RPI and series sweeps

How does a 3 game series affect the teams RPI?
Is your RPI rating based on each game you win against a team in a 3 game series? Does a teams RPI rating increase after after each win? Does it increase based on the series as a whole?

Yep. Each game matters. LSU’s record and their opponents’ records plug into the formula three times, once for each game (same goes, or went, for a home and home in basketball when they still used RPI). It’s treated as three separate games, so if you win two out of three, that goes into the calculations.

Cliff’s Notes RPI formula, minus some minor tweaks:

25% of your RPI is your own winning percentage, with adjustments for home, road and neutral site. If you win a home game, it’s worth 0.7; if you lose a home game, it’s worth 1.3. Those numbers are reversed for road games. Neutral site games are straight up. Sometimes you’re not sure if they count a game as home or neutral. Since UCA is closer to NLR than Fayetteville is, I’m pretty sure that game will count as neutral, whereas in the past it might have been considered a home game. So with those adjustments, our “record” for RPI purposes is 22.6-7.3, according to Boyd Nation’s RPI simulation page. If we had lost one game to LSU, we’d be 21.9-8.6, since a home loss is worth 1.3.

50% is your opponent’s winning percentage, minus the games against you. LSU plugs in at 23-9, its record in all other games except against us.

And the other 25% is the opponents’ opponents’ winning percentage. So if you play and beat a team with a good record who has been killing cupcakes, their record will help you, the cupcakes’ records will hurt you.

Sounds to me like RPI is basically worthless.

1 Like

So, are you saying the same RPI values remain for each win against the same team for each game they lose.

RPI is intended for tournament selection and seeding purposes. Thus its value – or its worth – comes at the end of May when the committee is picking 64 teams. In mid-April, it doesn’t mean much.

They will be looking at our entire resume, which includes three games against LSU (maybe more at Hoover), four against UIC, etc. RPI helps them determine who achieved their success against a better schedule. That’s why 75% of RPI is what teams did when they weren’t playing you.

If you’re looking at RPI to identify the best teams, it doesn’t. It is largely a strength of schedule metric, which is why they stopped using it for basketball. Will the NCAA adapt NET for baseball? Wouldn’t surprise me

Let’s make up some numbers for LSU to assess what they contribute to our RPI.

We know that our RPI “record” is 22.6-7.3, which is a .755 winning percentage. That’s the 25% which comes from our record. Tennessee’s adjusted record is 27.3-3.3 (.892) and thus that 25% for them is huge right now.

LSU is 23-9 (.718) against everyone not named Arkansas, so their contribution to our opponents winning percentage portion of the RPI (which is 50% of the total) is .718, x 3. Of course, every other team we play makes their contribution to that 50% through the regular season and conference tournament, which could be 60 games or more, but probably more like 57 this year. Obviously, if you play tough teams and beat them, like we did last year, this is where that shows up. It’s better to beat LSU than it is to beat UAPB for that reason.

Let’s say that in those 32 games against everyone else, LSU’s opponents have a total record of 465-495. Thus the LSU contribution to the OOWP is .484 (465 divided by 960) x 3, because again we would have played them in 3 of the 50+ games of the season.

Good info thanks.

RPI has always been A tool, not THE tool. Another data point, a way to compare teams who didn’t play each other.

For example, in the current RPI, Virginia is slightly above us (18th to our 20th). If the committee were meeting on a Monday night in April instead of Sunday, May 29, they might compare the Hoos to us to see who might be worth receiving a regional at home. RPI would be part of that consideration, but so would the info on the so-called “nitty gritty” sheets, which Warren Nolan also simulates. Although UVa’s RPI is better, their road record is actually below .500. And they are 10-8 in the #2 conference (again according to RPI) while we are 11-4 in the #1 conference. Our NC strength of schedule is actually a virtual deadlock according to Nolan, but their NC RPI is better because they have a better record (17-1 to 17-3). We have two bad losses (Illinois State and Southeastern Louisiana), they have one (a conference loss at Duke).

