power Points Index. The whole thing is up where I originally put the Richmond rankings, http://jdfrew.people.wm.edu. Enjoy!

PS- Two notes:

- In the Power of Region formula, 2.425501615 was derived from taking the log of the highest SoR, which belonged to both Pitt 610 and WM 336. This was done to help set a maximum value on the PoR scores at .25.

-For a few of the St. Louis regional qualifiers, as well as for Loras at St. Paul, I ran into a situation where one of the qualifying teams' opponents did not compete in all 4 rounds. While this wasn't a problem when needing to figure out their record, it is when trying to calculate CS. After giving it some thought, I decided to throw the team who did not finish the tourney out of the Opp. Avg. CS calculations, as there is no way to accurately calculate what their CS would have been, had they continued throughout the tourney.

MarshallWythe · « **Reply #29 on:** March 10, 2004, 01:46:05 pm »

bump... they're done

TenaciousDRB · « **Reply #30 on:** March 10, 2004, 01:53:03 pm »

Quote from: MarshallWythe

-For a few of the St. Louis regional qualifiers, as well as for Loras at St. Paul, I ran into a situation where one of the qualifying teams' opponents did not compete in all 4 rounds. While this wasn't a problem when needing to figure out their record, it is when trying to calculate CS. After giving it some thought, I decided to throw the team who did not finish the tourney out of the Opp. Avg. CS calculations, as there is no way to accurately calculate what their CS would have been, had they continued throughout the tourney.

Loras' opponent would have gone 0-8, if that helps.

MarshallWythe · « **Reply #31 on:** March 10, 2004, 01:57:23 pm »

Quote from: TenaciousDRB

Loras' opponent would have gone 0-8, if that helps.

Yeah, I know, but that's not the problem. The problem rests with figuring out who they would've faced in the last two rounds. The other option I considered was doubling their CS, but that still would have been too much guesswork, and not enough accuracy. I'm more comfortable with just throwing their CS out of the averaging.

Mock Captain · « **Reply #32 on:** March 12, 2004, 10:45:54 pm »

HAVE AN UPDATED LIST OF GOLD TEAMS RANK....INCLUDING THOSE WHO QUALIFIED FOR GOLD AT FLORIDA?

Golden Skull · « **Reply #33 on:** March 13, 2004, 09:30:18 pm »

It's ND's spring break and I'm sure Anderson has other things to do, for once - he'll probably get to it soon.

Mock Captain · « **Reply #34 on:** March 19, 2004, 11:07:22 pm »

Quote from: Mock Captain

HAVE AN UPDATED LIST OF GOLD TEAMS RANK....INCLUDING THOSE WHO QUALIFIED FOR GOLD AT FLORIDA?

And Kentucky...

(I was wondering if there was a list of ALL teams going to DM, and what rank they are....)

Golden Skull · « **Reply #35 on:** March 20, 2004, 03:03:53 am »

Well, I guess Anderson has other stuff to work on - if anyone else wants to finish out the list though, they're free to do so, the formulae are all there.

Jaggers27 · « **Reply #36 on:** March 21, 2004, 01:10:32 am »

It's not a matter of being too busy, it's a matter of losing motivation since being chewed out by my coach over the existance of said rankings....

Golden Skull · « **Reply #37 on:** March 21, 2004, 02:26:43 am »

In all fairness, he may not have understood what exactly the rankings meant. But yeah - rankings may be on hiatus for a while.

nwt · « **Reply #38 on:** March 21, 2004, 03:04:32 am »

wait, what the hell business is it of your coach what you do in your private time?

ArizonaMock · « **Reply #39 on:** March 21, 2004, 03:08:14 am »

Quote from: nwt

wait, what the hell business is it of your coach what you do in your private time?

a-fucking-men

nwt · « **Reply #40 on:** March 21, 2004, 01:33:21 pm »

Quote from: ArizonaMock

Quote from: nwt
wait, what the hell business is it of your coach what you do in your private time?

a-fucking-men

...an interestingly worded response.

what would freud say!?

ArizonaMock · « **Reply #41 on:** March 21, 2004, 03:34:35 pm »

*Lewd Wink*

princecaspian · « **Reply #42 on:** March 22, 2004, 11:08:12 am »

Quote

It's not a matter of being too busy, it's a matter of losing motivation since being chewed out by my coach over the existance of said rankings....

Maybe your coach was pissed off about where your program ended up in the rankings.

Tell him that there are dozens of students and coaches who find the rankings interesting and that there is absolutely no harm in them whatsoever. And tell him to show a little love.

TenaciousDRB · « **Reply #43 on:** March 22, 2004, 01:14:11 pm »

Quote from: princecaspian

Quote
It's not a matter of being too busy, it's a matter of losing motivation since being chewed out by my coach over the existance of said rankings....

Maybe your coach was pissed off about where your program ended up in the rankings.

What, fifth?

KingsfieldNot · « **Reply #44 on:** April 05, 2004, 12:02:45 pm »

Quote from: Jaggers27

Finally, an ordinal ranking of regional strength:

1) College Park Regional: 266.38 points
2) Macalaster Regional: 235.08 points
3) St. Louis Regional: 208 points
4) Princeton Regional: 194.29 points
5) Marquette Regional: 193.75 points
6) Peoria Regional: 193.33 points
7) Furman Regional: 185 points
8) Samford Regional: 175 points
9) Notre Dame Regional: 173.25 points
10) Lewis Regional: 157 points
11) Tallahasse Regional:150.75 points
12) Washburn Regional: 145.75 points
13) Manchester Regional: 137.25 points
14) Weber State Regional: 135.92 points
15) St. John's Regional: 132.42 points
16) Indiana, PA Regional: 124.33 points
17) Miami OH Regional: 109.42 points
18) Portland Regional: 97 points
19) So. California Regional: 87 points

Hmm. UCLA and Cal came out of which regions?

News:

Author Topic: Ranking of Gold Qualifiers Revised and Completed (Read 4971 times)