If You Could Trade . . .

[UberZagFan]

two wins (UNC and Stanford) for 4 losses (Georgia, Nevada, St. Mary's and LMU), would you?

It would put GU at 19-6 (9-0) but no real great quality wins except Nevada. Assume a loss to Memphis and a loss in the WCC championship, putting the final record at 25-8. Would GU dance?

Do you really think those 4 losses outweigh the two wins? GU would have only one top 50 win. No good win since December and an even lower RPI. I say that 25-8 team would go to the NIT and the 23-10 team (if GU wins out except Memphis and WCC championship) would get an invite.

[tobizag]

Uber,
Very interesting proposition...The only thing I wonder about is how highly will people think of Texas at the end of the season? They certainly get better by the day, because they ride Durant's success. This might count as one big win. Could you explain the RPI drop? I'm not an RPI expert.

Overall, I'd stick with where we are now. This team is learning a lot and gaining valuable experience. I often have wondered in years past if our teams would have gone deeper into the tournament had they experienced some defeat throughout the course of the season.

Good post.

[dawgncarolina]

Do you really think those 4 losses outweigh the two wins? GU would have only one top 50 win. No good win since December and an even lower RPI.Your RPI would be higher, not lower.

RPI takes into account your record, the record of your opponents, and the record of your opponents' opponents.

RPI does not take into account who you beat or who you lost to.

From a purely RPI perspective you'd be much better off with those 4 wins than the 2 you have.

From a non RPI perspective you present an interesting debate. I'd be inclined to take the trade were I a Zag fan, though.

[tobizag]

Thanks for the RPI lesson oh Tar-Heeled one.

[dawgncarolina]

Thanks for the RPI lesson oh Tar-Heeled one.Sacrilege!

;)




And you're welcome, btw. I've found most people have no concept of how the RPI works. It's actually not really that complicated. I should add that it also factors in location of the games, but that wouldn't affect this trade one way or the other.

[UberZagFan]

Uber,
The only thing I wonder about is how highly will people think of Texas at the end of the season? They certainly get better by the day, because they ride Durant's success.

Well, I certainly count the Texas win as a good win, but right now Texas is below GU in RPI at 56 (GU is 53 according to http://www.kenpom.com/rpi.php ). So that win is still not a Top 50 win but UT may move into the Top 50 by season's end.

Could you explain the RPI drop? I'm not an RPI expert.
1/4 of the RPI is based upon winning percentage so the 2 more wins would raise the RPI a bit but 1/2 of the RPI is based upon opponent's winning percentage so 4 of GU's opponents would have worse records whereas only 2 would have better--a 2 game difference. So the 2 game plus in GU's win column would only account for 1/4 whereas the 2 game drop in GU's opponents win column would account for 1/2 of the RPI. It would likely only be a small drop, but a drop nonetheless.


I often have wondered in years past if our teams would have gone deeper into the tournament had they experienced some defeat throughout the course of the season.


I have wondered the same thing. I'm not so sure that it's defeat though as all GU teams have experienced some defeat--the bigger issue may be either defeat later in the season or tough competition later in the season.

[SleepyPigs]

Great discussion question, Uber. There is no way I'd trade those two wins for four losses. Those were the two most enjoyable games of the season. We'll never forget that UNC game. We beat UNC in the Garden! After all, that's what it is all about, right? Getting enjoyment from watching your team...

[tobizag]

Like I said, I am no expert. I just know the higher, the better according to ESPN...Thanks for the interesting point. You would take the trade?:D

[lothar98zag]

nevermind

[UberZagFan]

Your RPI would be higher, not lower.

RPI takes into account your record, the record of your opponents, and the record of your opponents' opponents.

RPI does not take into account who you beat or who you lost to.

From a purely RPI perspective you'd be much better off with those 4 wins than the 2 you have.

From a non RPI perspective you present an interesting debate. I'd be inclined to take the trade were I a Zag fan, though.DNC, I don't think you are correct and the simplest explanation is that 1/2 the RPI is opponent's winning percentage compared to only 1/4 being your own winning percentage. GU would only gain two games in the win's column but the opponent's would lose two as well. That would only be a small difference but it would be a negative difference.

[tobizag]

My head is spinning now with all of these different explanations!!!

[UberZagFan]

I wouldn't take the trade because of what Sleep said. I am a GU fan and I still want to see the Zags win games like UNC or Mich. St. last year. Those are games that make memories.


P.S. I agree Lothar is an idiot.

[dawgncarolina]

DNC, I don't think you are correct and the simplest explanation is that 1/2 the RPI is opponent's winning percentage compared to only 1/4 being your own winning percentage. GU would only gain two games in the win's column but the opponent's would lose two as well. That would only be a small difference but it would be a negative difference.Those two games have a MUCH bigger impact on the 25% that goes into your own winning percentage than the 50% that goes into your opponents winning percentage. Your winning percentage is (currently) based on 25 games. Your opponents' winning percentage is based on roughly 625 games.

If what you're saying were true it would always be better for your RPI if you lost.

[UberZagFan]

Either way it's a slight change in the RPI.

I'd still take the UNC win!!!

[lothar98zag]

DNC, I don't think you are correct and the simplest explanation is that 1/2 the RPI is opponent's winning percentage compared to only 1/4 being your own winning percentage. GU would only gain two games in the win's column but the opponent's would lose two as well. That would only be a small difference but it would be a negative difference.

The 50% opponent's records portion of the RPI formula does not include the game against GU, so it doesn't matter if GU won or loss certian games versus certain opponents.

For example:

UNC is currently 21-3, but for GU's RPI UNC's record is calculated at 21-2
&
Duke is currently 18-6, but for GU's RPI Duke's record is calculated at 17-6

[dawgncarolina]

The 50% opponent's records portion of the RPI formula does not include the game against GU, so it doesn't matter if GU won or loss certian games versus certain opponents.

For example:

UNC is currently 21-3, but for GU's RPI UNC's record is calculated at 21-2
&
Duke is currently 18-6, but for GU's RPI Duke's record is calculated at 17-6Thanks lothar.

I always wondered if that happened but wasn't sure.

It's a good rule, obviously.

[tobizag]

My head is spinning less now...thanks to all of you.

[UberZagFan]

Thanks Lothar. Glad that we settled that tangential issue raised in the initial post.

Again, I'd keep the UNC and Stanford wins.

[lothar98zag]

My head is spinning less now...thanks to all of you.
Relax now while you can. It'll get more complicated over the next 4 weeks.

BTW - if you are interested, here's one of the many places you can read/learn about the RPI: http://kenpom.com/blog/index.php/weblog/rpi_help/