Kinda reminds you of Vandy and Hofstra in hoops, huh? Although Vandy eventually worked its way off the bad loss list.

If Stanford starts to win more games, does it help our RPI since they beat us?

Any game won by a team who played us helps a bit. That goes into that 50% for opponents’ winning percentage. But the same applies for UAPB or any other cupcake. They just won’t win as many as Stanford will.

Yes - this is a common misconception about the RPI - it’s your overall W/L record that accounts for that first part of the RPI Jeff described. And then, the overall W’L record in the other two categories as well. It’s NOT so much that you just beat the #5 RPI team, per se. If you have a 33-12 record, it doesn’t matter WHICH of those teams you beat on the way to 33-12, with regard to how that affects the first of the 3 values.

It DOES help to beat the better teams, because you are increasing your winning percentage at the same time as you are adding their WP to the second category. But, had we been swept by LSU, and won 3 earlier games (against anyone of the other teams we lost to) instead, we’d have an identical W/L record against teams that have the same W/L record as they have today. (Or course, the W’s and L’s would need to have come at home, since that’s where we beat LSU, since home W’s count differently than Road W’s, as Jeff also pointed out).

All of that said, when the selection committee looks at a team’s “resume” at the end of the season, they also look to see how you’ve done against better teams, Conference foes, at home and on the road, etc. It isn’t JUST the RPI. It’s one of many metrics.

Still, the correlation between having a top 10 RPI and being selected as a national seed (top 8 seeds) is real.

Once you get into SEC play, the RPI tends to take care of itself, if you win a lot. Just win a lot and you won’t worry. And, the top 8 national seeds are going to be teams that win an awful lot. I don’t ever recall a national top 8 seed that didn’t win a lot. I know that’s over simplification, but it works out.

1 Like

Clay how many wins do you think the hogs need to be a top 8 national seed?

I will also add this. I understand DVH’s attitude toward playing the other in-state schools, and by and large I’m fine with it. The only real issue with it is that it’s a LOT of games with teams that are really, really low in the RPI, and that is an issue. A game or two is no big deal.

But when you play EIGHT games against teams whose best RPI is 192 (and most are in the mid to high 200’s), that’s enough to make an impact on our RPI that could mean the difference between hosting or not, be it a Regional of a Super, in a given season.

I think it may be wise to re-think the strategy and rotate these 4 other schools to play 2 each season. We can better absorb that hit to RPI while still playing them all frequently. In the meanwhile, if we can get a team like Oral Roberts to play instead of the other 2, that would be much better. They are currently 128, which isn’t great, but (a) they’re usually better than that and (b) even at that ranking, they are MUCH higher than our in-state foes.

The problem is who will come here this time of year?

1 Like

I’ve been advocating for us to build a dome for years.

All it takes is money, right?

More than most of the rest. Ha!

That’s a good place to start. If not there, at least replace the two Grambling State games each season. They’re always in the 275-300 range.

:bell: :bell: :bell:

Someone was posting on FB that they would like to see Arkansas play OU, OSU, etc. mid-week. That’s fine, it would be fun to see. BUT those schools must agree to those games. Once a school has played a game at Baum-Walker, they are not often eager to repeat the experience.

1 Like

I understand, and agree. It’s kind of like the folks who still complain about our N/C schedule in football and want us to play Southern Cal, Michigan, Florida State and Oklahoma State instead. It’s just not realistic.

But in recent years, we’ve played Memphis, ULM, Oral Roberts…none of which are powerhouses (although Oral Roberts has had some very good teams off and on), but they are teams in the 150-180 range of the RPI, not 250+. Also, we have in the past had some nice series in March and April vs. teams like Texas and Texas Tech. It’s clear to see the difference in our SOS (and, therefore, RPI) if you swap our 5 or 6 games from this year’s in-state schedule with games like those.