Thank you! How will it work next week when WP2024 (name change soon) has 4 squad members? Any squad with less has no chance by design.
@clark will you be able to share the scores dataset before Thursday?
Many thanks for putting this together and happy Labor Day all.
Good questions, let me think about balancing this out for next week, I have a request for another participant. In a “perfect” world (where Clark doesn’t have to think too hard), we can hopefully create another squad of 4.
Otherwise, we’ll probably need to figure out some way to normalize squad scores based on team size. In our first round, everyone who made picks scored within 10% of each other.
–
RE: dataset, I can probably just dump the data into a CSV if that works for whatever you’re trying to do.
Do you want the scores from this weekend or are you looking for the list of games for next weekend?
The scores of this weekend and hopefully a cumulative file in future weekends.
I suggested how to adjust the ‘win’ and ‘loss’ selections. My assumption is that you don’t have a single user that got those wrong. You want to add more variance in those categories. My suggestion is to have a multiplier on ‘upsets’ or based on ‘diff. rank.’ That is, more points if I one bets that #9 bears #10 versus #1 beats #20.
If you tweak those categories the variance will be larger and the results over multiple weekends will start to reflect skills versus luck.
I am sure there are many more ways to increase the variance.
Will update this periodically, please flag if there are inaccuracies.
I omitted any of the TBD games where both opponents were not marked as well as some schools that were playing non-NCAA opponents. I was also not opinionated on team1 vs team2, it mostly came down to which team I started parsing first.
–
@wp2024 Agreed that rank would provide more variance but I’d need to figure out a way to crunch the rankings out of a matchup as well. I also think the next couple weekends are going to have less variance in general as the game inventory this weekend was huge (looks similar-ish next week).
As we move later into the season, I think there could be another layer where you might only have 6 of your 8 teams playing. I’ve been toying with the idea of adding a wrinkle for win/loss prediction where you *lose* points if you get it wrong, because it should be your most stable point getter.
Reminder to get your picks in for this week, picks close tomorrow at 4PM PT (Iona plays La Verne at that time)
This has been updated with the weekend’s scores and squads should be finalized as well.
Big shoutout to WP2024 Squad who really pulled away on this round by scoring 211 points!
Mod Squad trailed with 142 points and we had a couple of solo participants put up respectable showings.
–
We had several ties for most points in a selection with 20 (all on points scored).
Anyone with Johns Hopkins might regret keeping them inactive as they beat Washington and Jefferson with a score 31 to 11!
–
As a straw poll, do y’all have preference on how we should normalize squad scores against each other?
As a default, I was just going to compare average scores but I’m wondering if anyone has a more clever idea.
How many members ‘Mod Squad’ has?
I need to think about the math of normalizing given a constrained selection each member of a larger squads started with.
Another solution given the variation in sizes and only 4 squads is to continue as individuals. I like the concept of squads but might be complicating the tally of the game at this point.
We have 4 on our squad.
If everyone is reasonably comfortable with merging all of the squads and competing as individuals, I’m fine to do it. TBH, it’d make the weekly recaps easier cause I wouldn’t have to do ~~mAgiC QueRies~ against the database 
–
If there’s objections, please feel free to speak up here or message me directly to voice concerns / opt-out.
Otherwise, I can merge us all into a big happy league in the next couple days.
I am good with it! Still learning, next week I should do better. I thought picking UCLA for points against meant the team playing against UCLA… UCLA won 21-6 but I had it backwards and got 1 point.. All good made changes for next week now that i fully understand.
With so many teams doing the MPSF invite, how will that week work? Have not seen any schedule or bracket yet, hope it comes out soon.
RE: MPSF Invite
I was thinking of doing a “mini-game” because I think we’d struggle to keep up with picks as the schedule updates based on results.
I’d have to look at the schedule to see if this is feasible but I think we can do something where each player picks a single team that’s participating and we base it off goal differential for the weekend.
ie, if someone picks UCLA and they win 18-10 (+8) and 12-4 (+8) the first day, you’d get 16 points.
The next day, they might win 12-10 (+2) but then lose to USC 
6-10 (-4), you’d get -2 points for a total of 14 points.
-
If any of y’all have the schedule, I can look at setting it up or take suggestions.
This year, there are 16 teams in the MPSF invite field. The last time there were this many teams was in 2023, when it was again hosted by USC and UCLA. The schedule likely won’t come out till next Wednesday, as it is based off that week’s CWPA poll. Each team will play 2 games within their group and then do 1-4, 5-8, 9-12, and 13-16 classification. The first group will be 1, 8,9, and 16. Second, 2, 7, 10, 15. So on and so forth. Highest plays lowest seed and winners play for first in group. The notable teams not playing in it are Fordham, San Jose State, and Pacific. For example, if the tournament was held this past weekend, the first group would be UCLA, Davis, Pepperdine, and Claremont-Mudd. Second group would’ve been USC, Princeton, UCSD, and Redlands.
There’s 10 users who are active in fantasy right now, I think the single team pick is feasible.
I’m trying to figure out if there’s a way to reward picking a non-top 4 team. I want to avoid a Bills vs Chiefs coin-flip situation of ‘22 where “first pick probably wins”.
Maybe something like, if you pick up the 8th seed, you start with 8 bonus points?
I did the same thing with the CBU 5 - 21 loss to UCLA and got 1 point.
@clark I am good with doing individual, especially as my pathetic points total is dragging our squad down!
Although it will be like spilling out the ‘GSW’ of Fantasy WP, I am good going solo. Heads up, WaPo is the GOAT and the one to beat.
Would we be able to redraft our teams to level the playing field? Let’s treat 2025 as a “dry run” for Fantasy WP to iron out all the wrinkles.
Yes, this is all good.
I’ll start merging everyone into a single league.
Timeline is something like:
A quick caveat that will change strategy: since we’re going to have more people, each player will get less teams (likely less than 5). This means that you will need be more strategic on which scoring format you want to use . If you’re a water polo sicko, you can do some schedule analysis to figure out your optimal pick order based on available game inventory.
–
A couple other notes:
Thanks - I have really enjoyed playing so far (although a little disappointed to lose my points after getting very lucky over the first two weeks 
)
I actually really like the idea of only having 5 teams for the five categories. It will make you have to think about where to put them each week. It would be fun if you could trade teams with other managers or even go to a waiver wire to swap a team etc. Maybe this is for 2026.
Quick clarification:
I have 53 teams listed so if we get more than 10 players, we’ll get 4 teams each instead of 5.
There’s also not a guarantee that teams will play every week, so there’s def going to be times when schedules determine how many matches you can actually select.
I see two options here:
What about 4 categories and 4 teams per squad. That will allow 13 squads
(sorry if I’m using squad incorrectly - that name confuses me for some reason).
Yes, that’ll work.
I’ll wait until tomorrow night to see if we have any late-joins. Based on number of users playing, I’ll get us the most amount of teams.