Wow. We jumped 17 spots in the RPI yesterday.

[DawgatAuburn]

From 53 o 36. Bama dropped from 15 to 22.

<font size="2"><br style="font-style: italic;"><span style="font-style: italic;">eta the text of the post, which I typed the first time but it disappeared</span></font>

 

[DawgatAuburn]

From 53 o 36. Bama dropped from 15 to 22.

<font size="2"><br style="font-style: italic;"><span style="font-style: italic;">eta the text of the post, which I typed the first time but it disappeared</span></font>

 

[CadaverDawg]

can someone explain something to me...

I saw where one of the rpi guys projected us to finish the year around where we are now in the RPI, but he also had USM projected to finish in the top 20 I think (i don't remember the exact spots he had either of us). How is it that USM can hold their rpi ranking while playing some terrible teams in CUSA? I mean, when you look at the two schedules right now, our overall schedule will be far and away tougher than USM's. I'm not bitching, I'm just seriously wanting someone to explain what I'm missing. For RPI to be so important to the committee, it just doesn't make a lot of sense to me.

 

[drt7891]

Memphis and UAB are usually regulars in the NCAAT. Also, the RPI molds and solidifies as the season continues. As teams we beat continue to win, our RPI increases. It also helps as teams that beat us continue to win. Same applies to USM. USM has a VERY respectable team this year and USM has a good chance of winning most of, if not all of their remaining games. ANY team with 4 to 6 losses in the regular season is going to be very high in the RPI standings, and USM will probably fall in that category. The RPI is merely a statistical indicator for the committee to use when selecting teams, so don't knock them too hard for a high RPI. <div>
</div><div>With all that said, I don't expect them to be chosen (if they are) much higher than a 12 seed. </div>

 

[CadaverDawg]

but I still don't get it. I mean, I understand how you're saying RPI works, but it still doesn't show me how USM's can be so high. They have not played as strong of a schedule as we have, and the teams they have played aren't all winning (for instance: Ole Miss). They seem to be a respectable team, and may even make the dance, but I just do not see any explanation as to why their rpi could be that high....or stay that high. They don't even have a conference team ranked, but they are supposed to stay in the top 20? How? That would go against your argument right?

Just curious because it bugs me

 

[drt7891]

their RPI goes up... regardless of how "bad" those teams are. Here is an example formula from Wikipedia (not a credible source, I know... but this illustrates how it works).

RPI = (WP * 0.25) + (OWP * 0.50) + (OOWP * 0.25)
where WP is Winning Percentage, OWP is Opponents' Winning Percentage and OOWP is Opponents' Opponents' Winning Percentage.
The WP is calculated by taking a team's wins divided by the number of games it has played (i.e. wins plus losses).

So as USM wins, their RPI increases. As their opponents win games (regardless of how "strong" they are... like how the BCS formula takes into account Strength of schedule), that part of the formula increases
(this is the part with the most weight... which for USM, holds true). As the opponents of their opponents win, that increases their RPI even further. USM has beat teams that are winning, albeit in ****** conferences. Keep in mind, RPI does not consider "how good" teams are... so a win over a top 5 RPI team weigh the same amount as a win over a bottom 300 team. Also, WP takes into account road wins, home wins, and neutral court wins (since they weigh differently from one another).

Don't put so much stock in RPI, because as the season goes on, the RPI will begin to really take shape... and it is also not the only thing the committee considers in the selection.... like I said, it is only a statistical indicator... and does not work the same as BCS rankings... which is probably what you are trying to compare it to.<div><p style="margin-top: 0.4em; margin-right: 0px; margin-bottom: 0.5em; margin-left: 0px; background-color: rgb(255, 255, 255); "></p></div> <div>
</div><div>ETA: RPI takes into account how much your team wins and how much your opponents win, not necessarily SOS... that's what I meant when I said RPI does not take into account how "good" teams are. </div>

 

[maroonmania]

you would THINK that USM's schedule is weaker given they have played no ranked teams or even big name teams BUT according to real time rpi their schedule strength is 58th while ours is 89th. They are benefitting by having some real no names on their schedule be high RPI teams including a Murray State team that is still undefeated. But look at these on their schedule:

Denver L - RPI #48
Colorado St. W - RPI #23
NM State 2 Ws - RPI #70
OM W - RPI #57
Murray St. L - RPI #28 (they are a good team this year)
Memphis L - RPI #29
South Floriday W - RPI #96

Tomorrow they play Marshall, yes Marshall, who is RPI #33 so even if they lose it won't affect their rating much. Their best win though this year though is Colorado St. but yet they maintain a Top 20 RPI, that tells you how screwed up the system is.

