RPI for 2008

Maybe you should send your analysis to the coaches from the strong regions - especially those on the bubble.

Let THEM do the cajoling/embarrassing/weedling/ bitching/ and forcing for you.

I did that after the season last year and I'll do it again in two weeks, with this year's data.

What I see is the West and the Southeast are the only regions that have more wins than losses. So I am not sure what makes the West region clearly the strongest region

Come play in my casino

I was, of course, being flip. However, here's one way to look at it.

UP's unadjusted RPI in UPSF's last work-up was .7059

The difference between The West region and the nearest region is .0505
You could look at it as if all the teams in the West were being devalued by that amount. which means an RPI adjusted for region strength would give UP something like .7564


A team .0505 down from UP is about 11 or 12 places down the RPI. it would make the top 10 in the RPI an almost exclusively Western affair, and could make the 16 seeds almost all Western.


It doesn't matter right now because we are #1 in the RPI, but just suppose we had Colorado's RPI of .6557 which is 11th.
That .0505 would put us back in first place, instead of 11th, assuming the teams in front of us weren't from the West.

That could be the difference between a #1 seed and no seed at all. It would have given us a home advantage last year instead of North Carolina, perhaps, which meant a comfy ride through the playoffs instead of trips to high altitude.

Since there are only about 34 teams that get at large bids, it could be the difference between tournament and no tournament for a lot of western teams.

It's a big deal.

Just for fun, I ran the totals, by region, for wins-losses-ties, for games through Sunday, September 21. I also calculated the average number of out-of-region games per team for each region. Here are the results:

Central: 78-77-15 (2.98 out-of-region games per team)
Great Lakes: 91-95-16 (3.26)
Mid Atlantic: 66-81-24 (3.56)
Northeast: 54-66-18 (3.37)
Southeast: 90-81-21 (3.15)
West: 77-56-22 (3.04)

If we had played Oregon State on Sunday, would we still be on top of the RPI standings?

harryb wrote:If we had played Oregon State on Sunday, would we still be on top of the RPI standings?

It looks to me that we would drop to #5. Element #1, our winning percentage, would be higher. But in terms of strength of schedule, Element #2 (opponents' winning percentage) would be lower; and Element #3 (opponents' opponents' winning percentage) would be lower by almost the exact amount that Element #1 would be higher. The net decrease in our RPI would drop us down.

What would happen now is not that important, however, since we're not even half way through the season yet. The bigger issue would be at the end of the season: Oregon State may win two more non-conference games, but after that they could end up 0-9 in the Pac 10. At that point, instead of contributing their current 5-3 record to our RPI, they could be contributing a 7-12 record, which in all likelihood would more than offset the benefit of the additional win assuming we would have beaten them.

Boy, UPSF, It's sure seems like your opponent's won-loss record is really critical to your chance of a high seed. You really have to have a great season, meaning very few losses, and your opponent's have to also have very good seasons.

In theory, couldn't you lose only 2 or 3 games and if your opponent's records are lousy, then you might not even end up in the top ten.

A team definitely could lose only two or three games and end up not in the top 10. Right now, Minnesota is 10-0-0 but is #55 in the RPI rankings. Its rank is so low because its opponents, on average, have won only a little over 1/3 of their games.

This early in the season, with teams having played few games, one win or loss by an opponent can have a significant effect. To illustrate -- I hope you enjoyed staring at the post with Portland at #1 in the RPI -- Colgate lost tonight, moving the Pilots to #2, just below Texas A&M. It was fun being at #1, and if the Pilots can win two this weekend, maybe we'll get back there.

Towards the end of the season, when teams will have played a little more than twice the games they have played now, an opponent's win or loss will have a reduced effect. Using the Pilots as an example, by the end of the regular season, each opponent will be making a 1/18 contribution to the Pilots' strength of schedule elements vs the 1/8 contribution each makes today. The strength of schedule elements also will be subject to some pressure to move towards .5000 once conference play begins, since for each win of a conference opponent, another conference opponent will be recording a loss. (Are all these details driving you crazy? Welcome to the club!)

Crazy ain't strong enough a word, UPSF. But I do appreciate all your hard work and know that it takes you a lot of time to do this.

But you offer a special niche to this forum. We will actually have a pretty good idea where we stand come playoff seeding time. That is a great contribution by you, along with anyone else who is deeply involved, like Geezer and anyone else I have forgotten.

But it is your baby. So keep up the great work!

Interesting to see what will happen once WCC play starts since it seems we are currently the only ones in the top 25 (so strange)... Would have been nice for our conference mates to do a little better in out of conference play, geez, its like they want us to go on the road for the playoffs...

