Arch madness seeding

[Cdizzle]

Based on what I've seen recently....

ISU will play at noon. WSU will play at 6.

[Hacksaw]

Here is today's update:
Tip of the cap to you sir, that is pretty awesome that you do that. Question for you, does it have the tiebreakers built into it? Like if ISU ties for 1st, they get it...or if UNI ties for 3rd, they get it?

This does have all tiebreakers built into the projections. Correct on the ISU assumption, they definitely control their own destiny for that #1 seed, but Sagarin gives them less of a chance to run the table than WSU... with each win they pick up, those odds of #1 would increase.

[sivert]

Based on what I've seen recently....

ISU will play at noon. WSU will play at 6.

This will most likely be determined in Cedar Falls. Barring a bizarre letdown where GM tries out the motto Play Nicely, WSU should win out the season. (Though, Go Panthers on the 18th!)
ISUr will be at UNI on the 25th. I think we have a good chance in that game.

[Hacksaw]

Out of curiosity, I did some quick probability calculations on whether IlSU will end up with at least 1 more loss than the WSU (since a tie or more WSU losses means IlSU gets the 1 seed). I ended up with a 50.9% chance that IlSU has at least one more loss than WSU using the final record probabilities from rpiforecast.com.

Note that it has been a while since I've done probability calculations and might have made a mistake.
EDIT: I just realized that I did. The probabilities on there for WSU haven't been updated to account for their victory last night. I tried to correct for that and I now get a 52.5% chance that WSU has one fewer loss than IlSU.

Your new math looks similar to mine. In the Monte Carlo model, there is a feature where I sensitize RPI and allow it to float with a set standard deviation from where it each team is currently at. This is more important among the teams in the middle of the pack where shifting of RPI rank is much more frequent. The RPI standard deviation I had the model set at was probably a bit high for this late in the season (for the teams near the ends of the RPI curve). For instance, and agreeing with your calculation, if I drop it to no shifting in RPI, I get 52%-54% odds for a WSU one seed.

Many thanks though - this is a great point and an enhancement I will plan to add to the model. Each team's RPI should float more precisely based upon simulated wins/losses, which in theory should create narrower RPI ranges for the top and bottom teams.

[UNIFanSince1983]

After starting 0-5 we looked almost certain to be a PIG team for the first time under Jacobson. Now we only have a 2% chance of being a PIG team. And currently a 46% chance of being the 3 seed. Now considering we do have a game at WSU and home against ISUr left I am not holding my breath for the 3, but I think Top 5 is probably likely again at this point. Crazy season.

COY

I am sure you are being sarcastic, but if he truly was COY we wouldn't have started 0-5 in conference play to being with

[Stickboy46]

After starting 0-5 we looked almost certain to be a PIG team for the first time under Jacobson. Now we only have a 2% chance of being a PIG team. And currently a 46% chance of being the 3 seed. Now considering we do have a game at WSU and home against ISUr left I am not holding my breath for the 3, but I think Top 5 is probably likely again at this point. Crazy season.

COY

I am sure you are being sarcastic, but if he truly was COY we wouldn't have started 0-5 in conference play to being with

Nah, He's just following the formula of the last COY. Set the bar low so you can "overachieve" and win COY. UNI doesn't suck as much as they did a month ago .. Must be great coaching!!!

[BirdsEyeView]

Based on what I've seen recently....

ISU will play at noon. WSU will play at 6.

This will most likely be determined in Cedar Falls. Barring a bizarre letdown where GM tries out the motto Play Nicely, WSU should win out the season. (Though, Go Panthers on the 18th!)
ISUr will be at UNI on the 25th. I think we have a good chance in that game.

I just want to be healthy. A healthy ISUr team easily wins all remaining regular season games...unfortunately we are not.

Edge to WSU

[AndShock]

If ISUr @ UNI's o/u is anything more than 100 points I am liquidating all assets and betting the under. ISUr will win that game by 5 but it might as well be 55. I feel like every UNI game gets more and more boring as the season goes on.

[Wufan]

I am sure you are being sarcastic, but if he truly was COY we wouldn't have started 0-5 in conference play to being with

Nah, He's just following the formula of the last COY. Set the bar low so you can "overachieve" and win COY. UNI doesn't suck as much as they did a month ago .. Must be great coaching!!!

He still set the bar too high to start with...see Drake.

[LanceShock]

Your new math looks similar to mine. In the Monte Carlo model, there is a feature where I sensitize RPI and allow it to float with a set standard deviation from where it each team is currently at. This is more important among the teams in the middle of the pack where shifting of RPI rank is much more frequent. The RPI standard deviation I had the model set at was probably a bit high for this late in the season (for the teams near the ends of the RPI curve). For instance, and agreeing with your calculation, if I drop it to no shifting in RPI, I get 52%-54% odds for a WSU one seed.

Many thanks though - this is a great point and an enhancement I will plan to add to the model. Each team's RPI should float more precisely based upon simulated wins/losses, which in theory should create narrower RPI ranges for the top and bottom teams.

Glad to help. Love the work you put into this. It helps see how things are likely to end up.

Only looked into it because I thought it was odd that WSU would be at 60% when IlSU will have the tiebreaker. 52-54% still seems a little high, but it is what the Sagarin projection indicate. I hadn't even realized until just a week or so ago that the tiebreaker had changed to RPI from noncon RPI SOS.