Flop The Nuts

If wrestler A loses to 3 guys with statistically equal or better records, and wrestler B's losses were to statistically worse opponents, wouldn't wrestler B have a slightly lower RPI because of those losses?

Assuming that ISU's three criteria are correct, and because the set of wrestlers both A&B wrestled are the same, I agree with ISU here.

Wrestler A has "quality" losses, in that he lost to three people that had equal or better records. But, that also means that wrestler A had three fewer quality wins than B.

Wrestler B has three more "quality" wins than A, but he also has three bad losses.

architeuthis

Ok, That makes sense. But, Shouldn't Statistically worse wins carry more weight than statistically worse losses. A win is still a win. Despite the fact that the competition is weaker, Wins should be worth more.

I get that Wrestler B has 3 more statistically better wins, but shouldn't the 3 startistically worse losses weigh more?

This example is weird because it uses fixed data where everyone wrestles everyone else. I know it isn't like that in the real world. Thanks ISU for putting the numbers and formulas there so we can see it. It makes way more sense.

I just can't get my brain around the fact that technically, you could get pinned 3 times (whether it is a fluke or not) by guys with losing records, and still have the same RPI as the fourth best guy who lost squeakers to the three best.

If it were only a word problem.....sigh

M

__________________
"It is the difference between humble and hubris that distinguishes their positions." -TLV

ptoc-nwca

RPI is only one component in the whole qualifier formula. RPI does factor in margin of victory into the formula. However, there is a coaches poll which will clearly factor in margin of victory.

When factoring in the RPI, you exclude the head to head match. In other words, if the guy I compete against is 9-1 and I beat him, his win percentage is still calculated at .900 and not at .818.

To get the opponen's winning percentage you take each opponent's won-lost percentage against other Division I wrestlers (excluding the wins or losses against the wrestler in question), then average these percentages. You do not add up the total wins and losses for the opponents and then get the winning percentage.

architeuthis

Thanks ptoc.

How is the coaches poll supposed to factor into qualification then? Is the RPI supposed to play into rankings then?

Maybe, since I'm the one asking all the stupid questions, someone has a link to this information? That would probably make all this easier.

__________________
"It is the difference between humble and hubris that distinguishes their positions." -TLV

RYou

I wonder how much they paid for this computer program? They prabably could have just bumped the total invitees to 360 and paid for them all, allotted a few more slots for wild cards and sat pat with the current scheme. Surely, someone is not there with pencil, paper and an abacus.

__________________
19-0 ? Hell No ! 18 Wins and 1 GIANT Loss !

ptoc-nwca

The issue is that the NCAA mandated a change from using historical data so a new system to allocate the qualifiers had to be made.

The coaches poll is completely independent of the RPI. They are two different components of the qualifier system and do not relate to each other.

The RPI was run and used for the last season as a test.

ISU2008

OK, did not know that.

So Wrestler A's RPI goes down some from my previous calculations because the good guys he lost to, lose some wins off their record (their win % goes down); and it goes up some because the bad guys he beat, lose some losses off their record (their win % goes up).

Wrestler B's RPI goes down some because the bad guys he lost to, lose some wins off their record (their win % goes down); and it goes up some because the good guys he beat, lose some losses off their record (their win % goes up).

But I don't think you can determine how much the RPI's change (how much their opponent's W-L records change) without exact numbers.

Am I right? (Probably not)

__________________
I don't wanna sound like a queer or nothin, but I think unicorns are pretty kick-ass.

architeuthis

Thanks again. Thats exactly the point I was trying to get at. Stupid BCS

__________________
"It is the difference between humble and hubris that distinguishes their positions." -TLV

ptoc-nwca

It is hard to factor without having exact numbers. It is fairly easy to figure out factors I and II of the RPI by hand. It gets a lot more complicated when you have to calculate Factor III (opponent's opponent).

RYou

Well at least knows something about this blasted formula

__________________
19-0 ? Hell No ! 18 Wins and 1 GIANT Loss !