The weakness of the WCC definitely is a concern this year, for strength of schedule purposes. So let's hope they win as many of their remaining non-conference games as possible. The good side of things for the Pilots is that our non-conference schedule may contribute more to our strength of schedule than in prior years. Having UCLA, USC, Florida, Rutgers, and Penn State among our non-conference foes, plus also possibly Oregon and Kennesaw State -- and maybe UW -- could help us more than the balance of the WCC will hurt us.

PS - On going back and looking at the numbers, the WCC teams so far this year, excluding the Pilots, are 28-26-9, which is not as bad as I thought. Interestingly, Santa Clara at 3-5-1 has the worst record. Looking at the strengths of schedule they have been playing, they are pretty tough. Their opponents have won approximately 60% of their games. The only team whose opponents have a cumulative losing record is San Francisco. So this reconfirms that if the WCC teams can do well in their remaining non-conference games, the conference's off-year this year won't hurt that much. Of course, this all is relative to how the other key conferences are doing in their non-conference games.

UPSF... here's a theoretical question I've been meaning to ask for a while.

Once all the non-conference games are over, is it basically a wash in terms of what happens with the WCC? What I mean is... no matter what happened in the pre-season, would it even matter if UP beat two 6-0-0 WCC teams but also had to play two 0-0-6 WCC teams? Would that ultimately be the same as beating four 3-3-0 teams? I ask because at the end of the day, the total number of wins is going to equal the total number of losses when it comes to the WCC teams. So does the gross effect of all that winning and losing ultimately just balance out?

Does that make sense? I realize there are bonuses awarded for beating top 50 etc. RPI teams so in that sense we want the better teams to continue to win, but what about the expense of having the lower teams really dragging us down?

I realize I'm not making myself very clear, but do you get my point?

Said another way... once non-conference play is done, does it even matter for UP's RPI what the rest of the WCC does in terms of playing each other?

Stonehouse wrote:UPSF... here's a theoretical question I've been meaning to ask for a while. ... Said another way... once non-conference play is done, does it even matter for UP's RPI what the rest of the WCC does in terms of playing each other?

Your observation is correct. Once conference play starts and the non-conference games are done, the intra-conference results don't really matter. For every Team A win, there will be a Team B loss; and for every Team A tie, there will be a team B tie. In either event, the net effect will be to pull the other conference teams' RPIs towards .5000. What does matter is what the conference teams' non-conference records and strengths of schedule are.

I went ahead and ran the win-loss-tie records by conference to get a feel as to how the WCC stacks up right now against the other conferences. This presumably gives a rough picture of how intra-conference play will affect teams' RPIs. If you want to think about a specific team and how its conference games roughly would contribute to its RPI, remember that you have to delete that team's record from the totals. (I say "rough," by the way, because for strength of schedule the RPI doesn't simply add up opponents' wins, losses, and ties and then compute an opponents' winning percentage. Instead, it computes a winning percentage for each opponent and then calculates the opponents' average winning percentages. The numbers are different, due to differences in the numbers of games played.) A comparison of the conferences makes the WCC look like a problem for the Pilots.

ACC 67-19-6
Big East 74-43-15
Big 10 61-31-7
Big 12 58-27-4
Pac 10 57-16-9
SEC 52-35-10
WCC 35-27-9

I've posted a new RPI report covering games through September 28 on BigSoccer. The report is an attachment to the cpthomas entry at 3:45 pm on September 28. No surprise, the Pilots are at the top. Here's the link: http://www.bigsoccer.com/forum/showthread.php?t=650592

UPSF

I've tried to log in to Big Soccer. Have username and password and keep getting the sign on screen and cannot access your document. Grrrr.

You may have my email address = firstnamelastname@aol.com

Could you send it to me?

Thanks!

ps. Excited to know that UP is still #1

This is great. Getting the RPI each week as it shakes into alignment really demonstrates how it works.

You have to think USC is kicking themselves for scheduling 1-8-2 PSU the same weekend as it played UP.

Those two games has to be what puts them all the way down the chart at #14.

UPSF, have you done a run-down by conference or region yet? Are you at least contemplating that at the end of our out-of-conference season?

I'll post a comparison of the regions when I get back in Portland later this week. If I have time, I'll also do it by conference, as that will be an important factor from here on out, since most teams' further games will be purely in-conference so that the balance of their strength of schedule elements will be determined by strength of conference. I'm guessing, without knowing for sure, that this not help the Pilots in relation to the other top teams. However, our non-conference strength of schedule has been very good, so that will help.

Good news. I await the results.