 

[drt7891]

Those teams you pointed out will not all have that high of an RPI as the season wears on. The RPI is all over the place until middle of February or so when conference play really gets underway. I imagine if USM finishes with 4 or 5 losses, they will stay in the top 30, but if they lose a few more games, that can easily drop as much as 20 spots.

 

[engie]

is playing on the road against good teams a lot. The RPI actually assigns different weight to home/road wins. If memory serves me correct, a home win only counts as 0.6win, whereas a home loss counts as 1.4loss. Losing to a GOOD team on the road often helps your RPI much more than beating a bad team at home. Let's remember that 75% of RPI is essentially SOS.

USM has played 8 games vs rpi top 100, we've played 6. They've played 5 at home and 3 on the road, won a top25rpi game on the road, as well as 3 rpi 51-100 wins at home. They have NO bad losses according to RPI. We have 2 top25rpi wins at home, a neutral site top 25 loss(counts even), as well as a road and home top 100 loss, and a neutral site top 100 win. We also have NO bad losses according to RPI.

Basically, the "block" that's easy to experience toward USM is because we tend to look at the team names they are playing and apply bias...if you strictly look at the RPI numbers, which eventually work out to be completely unbiased and fair, it's easy to see why they are ranked more highly than us. That said, I expect them to drop as the season goes along, regardless of the "predictors", because their SOS gets much worse in conference...

 

[drt7891]

and what makes them up, but you make a very good point. The most important thing is to win at home... and grab some wins on the road if we can.

 

[CadaverDawg]

So unlike football, it is actually better to play good teams and lose, than to play average teams and win. Interesting. Does that mean we got more credit for losing to Baylor than beating Arizona basically?

 

[maroonmania]

because the RPI formula only gives a 25% weight to your winning percentage and a 75% weight to your opponent's winning percentage and your opponent's opponents winning percentage.

 

[engie]

So unlike football, it is actually better to play good teams and lose, than to play average teams and win. Interesting. Does that mean we got more credit for losing to Baylor than beating Arizona basically?

Actually, yes. Since they were both neutral site games, they actually go as 1.0 win and 1.0 loss, so the implications aren't as drastic as they "could" be in either case if it was a home/road game. Since we are rated higher than Arizona by a smaller margin than we're rated lower than Baylor, the Baylor loss is actually helping us slightly more than the AZ win, by a very small margin, due to SOS. This gets much harder to figure when it's not a neutral site game.

IMO, it's REALLY dumb for us to play Baylor in Dallas. We get a road atmosphere, without getting credit for a road game, so it goes as 1.0 loss instead of 0.6 loss in the scheme of things. Any neutral site game we schedule in the future needs to be at a truly neutral site...

 

[drt7891]

Got murdered in front of a Cincy home crowd, but it counted as a neutral court loss instead of a road loss. There should be limitation on that rule to avoid ******** like that because I was pissed. It didn't really affect the season because we won the SECT that year, but it could have cost us in seeding. It's stupid to have teams pay a penalty to go on the road in a situation like that and have it count as ESSENTIALLY a home loss. Stupid.

 

[MSUDawg25]

Keep it up, or start posting ridiculous ****. Doesn't matter. Makes it interesting either way.

 

[QuaoarsKing]

RPI won't take shape for another few weeks. Right now, there just isn't a large enough sample size to get a good ranking, so a few teams end up misrated, and then the teams that play them get somewhat misrated (but less so...) and so forth. Give it a few more weeks and it will mostly sort itself out.

 

[engie]

I don't mind playing the early-season tournaments at true neutral sites, but we should never EVER play a game like that Cinci game or the Baylor game, at a "neutral site" in hostile territory. It's actually a SMART move on Baylor's part to schedule like this. Instead of being 0.6win, they get 1.0win, so it actually helps their RPI more than just beating us at their place, and they get to keep their home atmosphere. The more I think about this, there HAS to be an NCAA rule regulating the amount you can do this, because you could easily manipulate the heck out of the RPI system by playing "home games" at a neutral site a few towns over, and this would be a lifeline for mid-major programs.

