Gonzaga should be #15-20, consistently go the way of their first halves

[pittpitt]

#10 Gonzaga, #18 UConn, #25 Texas A&M, and 6-0 (and about to be ranked high) Syracuse all just played each other close, proving to all basically be each others' equals. After all that I'll call Gonzaga about the 15th to 20th best team in the country.

As far as their scoring margins by half:
NoAriz, +15, +24, W by 39
MtStM, +12, +33, W by 45
(These last 3 have much more talent - notice the Zags after halftime)
Wash, +17, -1, W by 16
Tex A&M, -3, +2, L by 1
UConn, +16, -13, W by 3

Pitt wasn't just beating Gonzaga by 2 at halftime, Pitt was also the team much more negatively affected by the slick floor. We were clearly outplaying them, and we were going to win that game. Wiltjer was already going off, so it wasn't like he was going to save them in the 2nd half. Despite not getting the respect of pollsters yet even though Pitt has shown they're better than a legit top-20 team, it will take that same type of showing against #16 Purdue on Tuesday to really get people's attention. They also have a balanced, high-scoring offense, but they have two 7-footers patrolling the paint. Gonzaga had 3 guys between 6'10" and 7'1" all playing at once (1 expected to be an All American, another a top-10 draft pick), so nothing about Purdue is unknown or scary to this squad, and they're much sloppier with the ball than Gonzaga or ourselves.

 

[Fredact]

Here's the impact of not finishing that game:

At RPIforecast, Pitt's current projected RPI is 42 with a SOS of 30. Had they played and LOST the game with Gonzaga, their RPI would be projected to be 31, with a SOS of 19. Had they played that game and won, the RPI projection would be 28 with that SOS of 19.

Once again showing that the RPI is a joke of a statistic, but it is what it is. Not finishing that game will cost Pitt a seed line or 2.

 

[DT_PITT]

Here's the impact of not finishing that game:

At RPIforecast, Pitt's current projected RPI is 42 with a SOS of 30. Had they played and LOST the game with Gonzaga, their RPI would be projected to be 31, with a SOS of 19. Had they played that game and won, the RPI projection would be 28 with that SOS of 19.

Once again showing that the RPI is a joke of a statistic, but it is what it is. Not finishing that game will cost Pitt a seed line or 2.

All true, but the "word" is that the committee will take everything into account when seeding teams -- including the fact that this game was cancelled.

First things first -- let's get IN to the tourney.

 

[KiwiJeff]

#10 Gonzaga, #18 UConn, #25 Texas A&M, and 6-0 (and about to be ranked high) Syracuse all just played each other close, proving to all basically be each others' equals. After all that I'll call Gonzaga about the 15th to 20th best team in the country.

As far as their scoring margins by half:
NoAriz, +15, +24, W by 39
MtStM, +12, +33, W by 45
(These last 3 have much more talent - notice the Zags after halftime)
Wash, +17, -1, W by 16
Tex A&M, -3, +2, L by 1
UConn, +16, -13, W by 3

Pitt wasn't just beating Gonzaga by 2 at halftime, Pitt was also the team much more negatively affected by the slick floor. We were clearly outplaying them, and we were going to win that game. Wiltjer was already going off, so it wasn't like he was going to save them in the 2nd half. Despite not getting the respect of pollsters yet even though Pitt has shown they're better than a legit top-20 team, it will take that same type of showing against #16 Purdue on Tuesday to really get people's attention. They also have a balanced, high-scoring offense, but they have two 7-footers patrolling the paint. Gonzaga had 3 guys between 6'10" and 7'1" all playing at once (1 expected to be an All American, another a top-10 draft pick), so nothing about Purdue is unknown or scary to this squad, and they're much sloppier with the ball than Gonzaga or ourselves.

Nobody knows we would have won that game. A 2 point lead is not insurmountable. Maybe they make an adjustment, maybe we go cold shooting. Maybe we win by 20, maybe we lose by 4. Nothing is for sure in a 2 point game at the half between 2 good teams.

 

[R-M-U-h-o-o-p-s-f-a-n]

Nobody knows we would have won that game. A 2 point lead is not insurmountable. Maybe they make an adjustment, maybe we go cold shooting. Maybe we win by 20, maybe we lose by 4. Nothing is for sure in a 2 point game at the half between 2 good teams.

I agree, everyone just "assumes" that Pitt would have won the game!

What if Kyle W. went for 20 pts in the second half? Who knows but no one can say "we would have for sure won". We were up 2 at the half, not 20.

 

[steel_curtain]

I agree, everyone just "assumes" that Pitt would have won the game!

What if Kyle W. went for 20 pts in the second half? Who knows but no one can say "we would have for sure won". We were up 2 at the half, not 20.

Artist didn't play almost the whole first half.

 

[KiwiJeff]

Artist didn't play almost the whole first half.

True. They should cancel all games at the half then where the fans of the team that is up by 2 believe they are a shoo in to win. Why waste everybody's time playing the second half if it is a foregone conclusion?

 

[Harve74]

I agree, everyone just "assumes" that Pitt would have won the game!

What if Kyle W. went for 20 pts in the second half? Who knows but no one can say "we would have for sure won". We were up 2 at the half, not 20.

As Fred's post shows, even LOSING the Gonzaga game would have raised our RPI ten places.

 

[R-M-U-h-o-o-p-s-f-a-n]

As Fred's post shows, even LOSING the Gonzaga game would have raised our RPI ten places.