The conference hasn't been doing so good this year, it seems. We'll probably get dragged into the pack a bit. Our big hope is that the Pac10 teams beat each other up.

For those interested in comparisons of the six geographic regions, here are the average RPIs by region, taking into consideration only inter-regional games. Also included are the number of inter-regional games each region played, the number of teams in each region, and the inter-regional games per team for each region. There are 17 teams that played no inter-regional games so that their records made no contribution to these numbers. In addition, a significant portion of the remaining 303 teams have not played enough inter-regional games to allow the system to compute RPIs for them. The result is that it is possible to compute RPIs only for 175 teams, so those are the teams whose "inter-regional RPIs" have contributed to these numbers.

The numbers, in order, are average RPI, number of games, number of teams, and average games per team:

Central 0.5049 189 57 3.32

Great Lakes 0.5090 222 62 3.58

Mid Atlantic 0.5057 213 48 4.44

Northeast 0.5007 166 41 4.05

Southeast 0.4985 226 61 3.70

West 0.5462 162 51 3.18

I wonder if that means we are scheduling teams from the Southeast and Northeast regions next year....

Based on a suggestion from someone participating in the BigSoccer RPI thread, I've run the RPI program to compute the RPIs of teams using only inter-conference games. In other words, I've excluded intra-conference games. I've then averaged each conference's RPIs to come up with RPI-based conference ratings. The inter-conference RPI numbers give what I believe to be the best RPI-based comparison of conferences that one can come up with. (There are a few out-of-conference games remaining this coming weekend, after which we'll be done with inter-conference matchups except for a few strays later in the season. If I have time, I'll run these numbers again next week.)

Here are the conference average RPIs, in order:

Pac Ten .6146
ACC .6053
WCC .5734
Big Twelve .5719
Big Ten .5667
SEC .5593
Big East .5531
Ivy .5522
Big West .5431
Mtn West .5342
Colonial .5302
Conference USA .5257
Atlantic Ten .5146
Southern .4931
Horizon .4820
Mid American .4798
Missouri Valley .4692
Sun Belt .4686
Atlantic Sun .4685
Metro Atlantic .4629
Big Sky .4612
Big South .4609
America East .4589
Patriot .4550
WAC .4509
Northeast .4485
Southland .4348
Summit .4346
Ohio Valley .4187
Independent .4174
United .3968
Southwestern .3408

If you want to see each team's RPI based only on inter-conference games, check the cpthomas post at 4:59 pm today, at the following link: http://www.bigsoccer.com/forum/showthread.php?t=650592

I'm looking over the data related to the strength of schedule elements of the Pilots' RPI, trying to guess where that part of the Pilots' RPI will go from here. I figure all of us can make educated guesses about what the ultimate records of UCLA, USC, Oregon, and Washington are likely to be. So, I've been looking at where our other non-WCC opponents are likely to go from here, based on how their out-of-conference RPIs compare to the other teams in their conferences. Here's what I see:

Northern Arizona has the best out-of-conference RPI in Big Sky. They're a possibility to win it. So, I'd look at their winning percentage at the end of the year to be good. A + for the Pilots.

San Diego State is in the middle of Mountain West, which is looking like a tough conference. Let's hope they end up 50-50 in the conference. A slight - for the Pilots.

Colgate is in the middle of Patriot League, which is a fairly weak conference. Again, Let's hope they end up 50-50 in the conference. A slight -.

Kennesaw State is #2 in Atlantic Sun. They're a possibility to end up #1 or #2 in the conference, so a likely strong winning percentage. A + for the Pilots.

Rutgers is #2 in Big East. They're a possibility to end up at #2 in the conference (to Notre Dame), so a likely strong winning percentage. A + for the Pilots.

Penn State is #2 in the Big 10. They're a possibility to end up at #1 or #2 in the conference. Right now, they're 6-5 (I think), so I'm expecting their winning percentage to increase significantly. Most likely a big + for the Pilots.

Florida is #1 in the SEC. They're a likely #1 or #2 in the conference. A + for the Pilots.

What this tells me is that the Pilots are positioned to maintain high strength of schedule numbers through the balance of the season, notwithstanding that the WCC is a little off. Further, the conference RPI averages indicate the WCC is not as far off as we might have thought -- or, perhaps, there is more parity everywhere so that almost all strong conferences (except the Pac 10 and ACC) are not as dominant as they once were.

So, if the Pilots can win out -- which will take a lot of determination -- then their non-conference strength of schedule numbers look likely to balance out any decline in numbers due to the weaker-than-usual winning percentages of the other WCC teams. But still, let's also hope that the other WCC teams have good results this coming weekend