Calculating the difference between losing to Baylor on neutral court vs away:

RPI = (WP * 0.25) + (OWP * 0.50) + (OOWP * 0.25)

WP = total (weighted) wins/ total (weighted) games
WPmsu = (10.6/13.6) = .77941
BaylorRdWPmsu = (10.6/13.2) = .80303
WPmsu25% = .19485
BaylorRdWPmsu25% = .20076
difference of .00591 points in the cumulative WP column

Current RPImsu = .59560
Baylor road game RPImsu = .60151

According to Warren Nolan, that would jump us from our current 41st RPI, all the way to 35th. So, that "neutral-site" road game we played actually cost us SIX positions in the current RPI rankings vs a true road game. That's the difference between a bubble team making the NCAA's and sitting at home in March!

 

[engie]

is minimized over the course of an entire season due to the increase in total volume of data. We've played 13.6RPI games thusfar, with 13.4RPI games remaining, essentially meaning the season is at it's halfway point. Assuming the "Baylor effect" is double now what it will be at the end of the season, which it essentially will be, we still gave up ~ 3 RPI spots by allowing ourselves to play in that "neutral court" setup vs making it a true away game...

Without punching numbers, it probably ends up helping their RPI ~1-2 spots as well...

 

[drt7891]

That's what pissed me off so much about the Cincy game in 2008 because it's absolutely ABSURD to think that there is a true "neutral court" in the 17ing school we are playing's hometown. Same thing applies for the Baylor game this year. The folks who came up with the neutral court rules are idiots to think there is a true "neutral court" when you are 98 miles from one school compared to 530 miles from another. It's like us inviting them to come play us in a game at Starkville High and trying to claim it is a "neutral court." Complete and total ********. <div>
</div><div>Maybe a "scaled weight" for neutral court would be fair. So think about this... a home court win is .6 of a win and a road win is 1.4</div><div>wins. So do a "scaled weight," of sorts... do a ratio... so let's do the math...</div><div>
</div><div>The difference in a road win and a home win is .8 of a win. The number of miles that separates the schools is 98 for Baylor and 530 for State. So with this info, Should Baylor win (which they did), they should get credit for a home win, 0.6, PLUS 15.6% of the .8 difference... so .725 win. Had we pulled out the win, we should get the same... home win, 0.6 of a win, PLUS 83.4% of the remaining .8 of a win, or 1.275 wins. The same should go for the loss. </div><div>
</div><div>I know this seems complicated, but I literally set this up in 2 minutes in Excel... this can balance out the difference in "Home Court Advantages" in neutral court settings. Calculate the difference in mileage between the schools, and grant each school a weighted win based on how far they had to travel to the neutral court site. </div><div>
</div><div>Just an idea. </div>

 

[Seinfeld]

One of the biggest reasons that our RPI is so low is not because of a lack of quality wins or actual losses. It's this...

LA Monroe - (2-16) RPI 334
UT Martin - (1-16) RPI 298
Troy - (5-10) RPI 202
Fla Atlantic - (5-12) RPI 174

We could have beaten those teams by 100 pts each, but they still would have killed our RPI. That is what is so messed up about this whole system to me. For a formula that is supposed to be all about wins and losses, it always seems to truly be more about who you play rather than whether you actually win or lose.

 

[tenureplan]

we would have a top 50 SOS having not played those games and played 4 fewer games than the rest of the teams at this point.</p>

 

[DawgatAuburn]

Or does that count as a home game for us?

 

[engie]

Warren Nolan lists that as a home game for us, but there is a discrepancy between his numbers and realtimerpi's numbers on us, and I bet this is it. In my analysis, i used it as a home game for us, but I believe it should go as a neutral site game, which would explain why realtime has us ranked 36th and nolan has us 41st...

 

[drt7891]

played a quality opponent there that I can remember. *ETA- My math was backwards*<div>
</div><div>If anyone remembers a few years ago when we played MVSU at the Hump, their RPI was in the 290s at the time we played them, and by beating them, our RPI fell 6 or 8 spots. </div>