Yea, true. The RPI is weird that way. Too bad the game couldn't have been played in a gym with A/C.

 

[Joe the Panther Fan]

As Fred's post shows, even LOSING the Gonzaga game would have raised our RPI ten places.

BUT, it would have raised our RPI ten places RIGHT NOW. When the number of games that we have played that counts towards the RPI is two. At the end of the season, when we have 30 or more of them, any one particular game, including the non-played Gonzaga game, will be reduced in importance. We can check at the end of the year, but it is much more likely that not finishing that game is going to cost us a spot or two at most, rather than it cost us anywhere close to ten spots.

 

[Fredact]

That's not correct Joe. I was using the RPI forecast site that was projecting end of year RPIs. The impact of not playing that 1 game is that big on the final RPI.

 

[Joe the Panther Fan]

That's not correct Joe. I was using the RPI forecast site that was projecting end of year RPIs. The impact of not playing that 1 game is that big on the final RPI.

Well doing some quick back of the envelope calculations on the actual results from last season had we added a game against Gonzaga at a neutral site to last season's schedule and won the game it would have improved our end of year RPI from 77 to 68. Had we played that game and lost it would have changed our end of the year RPI from 77 to, well, 77.

Now I don't have the opponent's opponents portion of last year's RPI, however with the conference schedule that Gonzaga plays my guess is that adding in their opponents would actually hurt our RPI a wee little bit, but not by any significant degree. We will have to see how it all plays out at the end, but I would happily bet whatever sum of money that anyone wants to put up that at the end of the season adding a loss in the Gonzaga game to our record is not going to improve our final RPI by 11 spots.

 

[Harve74]

Well doing some quick back of the envelope calculations on the actual results from last season had we added a game against Gonzaga at a neutral site to last season's schedule and won the game it would have improved our end of year RPI from 77 to 68. Had we played that game and lost it would have changed our end of the year RPI from 77 to, well, 77.

Now I don't have the opponent's opponents portion of last year's RPI, however with the conference schedule that Gonzaga plays my guess is that adding in their opponents would actually hurt our RPI a wee little bit, but not by any significant degree. We will have to see how it all plays out at the end, but I would happily bet whatever sum of money that anyone wants to put up that at the end of the season adding a loss in the Gonzaga game to our record is not going to improve our final RPI by 11 spots.

Logically Joe, I've got to think your point makes sense, but I can't fault Fred for posting what the site he quotes said either. For a simple formula, the RPI seems to give confounding results sometimes.

Ultimately, the slippery floors might cost us a couple lines in the seeding. We can just hope the difference us , say, between a 4 and a 6 , not between the bubble and out.

 

[Joe the Panther Fan]

I can't fault Fred for posting what the site he quotes said either.

I certainly don't fault him for it. Admittedly I misunderstood his original post, but I still think that they are overstating what the actual final difference will be. My guess is that because there are still so many games left that there is too much reversion to the mean (as it were) going on, and that the team's "final" RPIs are being shown as much more closely grouped together than they actually will be when the actual games are played. Which means that they are showing a change of say .0080 in the RPI moving someone a lot more positions than actually moving .0080 will in the final standings.

That would especially be true at the top end of the RPI "standings" where the reality is that there is normally more of a difference between the average spots on the list than there are in the middle of the standings. For instance last season the difference between 20 and 30 was .0215. The difference between 70 and 80 was .0066. One result can move you a heck of a lot more places in the standings if you are in the 70-80 range than it can when you are in the 20-30 range. So the higher Pitt's RPI is at the end of the year the smaller the impact on finishing the Gonzaga game will have had.

 

[Fredact]

I certainly don't fault him for it. Admittedly I misunderstood his original post, but I still think that they are overstating what the actual final difference will be. My guess is that because there are still so many games left that there is too much reversion to the mean (as it were) going on, and that the team's "final" RPIs are being shown as much more closely grouped together than they actually will be when the actual games are played. Which means that they are showing a change of say .0080 in the RPI moving someone a lot more positions than actually moving .0080 will in the final standings.

That would especially be true at the top end of the RPI "standings" where the reality is that there is normally more of a difference between the average spots on the list than there are in the middle of the standings. For instance last season the difference between 20 and 30 was .0215. The difference between 70 and 80 was .0066. One result can move you a heck of a lot more places in the standings if you are in the 70-80 range than it can when you are in the 20-30 range. So the higher Pitt's RPI is at the end of the year the smaller the impact on finishing the Gonzaga game will have had.

It turns out you CAN fault me! Mostly because I didn't read one of the disclaimers. It seems that RPIforecast normally forecasts all the way through the conference tournament, but when you use their RPI Wizard, it only forecasts through the end of the season. So I had some apples and oranges going on in the original post.

Here is the correction with the numbers as they now stand:

Through end of regular season as now projected: RPI: 30 SOS: 24
If Pitt had played Gonzaga and lost: RPI: 34 SOS: 19
If Pitt had played Gonzaga and won: RPI: 27 SOS: 19


So going back to my original post, as Emily Litella used to say "never mind".


Thanks Joe for making me think more about this.

 

[Joe the Panther Fan]

Thanks Joe for making me think more about this.

No big deal. It will be interesting to see at the end of the year how it actually turns out. The better we do actually the less important the Gonzaga non-game will become, at least as far as RPI goes. It would actually make more difference if we end up on or near the bubble than if we are clearly in and are just looking at seeding. But we will see.