.png)
Sanity Check
As I was going to sleep last night, my thoughts turned to volleyball as they often do. I started thinking about these schedule analysis articles and it occurred to me I could do a sanity check on the numbers by looking at the 2024 teams. I mean I’m spitting out a lot of numbers but do they mean anything? (The short answer is that they never mean anything. Only the matches mean something.) I then worried that the numbers wouldn’t work with a 2024 team and everything I’d written was trash. It was at this point I turned over and started thinking about leprechauns and unicorns and went to sleep.
This morning, I went down the list of 2024 teams and settled on WashU as my test case. I would do the same analysis using the 2024 numbers on their 2024 schedule. Thankfully, finding the 2024 schedules is much easier than the 2025 schedules. Here is their stat line:
These numbers are based off of their regular season and do not include the UAA conference tournament. The interesting thing is that I actually used the conference tournament matches to start with and a measure of how tough/good the UAA is that WashU went from 9 maximum losses to 11 just by playing three additional games. (They won against NYU and CWRU before losing in the finals to Emory.) This is a huge advantage for any conference that is as deep as the UAA was last season.
Back to the above numbers and you can see they had a good maximum NPI. It would place 6th on the teams I’ve looked at for 2025. Remember, WashU finished in 11th place according to NPI in 2024. I think this maximum NPI would have gotten them to a Top 8 spot but they just lost too many matches. Their “Loss n” would have been concerning as this means if they had lost to all of the teams above the cut-line, they would have finished without an at-large bid. In reality, this didn’t happen and WashU only lost 5 matches during the regular season. What I’m thankful here is that their actual losses didn’t exceed their maximum loss number, which would have made everything I’ve done pointless. It probably also means I should do another team closer to the cut-line.
A little more on WashU and what actually happened in 2024. Their losses were to Hope, Northwestern, Emory, CWRU and Millikin. That’s four of the elite programs they played and one (Northwestern) in the 55-60 range. Winning half of your elite matches definitely pushes up the NPI and you can see why, despite 5-losses that they were so close to the Top 8.
Our last team above the cut-line in 2024 was Centre so they make sense to check out here, and I’m not fibbing by writing that I’m a little nervous about this.
Whew! Centre went 21-5 during the regular season so their actual losses matched their maximum losses. Of course, I assume the best opponents will deliver the losses but that wasn’t the case here. They lost to La Verne, Transylvania, Emory, Asbury (!) and Berry. That’s 2 of the elite teams, 2 in the 55-60 range and 1 in the 50-55 range. Their elite win was against CMS and they beat 5 teams in the 55-60 range. Winning those elite matches can cover up some blemishes and that certainly helped Centre in 2024. It also helps to have a higher “Loss n”, in my opinion. I think that also covers up for lower losses.
Just looking at this schedule in 2024, and it was a sweet NPI schedule. I like when you can double-up your ranges and Centre went from 3 to 7 and then increased to 10. Not too bad. Their “Loss n” was basically 60, which is good, and their maximum losses was two more than their elite opponents. Their 2024 schedule allowed for some wiggle room and Centre used all of that wiggle. I’m feeling really good about these schedule analysis articles now! I can’t believe you all doubted me!
Just to really test things, I went ahead and added in the SAA conference matches and this is where things went a little sideways. The “Loss n” number climbed to 60.20 but the maximum losses didn’t change. If I added a 6th loss, the Centre NPI dropped to 59.23 and below the cut-line. Of course, Centre did take that 6th loss to Berry in their conference final. That loss was slightly worse than the simulated loss I gave them. How did Centre go from a simulated 59.23 up to an actual 59.62? It’s in the makeup of the wins and losses. I simulated a CMS loss but they took an actual win here. I also simulated two losses to Transy but they split that series. Taking “bad” losses hurts but taking “good” wins is huge especially when pulling down a Quality Win Bonus (QWB). One of the NPI improvement suggestions (from “Ned”) is to increase the QWB multiplier to give more credit to teams beating the best teams. We can actually see why that would work here. (To be clear, increasing the multiplier will favor teams playing tougher schedules and earning quality wins. It would eliminate teams sneaking in by gaming the formula.)
In the end, I think my numbers are solid but we’ll see some crazy as we get closer to that cut-line. I think that makes sense. After all, Centre made the tournament because Smith beat MIT in the last match of the regular season. I mean, that was a little crazy.
Comments